Pharmacist response to alerts generated from medicaid pharmacy

Transcription

volume eleven • number seven
september 2005
Peer-Reviewed Journal of the
Academy of Managed Care Pharmacy
JMCP
JOURNAL OF MANAGED CARE PHARMACY®
Page
550
Gastrointestinal Bleeding Rates Among
Managed Care Patients Newly Started on
COX-2 Inhibitors or Nonselective NSAIDs
559
Relationship of Glycemic Control to Total
Diabetes-Related Costs for Managed Care
Health Plan Members With Type 2 Diabetes
565
Economic Burden of Anemia in an Insured
Population
575
Pharmacist Response to Alerts Generated
From Medicaid Pharmacy Claims in a
Long-term Care Setting: Results From the
North Carolina Polypharmacy Initiative
AMCP HEADQUARTERS
100 North Pitt St., Suite 400
Alexandria, VA 22314
Tel: (703) 683-8416 • Fax: (703) 683-8417
Volume 11, No. 7
BOARD OF DIRECTORS
C O N T E N T S
■ ORIGINAL RESEARCH
550 Gastrointestinal Bleeding Rates Among Managed Care Patients
Newly Started on COX-2 Inhibitors or Nonselective NSAIDs
Karen Stockl, PharmD; Lori Cyprien, MS; and Eunice Y. Chang, PhD
559 Relationship of Glycemic Control to Total Diabetes-Related Costs
for Managed Care Health Plan Members With Type 2 Diabetes
Sharashchandra Shetty, PhD; Kristina Secnik, RPh, MPH, PhD; and Alan K. Oglesby, MPH
565 Economic Burden of Anemia in an Insured Population
ADVERTISING
Allen R. Nissenson, MD, FACP; Sally Wade, MPH; L. Tim Goodnough, MD;
Kevin Knight, MD, MPH; and Robert W. Dubois, MD, PhD
Advertising for Journal of Managed Care
Pharmacy is accepted in accordance with the
advertising policy of the Academy of Managed
Care Pharmacy.
■ CONTEMPORARY SUBJECT
575 Pharmacist Response to Alerts Generated From Medicaid
For advertising information, contact:
Professional Media Group, Inc., P.O. Box 189,
40 N. Woodbury Rd., Pitman, NJ 08071
Tel: (800) 486-5454 or (856) 589-5454
Fax: (856) 582-7611
E-mail: [email protected]
Pharmacy Claims in a Long-term Care Setting:
Results From the North Carolina Polypharmacy Initiative
Troy K. Trygstad, PharmD, MBA; Dale Christensen, RPh, PhD; Jennifer Garmise, PharmD;
Robert Sullivan, MD, MPH; and Steven E. Wegner, MD, JD
■
DEPARTMENTS
530
Cover Impressions
Nashville Rocks! (2001)
Dave Newman
590
Sheila Macho
Cover Editor
584
Editorials
• Fish Oil for Heart Disease—
Happy to Be in Second Place
Editorial Subjects—In This Issue
and in Previous Issues
• COX-2 Inhibitors: Little or No GI
Protection, Increased Risk of
Cardiovascular Events, High Cost,
and Other Class-less Effects
Frederic R. Curtiss, PhD, RPh, CEBS
Editor-in-Chief
Brian K. Crownover, MD, FAAFP,
Lt. Col., MC, USAF
Associate Editor
• Medication Therapy Management
Services for Long-term Care
Patients: No Road Maps for
Those Trying to Find Their Way
Joshua J. Spooner, PharmD, MS
594
Letter
596
Abstracts From Professional
Poster Presentations at AMCP’s
2005 Educational Conference
620
Managed Care Pharmacy
Residencies, Fellowships,
and Other Programs
Questions related to editorial content should be
directed to JMCP Managing Editor Tamara C. Faggen:
[email protected]; (703) 323-0170.
Manuscripts should be submitted electronically
at jmcp.msubmit.net. For questions about
submission, contact Peer Review Administrator
Jennifer A. Booker: [email protected];
(703) 317-0725.
SUBSCRIPTIONS
Annual subscription rates: USA, individuals,
institutions–$60; other countries–$80. Single
copies cost $10. Missing issues are replaced free
of charge up to 6 months after date of issue.
Send requests to AMCP headquarters.
For article reprints, contact Diana Sholl, Reprint
Management Services, (717) 560-2001, ext. 162;
[email protected]. Microfilm and microfiche
editions of Journal of Managed Care Pharmacy are
available from University Microfilms, 300 N. Zeeb Rd.,
Ann Arbor, MI 48106. Reprint Guidelines are available at www.amcp.org.
All articles published represent the opinions of the
authors and do not reflect the official policy or views
of the Academy of Managed Care Pharmacy or the
authors’ institutions unless so specified.
Joanne LaFleur, PharmD
Associate Editor
JMCP
EDITORIAL
REPRINTS
• Evaluating the Relationship
Between Diabetes Treatment and
Health Care Costs—Measuring the
Atom With a Yardstick?
524 Journal of Managed Care Pharmacy
President: Dianne A. Kane Parker, PharmD, Amgen
Inc., Fallbrook, CA
President-Elect: Steven W. Gray, PharmD, JD, Kaiser
Permanente, Downey, CA
Past President: James R. (Rusty) Hailey, PharmD, DPh,
MBA, Coventry Health Care, Inc., Franklin, TN
Treasurer: Cathryn A. Carroll, PhD, Children’s Mercy
Hospital & Clinics, Smithville, MO
Secretary: Judith A. Cahill, CEBS, Academy of
Managed Care Pharmacy, Alexandria, VA
Director: Elaine Manieri, BSPharm, Caremark Rx, Inc.,
Hunt Valley, MD
Director: Janeen McBride, RPh, MedImpact
Healthcare Systems, Inc., San Diego, CA
Director: Mark J. Rubino, RPh, MHA, Aetna, Inc.,
Hartford, CT
Director: Brian Sweet, RPh, MBA, WellPoint
Pharmacy Management, Grand Island, NY
(JMCP liaison)
Director: Richard A. Zabinski, PharmD,
UnitedHealthcare Corporation, Edina, MN
September 2005
Vol. 11, No. 7
www.amcp.org
Copyright© 2005 Academy of Managed Care
Pharmacy, Inc. All rights reserved. No part of this
publication may be reproduced or transmitted in any
form or by any means, electronic or mechanical,
without written permission from the Academy of
Managed Care Pharmacy.
J M C P E D I TO R I A L P O L I C Y
■■ Editorial Content and Peer Review
All articles and editorials in JMCP undergo
blinded peer review. Letters may be peer
reviewed to ensure accuracy. The fundamental departments for manuscript submission
are:
• Original Research
• Subject Reviews
• Formulary Management
• Contemporary Subjects
• Editorials
• Letters
For manuscript preparation requirements,
see “JMCP Author Guidelines” in this Journal
or at www.amcp.org.
■■ Original Research
These are well-referenced articles based
on original research that has not been
published elsewhere and reflects use of
the scientific method. The research is
guided by explicit hypotheses that are
stated clearly by the authors.
■■ Subject Reviews
These are well-referenced, comprehensive
reviews of subjects relevant to managed
care pharmacy.
■■ Formulary Management
These are well-referenced, comprehensive
reviews of subjects relevant to formulary
management methods or procedures in
the conduct of pharmacy and therapeutics (P&T) committees and may include
description and interpretation of clinical
evidence.
■■ Contemporary Subjects
These are well-referenced submissions that
describe pilot projects or other subjects
that are not intended to be comprehensive
reviews of the subject.
■■ Editorials
These submissions should be relevant to
managed care pharmacy and address a
topic of contemporary interest.
■■ Letters
If the letter addresses a previously published
article, an author response may be appropriate. (See “Letter to the Editor” instructions at www.amcp.org.)
EDITORIAL MISSION AND POLICIES
JMCP publishes peer-reviewed original research manuscripts, subject
reviews, and other content intended to advance the use of the scientific
method, including the interpretation of research findings in managed care
pharmacy. JMCP is dedicated to improving the quality of care delivered to
patients served by managed care pharmacy by providing its readers with the
results of scientific investigation and evaluation of clinical, health, service,
and economic outcomes of pharmacy services and pharmaceutical interventions, including formulary management. JMCP strives to engage and
serve professionals in pharmacy, medicine, nursing, and related fields to
optimize the value of pharmaceutical products and pharmacy services delivered to patients.
JMCP employs extensive bias-management procedures that include (a) full
disclosure of all sources of potential bias and conflicts of interest, nonfinancial as well as financial; (b) full disclosure of potential conflicts of interest by
reviewers as well as authors; and (c) accurate attribution of each author’s
contribution to the article. Aggressive bias-management methods are
necessary to ensure the integrity and reliability of published work.
Editorial content is determined by the Editor-in-Chief with suggestions from
the Editorial Advisory Board. The views and opinions expressed in JMCP do
not necessarily reflect or represent official policy of the Academy of Managed
Care Pharmacy or the authors’ institutions unless specifically stated.
■■ Advertising/Disclosure Policy
A copy of the full advertising policy for
JMCP is available from AMCP headquarters
and the Advertising Representative. All
aspects of the advertising sales and solicitation process are completely independent of
the editorial process. Advertising is positioned either at the front or back of the
Journal; it is not accepted for placement
opposite or near subject-related editorial
copy.
Employees of advertisers may submit
articles for publication in the editorial
sections of JMCP, subject to the usual
peer-review process. Financial disclosure,
conflict-of-interest statements, and author
attestations are required when manuscripts
are submitted, and these disclosures
generally accompany the article in
abstracted form if the article is published.
See “Advertising Opportunities” at
www.amcp.org. Contact the Advertising
Representative to receive a Media Kit.
www.amcp.org
Vol. 11, No. 7
September 2005
Editorial Office
Academy of Managed Care Pharmacy
100 North Pitt St., Suite 400
Alexandria, VA 22314
Tel: (703) 683-8416
Fax: (703) 683-8417
[email protected]
Advertising Sales Office
Professional Media Group, Inc.
40 N. Woodbury Rd.,
Pitman, NJ 08071
Tel: (800) 486-5454
or (856) 589-5454
Fax: (856) 582-7611
JMCP
Journal of Managed Care Pharmacy 527
EDITORIAL MISSION
JMCP publishes peer-reviewed original research manuscripts, subject reviews, and other content intended to advance the use of the scientific
method, including the interpretation of research findings in managed care pharmacy. JMCP is dedicated to improving the quality of care delivered to patients served by managed care pharmacy by providing its readers with the results of scientific investigation and evaluation of clinical,
health, service, and economic outcomes of pharmacy services and pharmaceutical interventions, including formulary management. JMCP strives
to engage and serve professionals in pharmacy, medicine, nursing, and related fields to optimize the value of pharmaceutical products and
pharmacy services delivered to patients.
JMCP employs extensive bias-management procedures intended to ensure the integrity and reliability of published work.
EDITORIAL STAFF
Editor-in-Chief
(361) 749-0482, [email protected]
Managing Editor, Tamara C. Faggen, (703) 323-0170, [email protected]
Associate Editor, Brian K. Crownover, MD, FAAFP, Lt. Col., MC, USAF,
(850) 883-8288, [email protected]
Associate Editor, Lida R. Etemad, PharmD, MS, (952) 833-7036, [email protected]
Associate Editor, Joanne LaFleur, PharmD, (801) 585-3794, [email protected]
Peer Review Administrator, Jennifer A. Booker, (703) 317-0725, [email protected]
Graphic Designer, Laura J. Mahoney, (703) 917-0737, ext.101, [email protected]
Cover Editor, Sheila Macho, (952) 431-5993, [email protected]
Publisher
Judith A. Cahill, CEBS, Executive Director, Academy of Managed Care Pharmacy
EDITORIAL ADVISORY BOARD
The JMCP Editorial Advisory Board is chaired by Marvin D. Shepherd, PhD, Director of the Center for Pharmacoeconomic Studies of
the College of Pharmacy at the University of Texas at Austin. Dr. Shepherd and the other advisers review manuscripts and assist in the
determination of the value and accuracy of information provided to readers of JMCP.
Emily Ann Baker, PharmD, Northstar Health Care Consulting, LLC,
Cumming, Georgia
John P. Barbuto, MD, HealthSouth Rehabilitation Hospital, Sandy, Utah
Eliot Brinton, MD, School of Medicine, University of Utah, Salt Lake City
Diana I. Brixner, RPh, PhD, Department of Pharmacotherapy, University
of Utah, Salt Lake City
Scott A. Bull, PharmD, ALZA Corporation, Mt. View, California
Jeanne Carlson, CPA, Blue Care Network, BlueCross BlueShield
of Michigan, Southfield
Tara R. Cockerham, PharmD, Clinical Pharmacist, Atlanta, Georgia
Eric J. Culley, PharmD, Highmark BlueShield, Pittsburgh, Pennsylvania
Lisa A. Edwards, PharmD, Clinical Pharmacist, North Kingstown, Rhode
Island
Leslie Fish, PharmD, Fallon Community Health Plan, Worcester,
Massachusetts
Feride Frech, MPH, Novartis Pharmaceuticals Corp., East Hanover,
New Jersey
Zafar Hakim, PhD, Hoffman-La Roche, Nutley, New Jersey
Joel Hay, PhD, School of Pharmacy, University of Southern California,
Los Angeles
Katherine Knapp, PhD, College of Pharmacy, Touro University,
Vallejo, California
Christina Meyer, MHS, Caremark, Hunt Valley, Maryland
Robert P. Navarro, PharmD, Campbell Alliance, Raleigh, North Carolina
Eduardo Ortiz, MD, MPH, VA Medical Center, Washington, DC
Steven Pepin, PharmD, BCPS, MGI Pharma, Inc., Bloomington, Minnesota
Steven R. Peskin, MD, MBA, Pharmaceutical Research Plus, Severna Park,
Maryland
Cathlene Richmond, PharmD, Drug Information Services, Kaiser
Permanente, Oakland Regional Offices, California
Madeline Ritchie, PharmD candidate, College of Pharmacy, Idaho State
University, Boise
Fred L. Sego, Jr., JD, RPh, Poulsbo, Washington
Fadia T. Shaya, PhD, MPH, School of Pharmacy, University of Maryland,
Baltimore
Joshua Spooner, PharmD, MS, Advanced Concepts Institute, Philadelphia,
Pennsylvania
Marilyn Stebbins, PharmD, CHW Medical Foundation, Rancho Cordova,
California; University of California, San Francisco
Sean D. Sullivan, PhD, Pharmaceutical Outcomes Research and Policy
Program, University of Washington, Seattle
Kent H. Summers, PhD, School of Pharmacy, Purdue University, Lafayette,
Indiana
Sheryl L. Szeinbach, PhD, College of Pharmacy, Ohio State University,
Columbus
Robert J. Valuck, RPh, PhD, School of Pharmacy, University of Colorado
Health Sciences Center, Denver
Bill Yates, RPh, PhD, CaremarkPCS, Columbia, South Carolina
Journal of Managed Care Pharmacy (ISSN 1083–4087) is published 9 times per year and is the official publication of the Academy of Managed Care Pharmacy (AMCP),
100 North Pitt St., Suite 400, Alexandria, VA 22314; (703) 683-8416; (800) TAP-AMCP; (703) 683-8417 (fax). The paper used by the Journal of Managed Care Pharmacy
meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper) effective with Volume 7, Issue 5, 2001; prior to that issue, all paper was acid-free. Annual
membership dues for AMCP include $60 allocated for the Journal of Managed Care Pharmacy. POSTMASTER: Send address changes to JMCP, 100 North Pitt St., Suite 400,
Alexandria, VA 22314.
www.amcp.org
JMCP
September 2005
Vol. 11, No. 7
www.amcp.org
C O V E R
I M P R E S S I O N S
About our cover artist
Nashville Rocks! (2001) ■ Dave Newman
D
with an art agent who rented art for television
ave Newman creates a fascinating mixedshows. Newman’s artwork has been used in the
media work of art like a great chef concocts
background sets for shows such as “Seinfeld,”
a new dish. First, he chooses his basic
“Caroline in the City,” and “Felicity.”
ingredients: a piece of wood for the canvas and
His work is in many private and corporate
multiple hues of acrylic paint. Then he spices up
collections such as Ajo Al’s Restaurants in
the design with 3-dimensional objects culled from
Scottsdale, Arizona; Pfizer, Inc. in San Diego,
his studio filled with nostalgic memorabilia.
California; and Fender Guitar in Corona,
“I use photographs I’ve taken, metal parts, silkCalifornia. Newman recently painted two 10-feetscreened images, and ‘found objects’ of all kinds to
high Fender Stratocaster guitar sculptures that will
weave stories into my work,” Newman says. The
be donated to a Big Brothers Big Sisters of Central
finished collage is a delicious visual treat with the
Arizona charity auction in February 2006. The
perfect blend of color and texture.
commission was sponsored by the Best Western
Newman painted an acoustic guitar, musical
“My paintings are created with an
hotel chain and the Phoenix Airport Museum. His
notes, and running cowboys to convey the “Music
intentionally humorous, whimsical feel.”
guitars are part of “GuitarMania,” a Greater Phoenix
City” theme of his Nashville Rocks! collage.
public art project presented by Big Brothers Big
A 45-rpm record, dominoes, Scrabble tiles, scraps
of sheet music, Tennessee state maps, photographs, and pieces of old Sisters of Central Arizona, the Rock and Roll Hall of Fame and
license plates complete the composition. He created the artwork to be Museum, and Fender Guitar. The completed guitars will be displayed
used as the image for the 2001 Country Music Marathon’s official at the Phoenix Airport and elsewhere throughout the Valley for a fiveposter. Held in Nashville in April, the annual race began in 2000 and month period beginning in September 2005.
Newman’s Web site (www.davenewmanstudio.com) features
is organized and managed by Elite Racing, Inc., which also commissioned Newman to do the inaugural race poster. Country music bands original mixed-media collages and acrylic paintings. He says that he
play along the marathon route, and the participants are rewarded with enjoys using acrylic paint because it allows him to be impulsive and go
with his instincts as he creates each piece. The Newman Gallery
a star-studded postrace concert.
The original Nashville Rocks! was auctioned off at the end of the (www.newmangallery.net) showcases Newman’s art and represents
marathon, with proceeds going to the Leukemia & Lymphoma Society. 25 artists from around the country. It is located on historic Whiskey
Newman has a history of contributing to communities through his art- Row in the 100-year-old St. Michael Hotel in downtown Prescott.
work—he has also donated his art to the Heart Foundation, an Arizona Other galleries that feature his work are Atmosphere Art & Hair Salon
in Scottsdale, Arizona; Culture Shop Gallery in Venice, California; and
Montessori school, and museum auctions in Arizona and California.
Born in 1956 in Long Beach, California, Newman was reared in Terzian Galleries in Park City, Utah.
Newman says that one of his favorite themes is retro American
Anaheim, a mile away from Disneyland. Growing up so close to
Disneyland may have had an affect on his art, as his artist statement culture, especially Route 66. “Sometimes my art begins on a highway
reveals: “My paintings are created with an intentionally humorous, where there is a spectacular view or where I’ve found something interwhimsical feel.” Newman’s vibrant work also reflects his lighthearted esting on the side of the road,” he says. One of his upcoming projects
view of life. His collages are filled with so much detail that they almost involves the painting and collage of a 1977 Chevy pickup truck. He
always garner a second look—which is just what Newman desires. already has the personalized Arizona license plates, “ROAD ART.” That
“I want people to look again and again because there is always some- truck will undoubtedly blaze a colorful trail across the desert.
thing else to see,” he says.
Sheila Macho
Newman is a self-taught artist, and he maintains a studio, gallery,
Cover Editor
and residence in Prescott, Arizona. His interest in art began in his early
30s, when he operated a screen-printing business and started reading
about pop artists of the 1960s such as Andy Warhol and Jasper Johns. COVER CREDIT
Warhol frequently utilized screen printing in his work, and Johns’s Dave Newman, Nashville Rocks!, mixed media on wood. Prescott, Arizona.
Copyright© 2001.
paintings often included images and objects from popular culture.
By the age of 32, Newman had launched his own art career. “My
acrylic paintings on canvas and wood are spontaneous commentaries SOURCES
Interview with the artist.
on pop culture: coffee cups, hearts, television, and the spaces www.davenewmanstudio.com.
we inhabit,” he observed. When he lived in California, he worked www.newmangallery.net.
JMCP
September 2005
Vol. 11, No. 7
www.amcp.org
JMCP Author Guidelines
JMCP accepts for consideration manuscripts prepared according to the Uniform
Requirements for the Submission of
Manuscripts to Biomedical Journals.1
■■ Manuscript Preparation
Manuscripts should include, in this order:
title page, abstract, text, references, tables, and
figures (see Submission Checklist for details).
JMCP abstracts should be written narratives
that contain the information described for each
type of article shown below, where applicable.
For descriptions of editorial content, see “JMCP
Editorial Policy” in this Journal or at
www.amcp.org.
Original Research
An abstract is required in the format of:
• Keywords
• Objective • Results
• Methods
• Conclusion
Subject Reviews
An abstract is required, generally in the
format of:
• Objective • Conclusion
• Summary • Keywords
Formulary Management
An abstract is required in the format of
Original Research or Subject Reviews,
depending upon the subject matter.
Contemporary Subjects
An abstract is required in the format of
Original Research or Subject Reviews,
depending upon the subject matter.
Editorials
These submissions require no abstract
but should include references.
Letters
These submissions require no abstract
or title page.
Please note:
• A subsection in the Discussion labeled
“Limitations” is generally appropriate for
articles published in Original Research and
sometimes appropriate for articles published
in Subject Review, Formulary Management,
or Contemporary Subject.
• Most articles published in JMCP, particularly
Subject Reviews, should incorporate or at
least acknowledge the relevant work of
others published previously in JMCP (see
“Article Index by Subject Category” at
www.amcp.org).
• For articles in Original Research, a figure
is recommended for making the effects of the
inclusion and exclusion criteria clear to
readers (see JMCP examples in 2003;9(4):320
[Figure 1] or 2003;9(3):258 [Table 1]).
• Product trade names may be used only
once, for the purpose of providing clarity for
readers, generally at the first citation of the
generic name but not in the Abstract.
■■ Reference Style
References should be prepared following modified
AMA style. All reference number in manuscript
should be superscript (e.g., 1 ). See examples of
common types of references below:
The Journal of Managed Care Pharmacy is indexed
by Index Medicus/MEDLINE and
International Pharmaceutical Abstracts (IPA).
1. Standard journal article
(List all authors when 6 or less; if more than 6, list
only the first 3 and add et al.)
Lennard EL, Feinberg PE. Overview of the New
York State program for prescription drug benefits.
Am J Hosp Pharm. 1994;512:944-48.
2. No author given
Anonymous. Top 25 U.S. hospitals, ranked by
admissions, 1992. Manag Healthcare. 1994;4(9):64.
3. Journal paginated by issue
Corrigan PW, Luchins DJ, Malan RD, Harris J.
User-friendly CQI for the mental health team. Med
Interface. December 1994;7:89-92, 95.
4. Book or monograph by authors
Tootelian DH, Gaedeke RM. Essentials of Pharmacy
Management. St. Louis, MO: C.V. Mosby; 1993.
5. Book or monograph with editor, compiler, or
chairman as author
Chernow B, ed. Critical Care Pharmacotherapy.
Baltimore, MD: Williams & Wilkins; 1995.
6. Chapter in a book
Kreter B, Michael KA, DiPiro JT. Antimicrobial prophylaxis in surgery. In: DiPiro JT, Talbert RL, Hayes
PE, Yee GC, Matzke GR, Posey LM, eds. Pharmacotherapy: A Pathophysiologic Approach. Norwalk, CT:
Appleton & Lange; 1992:1811-12.
7. Government agency publication
Akutsu T. Total Heart Replacement Device.
Bethesda, MD: National Institutes of Health,
National Heart and Lung Institute; April 1974.
Report NIHNHLI-69-2185-84.
8. Paper (or Poster) presented at a meeting
Reagan ME. Workers’ compensation, managed care,
and reform. Paper (poster) presented at: 1995
AMCRA Midyear Managed Care Summit; March 13,
1995; San Diego, CA.
9. Newspaper
Winslow R. Lipitor prescriptions surge in wake of
big study. Wall Street Journal. March 18, 2004:D4.
10. Web site
National Committee for Quality Assurance. The
state of health care quality: 2004. Industry trends
and analysis. Available at: http://www.ncqa.org/
communications/SOMC/SOHC2004.pdf.
Accessed December 26, 2004.
■■ Manuscript Submission
A paper copy of the manuscript, including
originals of figures and tables and author attestation forms (see Submission Checklist),
should be submitted to the JMCP Peer Review
Administrator at the Academy of Managed
Care Pharmacy, 100 North Pitt St., Suite 400,
Alexandria, VA 22314; Tel: (800) 827-2627
or (703) 683-8416 or Fax: (703) 683-8417.
The paper copy is necessary to ensure proper
presentation and placement of text, figures,
tables, and graphs. Please submit the manuscripts electronically at jmcp.msubmit.net. All
text should be in a word processing program
(preferably Microsoft Word). Tables should be
prepared in a word processing program using
www.amcp.org
Vol. 11, No. 7
September 2005
the tab function to create columns, not using
“tables” or “cells.” Figures should be saved in
Photoshop or Illustrator and may be re-created
by us. Figures may also be prepared in a word
processing program (preferably Microsoft
Word). We can accept PowerPoint graphics.
Please identify the format (PC or MAC), all
programs used, and all file names. P values
should be expressed to no more than 3 decimal
points in the format 0.xxx.
Note: Please do not include author identification in the electronic manuscript document.
Cover letter: the corresponding (lead)
author should include a cover letter with
the manuscript, which
• briefly describes the importance and scope
of the manuscript,
• certifies that the paper has not been accepted
for publication or published previously and
that it is not under consideration by any
other publication, and
• identifies the nature and extent of any
financial interest or affiliation that any
author has with any company, product,
or service discussed in the manuscript.
All manuscripts are reviewed prior to peer
review. Manuscripts may be returned to
authors prior to peer review for clarification or
other revisions. Peer review generally requires
4 weeks but may extend as long as 8 weeks in
unusual cases. Solicited manuscripts are subject to the same peer-review standards and editorial policy as unsolicited manuscripts.
■■ Submission Checklist
Before submitting the paper copy of your manuscript to the Journal of Managed Care Pharmacy,
please check to see that your package includes the
following:
❑ Cover letter
❑ Manuscript: prepared in 12-point type, 1.5 line
spacing, including
❑ abstract: no more than 500 words
❑ keywords: follows the abstract
❑ references: cited in numerical order as they
appear in the text (use superscript numbers) and
prepared following modified AMA style; do not
include footnotes in the manuscript
❑ tables and figures (generally no more than a total
of 6). Submit referenced tables and figures on
separate pages with titles (and captions, as necessary) at the end of the manuscript; match symbols in tables and figures to explanatory notes, if
included. May use 10-point font.
❑ Disclosures and conflict-of-interest forms:
completed and signed author attestation forms
(available at www.amcp.org); clearly indicate
source(s) of funding and financial support.
Note: Please do not include author identification in
the electronic manuscript document.
For “Manuscript Submission Checklist” and “Peer
Review Checklist,” see www.amcp.org.
REFERENCE
1. International Committee of Medical Journal Editors.
Uniform requirements for manuscripts submitted to
biomedical journals. N Engl J Med. 1991;324:424-48.
JMCP
ORIGINAL RESEARCH
Gastrointestinal Bleeding Rates Among Managed Care Patients
Newly Started on COX-2 Inhibitors or Nonselective NSAIDs
KAREN STOCKL, PharmD; LORI CYPRIEN, MS; and EUNICE Y. CHANG, PhD
ABSTRACT
OBJECTIVE: While cyclooxygenase-2 (COX-2) inhibitors were introduced to the
U.S. market with the promise of less gastrointestinal (GI) toxicity than nonselective
nonsteroidal anti-inflammatory drugs (NSAIDs), additional research is needed to
examine this outcome in the naturalistic setting. The objective of this study was
to examine whether use of COX-2 inhibitors is associated with reduced risk of GI
bleed in a managed care population.
METHODS: Adult patients in a multistate managed care organization that were
initiated on a nonselective NSAID between January 1999 and August 2002 were
identified and matched using propensity scoring with patients in the same managed
care organization that were initiated on a COX-2 inhibitor. Matching variables
included age, gender, geographical state, comorbidity index, corticosteroid use,
warfarin use, arthritis indication, and history of recent GI bleed. Patients were
followed until they switched or discontinued their NSAID or COX-2 inhibitor,
disenrolled from the health plan, developed a GI bleed, or reached the end of the
1-year follow-up period. A GI bleed was defined as an inpatient hospitalization for
GI bleed or at least 2 medical claims with a primary diagnosis for GI bleed. The
relative risk (RR) of GI bleed was calculated using proportional hazards regression.
RESULTS: Overall, 35,007 pairs of COX-2 inhibitor and nonselective NSAID users
were evaluated. Mean age was 63 years, and 65% were female. There were 375
cases of GI bleed among 19,201 follow-up years for COX-2 users (19.5 cases per
1,000 person-years) versus 228 cases of GI bleed among 12,680 follow-up years
for NSAID users (18.0 cases per 1,000 person-years). The risk of GI bleed was not
significantly different for COX-2 users compared with nonselective NSAID users
(RR 1.07; 95% confidence interval [CI], 0.90-1.26). Even among high-risk
patients, there was no reduction in the risk of a GI bleed among users of COX-2
inhibitors (RR 0.995; 95% CI, 0.84 -1.19).
CONCLUSION: Overall, within this managed care population, COX-2 inhibitor users
did not have a reduced risk of a GI bleed compared with patients with similar
baseline characteristics using nonselective NSAIDs.
KEYWORDS: Cyclooxygenase, Nonsteroidal anti-inflammatory drugs,
Gastrointestinal bleed, Drug therapy
J Manag Care Pharm. 2005;11(7):550-58
Note: An editorial on the subject of this article appears on pages 590-93 of this issue.
Authors
KAREN STOCKL PharmD, was a research scientist, health economics and
outcomes research, Prescription Solutions, Costa Mesa, California, at the time
of this study; LORI CYPRIEN, MS, is a statistician, health economics and outcomes
research; EUNICE Y. CHANG, PhD, is chief statistician, health informatics and
outcomes research, Prescription Solutions, Costa Mesa, California.
AUTHOR CORRESPONDENCE: Karen Stockl, PharmD, c/o Eunice Y. Chang, PhD,
Chief Statistician, Health Informatics and Outcomes Research, Prescription
Solutions, 3515 Harbor Blvd., Mail Stop LC07-264, Costa Mesa, CA, 92626.
Tel: (714) 825-3732; Fax: (714) 825-3742; E-mail: [email protected]
Copyright© 2005, Academy of Managed Care Pharmacy. All rights reserved.
JMCP
September 2005
Vol. 11, No. 7
R
ecent reports of adverse cardiovascular outcomes with
two of the cyclooxygenase-2 (COX-2) inhibitors have
placed the risk-benefit profile of these COX-2 inhibitors
under public scrutiny; both of the COX-2 inhibitors have subsequently been withdrawn from the U.S. market).1-3 As a result,
clinicians have been encouraged to weigh the potential benefits
and risks of the nonsteroidal anti-inflammatory (NSAID)
medications prior to prescribing them.4 Managed care has been
particularly interested in examining the risks, benefits, and
cost-effectiveness of the COX-2 inhibitors.5-8 The present study
was designed to evaluate whether the use of COX-2 inhibitor
medications was beneficial in preventing gastrointestinal (GI)
bleeds compared with the use of nonselective NSAIDs in the
naturalistic managed care setting.
The early reported results of controlled clinical studies
suggested that the COX-2 inhibitors may have less GI toxicity
than nonselective NSAIDs,9-14 but the U.S. Food and Drug
Administration (FDA) permitted this claim only for rofecoxib,
which was later withdrawn from the U.S. market, on September
30, 2004, due to adverse cardiovascular events.1,15
In actual clinical practice, patients’ medication use is not
monitored as closely as in clinical trials. As a result, patients
may not take their COX-2 inhibitor medication under the same
conditions as those studied (e.g., they may take higher doses
than those prescribed or they may use COX-2 inhibitors
concomitantly with gastrotoxic substances such as alcohol),
which may lead to different outcomes than those observed in
clinical trials. Furthermore, the population prescribed COX-2
inhibitors in clinical practice may have more risk factors for a
GI bleed than the population selected to test the COX-2
inhibitors in clinical trials.
The objective of this analysis was to examine whether a
managed care population of patients who used COX-2
inhibitors in a naturalistic setting actually did have a reduced
occurrence of GI bleed compared with a population of patients
with similar baseline characteristics who received nonselective
NSAIDs.
■■ Methods
This was a retrospective analysis of electronic pharmacy and
medical administrative claims from a large managed care
organization and Prescription Solutions, a pharmacy benefits and
medical management company. Longitudinal claims data were
used from health plans (private as well as Medicare + Choice [now
Medicare Advantage]) within California, Oklahoma, Oregon,
Texas, and Washington, which consist of approximately 2.7 million
www.amcp.org
Gastrointestinal Bleeding Rates Among Managed Care Patients Newly Started on COX-2 Inhibitors or Nonselective NSAIDs
lives. This database has been used in previous research studies.16-20
An estimated two thirds of members within this managed care
population were subject to prior authorization for COX-2 drugs
that required patients to meet certain clinical characteristics prior
to receiving authorization for coverage of a COX-2 inhibitor.
Prior authorization approval for a COX-2 inhibitor was more
likely to be given to patients with older age, a diagnosis of
rheumatoid arthritis or osteoarthritis, history of GI bleed,
and/or concomitant use of oral corticosteroids or warfarin.
Patient Identification and Matching
Adult patients (aged 18 years or older) initiated on a COX-2
inhibitor (generic product identifier [GPI] code 661005xx,
which included rofecoxib, celecoxib, and valdecoxib) or a
nonselective NSAID (GPI code 661000xx, which consisted of
all prescription NSAIDs other than those classified as COX-2
inhibitors) during the 44-month period from January 1, 1999,
through August 31, 2002 (identification period), were identified.
The index date was defined as the date of each patient’s first
prescription fill of a COX-2 inhibitor or nonselective NSAID
during the identification period. Patients were excluded from
the analysis if they had a pharmacy claim for a COX-2 inhibitor
or a nonselective NSAID during the 6-month period prior to
their index date (i.e., the preperiod) or if they were not
continuously enrolled in the health plan during the preperiod
and at least 3 months after the index date. The first (earliest) patient
that could have been identified would have had an index date on
January 1, 1999, and a preperiod starting July 1, 1998. Hence,
some patients’ preperiods may have started prior to the FDA
approval of celecoxib. Celecoxib, the first COX-2 inhibitor to be
approved, was approved on December 31, 1998.21
From these identified patients, the final study cohort was
obtained by matching patients who received a nonselective
NSAID on the index date with those who received a COX-2
inhibitor on the index date on a 1:1 basis using the propensity score
method.22 A propensity score, which represents the likelihood of
receiving a COX-2 inhibitor rather than a nonselective NSAID, was
determined for each patient. Patients were matched based on
their propensity score.
The independent variables that were used to calculate the
propensity score included demographics (age at index date,
gender, geographical state of the health plan), Charlson
Comorbidity Index23 (calculated during the preperiod using a
method adapted for electronic claims databases),24 and the
following GI bleed risk factors (measured during the preperiod):
a prescription fill of a corticosteroid (GPI codes 2210xx, 2220xx,
2200xx), a prescription fill of warfarin (GPI code 83200030), a
medical claim representing a recent GI bleed (Table 1), history of
a GI bleed-related inpatient hospitalization (i.e., an inpatient
hospitalization with at least a 1-day length of stay and a diagnosis
code representing a GI bleed, Table 1), and arthritis indication
(osteoarthritis [International Classification of Diseases, 9th Revision,
www.amcp.org
TABLE 1
Diagnosis Codes Representing
Gastrointestinal Bleed
ICD-9-CM Code
Description
530.2
531.xx
532.xx
533.xx
534.xx
578.xx
Ulcer of esophagus
Gastric ulcer
Duodenal ulcer
Peptic ulcer site unspecified
Gastrojejunal ulcer
Gastrointestinal hemorrhage
ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical
Modification.
Clincial Modification (ICD-9-CM) code 715.xx], rheumatoid
arthritis [ICD-9-CM code 714.xx], or neither). Patients diagnosed with both osteoarthritis and rheumatoid arthritis were
classified into the rheumatoid arthritis group. If more than 1
patient who received a nonselective NSAID was identified as a
match, 1 patient was selected at random to be included in the final
study cohort. Patients who could not be matched were excluded
from the analysis.
Patients in the matched COX-2 inhibitor and nonselective
NSAID cohorts were further stratified according to their GI risk.
Patients were considered high-risk if they met any of the following
criteria: (1) age greater than 65 years; (2) GI-bleed-related
inpatient hospitalization during the preperiod; (3) pharmacy
claim for warfarin (GPI code 83200030) during the preperiod;
or (4) pharmacy claim for a corticosteroid (GPI codes 2210xx,
2220xx, 2200xx) during the preperiod. All other patients were
classified as low-risk.
Outcome Measures
The primary outcome of interest was the risk of developing a GI
bleed over the follow-up period. Patients were followed until
the first occurrence of one of the following events: (1) patient
discontinued (as defined below) the index COX-2 inhibitor or
nonselective NSAID medication; (2) crossover of medication of
interest (patient filled a prescription for a study medication in a
class [COX-2 inhibitor or nonselective NSAID] other than their
index class of medication); (3) patient disenrolled from the
health plan; (4) patient had a GI-bleed-related inpatient
hospitalization; (5) patient had 2 medical claims with a primary
(first-listed) diagnosis for Gl bleed (Table 1) during the follow-up
period (where the event date was defined as the date of the first
of the 2 claims); or (6) the end of the 1-year follow-up period.
A discontinuation was defined as a gap of at least 60 days
between the run-out date of the last index COX-2 inhibitor or
nonselective NSAID medication fill (fill date plus the days of
supply of that last prescription) and the end of the follow-up
period. Patients were considered to have a GI bleed event if they
experienced a GI bleed-related inpatient hospitalization (event
4 above) or had 2 medical claims with a primary diagnosis for
a GI bleed (event 5 above) within the follow-up time frame.
Vol. 11, No. 7
September 2005
JMCP
FIGURE 1
Study Population Identification
Patients Who Filled at Least 1 COX-2 Inhibitor
or Nonselective NSAID During the ID Period
(N = 1,038,437)
Patients Who Did Not Have a
COX-2 Inhibitor or Nonselective NSAID
Filled During the 6-Month Preperiod
(N = 958,075; 92.3%)
Patients Who Were Continuously Enrolled
in the Health Plan in the Preperiod
and at Least 3 Months After the Index Date
(N = 613,526; 59.1%)
Patients Who Were Aged
at Least 18 years or Older
(N = 585,634; 56.4%)
COX-2 Users
(N = 36,401; 3.5%)
Nonselective NSAID Users
(N = 549,233; 52.9%)
COX-2 Users Matched With
(N = 35,007; 3.4%)
Matched With COX-2 Users
(N = 35,007; 3.4%)
COX-2 = cyclooxygenase-2; GI = gastrointestinal; NSAID = nonsteroidal antiinflammatory drug.
To account for the different lengths of follow-up, the number of
GI bleed cases per 1,000 person years was calculated by dividing
the number of patients who experienced the GI bleed event
during the follow-up period by the sum of each patient’s observed
follow-up time (in years) and then multiplying by 1,000.
Statistical Analysis
Statistical analyses were performed using SAS version 9.1 (SAS
Institute Inc., Cary, NC). Chi-square tests and t tests were used
to compare baseline demographics and clinical characteristics
between the COX-2 inhibitor and nonselective NSAID cohorts.
Propensity score calculations were conducted using logistic
regression. Two-way interactions between independent variables
were tested and only significant interactions (P < 0.05) were
included in the final propensity model. The propensity score for
each patient was rounded to 0.0001 for matching. Proportional
JMCP
September 2005
Vol. 11, No. 7
hazards regression was used to determine relative risk of a GI
bleed event for the 2 cohorts, to study the relationship between
the event and specific risk factors, and to adjust for baseline
differences between the cohorts. Risk factors included in the
model were age, gender, state of health plan, preperiod
Charlson Comorbidity Index, preperiod corticosteroid use,
preperiod warfarin use, preperiod diagnosis of GI bleed in any
diagnostic field, and preperiod rheumatoid or osteoarthritis
indication. Separate models were created for the total study
population, the low-risk population, and the high-risk population.
Only the main effects were included in the models (interaction
terms were not tested for inclusion).
To adjust for baseline differences in the use of
gastroprotective agents between the cohorts, additional
proportional hazards regression models were performed after
adding preperiod use of a gastroprotective agent as a variable. To
assess the proportional hazards assumption, a plot of the scaled
Schoenfeld residuals by transformed time for each risk factor was
investigated and a test of zero slope of the plot was conducted.25-27
Nonzero slope would indicate a violation of the proportional
hazards assumption. If nonproportional hazards were found for
some risk factors, then stratified proportional hazards
regressions were conducted to evaluate whether the GI bleed
event outcome was changed by this stratification. All statistical
tests were 2-sided with an alpha of 0.05.
■■ Results
There were 1,038,437 patients who filled at least 1 COX-2
inhibitor or nonselective NSAID during the identification
period (Figure 1). Among them, 80,362 (7.7%) were excluded
because they had a pharmacy claim for a COX-2 inhibitor or
nonselective NSAID during the preperiod, and 344,549 were
excluded because they were not continuously enrolled in the
preperiod and at least 3 months after the index date. An additional
27,892 patients were excluded because they were younger than
18 years.
Overall, 585,634 eligible patients were identified; 36,401
(6.2%) used COX-2 inhibitors and 549,233 (93.8%) used
nonselective NSAIDs. A total of 70,014 patients (35,007 pairs
of COX-2 inhibitor users and nonselective NSAID users) were
matched according to propensity score and included in the final
study cohort.
Demographics and clinical characteristics were similar for
COX-2 inhibitor and nonselective NSAID cohorts (Table 2),
with the exception of a lower percentage of COX-2 inhibitor
users participating in a Medicare + Choice health plan (53.1%
versus 55.7%, P < 0.001); health plan type was not one of the
variables included in the propensity score match. When the
populations were stratified according to low and high GI bleed
risk, there were also statistical differences in the mean age for
the COX-2 inhibitor and nonselective NSAID cohorts, which
were apparently due to the large sample size since the mean
www.amcp.org
TABLE 2
Baseline Demographics and Clinical Characteristics of Study Population*
Total Population
COX-2
Inhibitor
(N = 35,007)
Age, mean (SD)
Nonselective
NSAID
(N = 35,007)
Low-Risk Population
COX-2
Inhibitor
(N = 14,897)
High-Risk Population†
Nonselective
NSAID
(N = 14,650)
COX-2
Inhibitor
(N = 20,110)
Nonselective
NSAID
(N = 20,357)
63.4 (16.3)
63.3 (16.3)
48.9 (10.5)||
48.6 (10.6)||
74.2 (10.3)||
73.9 (10.5)||
Female gender
22,909 (65.4)
22,906 (65.4)
9,219 (61.9)
9,142 (62.4)
13,690 (68.1)
13,764 (67.6)
State of health plan
California
Oklahoma
Oregon
Texas
Washington
24,090 (68.8)
1,771 (5.1)
1,555 (4.4)
3,843 (11.0)
3,748 (10.7)
23,936 (68.4)
1,756 (5.0)
1,560 (4.5)
3,961 (11.3)
3,794 (10.8)
10,726 (72.0)
951 (6.4)
724 (4.9)
1,373 (9.2)
1,123 (7.5)
10,506 (71.7)
902 (6.2)
729 (5.0)
1,361 (9.3)
1,152 (7.9)
13,364 (66.5)
820 (4.1)
831 (4.1)
2,470 (12.3)
2,625 (13.1)
13,430 (66.0)
854 (4.2)
831 (4.1)
2,600 (12.8)
2,642 (13.0)
Medicare + Choice health plan (%)‡
18,574 (53.1)¶
19,488 (55.7)¶
17,587 (87.5)¶
18,349 (90.1)¶
0.29 (0.95)
0.30 (0.98)
0.15 (0.69)
0.15 (0.71)
0.39 (1.09)
0.40 (1.12)
1,040 (3.0)
5,788 (16.5)
28,179 (80.5)
1,015 (2.9)
5,802 (16.6)
28,190 (80.5)
276 (1.9)
1,224 (8.2)
13,397 (89.9)
283 (1.9)
1,189 (8.1)
13,178 (90.0)
764 (3.8)
4,564 (22.7)
14,782 (73.5)
732 (3.6)
4,613 (22.7)
15,012 (73.7)
Warfarin use
2,264 (6.5)
2,262 (6.5)
0
0
2,264 (11.3)
2,262 (11.1)
Corticosteroid use
3,385 (9.7)
3,406 (9.7)
0
0
3,385 (16.8)
3,406 (16.7)
Medical claim for GI bleed (any field)
843 (2.4)
808 (2.3)
252 (1.7)
223 (1.5)
591 (2.9)
585 (2.9)
Inpatient hospitalization for GI bleed
132 (0.4)
119 (0.3)
0
0
132 (0.7)
119 (0.6)
Charlson Comorbidity Index, mean (SD)
Arthritis indication
Rheumatoid arthritis§
Osteoarthritis
None
987 (6.6)¶
1,139 (7.8)¶
* Values are number (%) unless specified otherwise.
† High-risk was defined as a patient with (a) age older than 65 years, or (b) recent history use of either warfarin or corticosteroid, or (c) a recent hospitalization for
a GI bleed.
‡ Variable was not used in the propensity score match.
§ Includes patients with medical claims for both osteoarthritis and rheumatoid arthritis.
|| P = 0.004 for the low-risk population and P = 0.02 for the high-risk population for the comparison of the COX-2 inhibitor cohort and the nonselective NSAID cohort.
¶ P < 0.001 for all 3 populations (total, low-risk, and high-risk) for the comparison of the COX-2 inhibitor cohort and the nonselective NSAID cohort.
COX-2 = cyclooxygenase-2; GI = gastrointestinal; NSAID = nonsteroidal anti-inflammatory drug.
(± standard deviation) age was similar for the 2 cohorts (48.9 ±
10.5 years versus 48.6 ± 10.6 years, respectively, for low-risk
users, P = 0.004; 74.2 ± 10.3 years versus 73.9 ± 10.5 years,
respectively, for high-risk users, P = 0.02). Although statistically
significant, the small difference in age between the groups
probably is not clinically meaningful. During the preperiod,
17.5% of COX-2 inhibitor users and 7.8% of nonselective
NSAID users filled at least 1 prescription for a GI protective
agent (i.e., a proton pump inhibitor or misoprostol, GPI codes
49250030xx, 6610990220, 4927xx, 6610990242).
Subjects in the nonselective NSAID cohort had a shorter
follow-up time (average 4.3 months per patient or 12,680 personyears, Table 3) than subjects in the COX-2 inhibitor cohort
(average 6.6 months per patient or 19,201 person-years).
A GI bleed event (defined as a GI-bleed-related inpatient
hospitalization or at least 2 medical claims with a primary
diagnosis for GI bleed) was noted in 19.5 patients per 1,000
person years in the COX-2 inhibitor cohort and 18.0 patients
per 1,000 person years in the nonselective NSAID cohort (Table 3).
The time to GI bleed event (median and mean) was shorter for the
nonselective NSAID cohort than for the COX-2 inhibitor cohort.
www.amcp.org
The proportions of patients in each cohort whose follow-up
period ended due to reasons other than a GI bleed event are
described below. Compared with COX-2 inhibitor users, a
greater percentage of nonselective NSAID users discontinued
their index medication (82.2% versus 63.6%), while a smaller
percentage of nonselective NSAID users had a crossover of
medication of interest (1.2% versus 4.6%). Nonselective NSAID
users were also less likely than COX-2 inhibitor users to
disenroll from the health plan (15.8% versus 30.1%) but were
less likely to reach the end of the 1-year follow-up period than
COX-2 inhibitor users (0.2% versus 0.6%).
Patients who had a GI bleed event had a mean age of 75.1 ±
11.3 years and 61% were female. The mean Charlson
Comorbidity Index was 0.84 ± 1.57. Among these patients,
15.3% used warfarin and 13.1% used corticosteroids during the
preperiod. During the preperiod, 16.1% of these patients had a
medical claim for a GI bleed (any field), and 3.6% had an inpatient hospitalization for a GI bleed. At the end of follow-up, use
of a GI protective agent was observed in 21.1% of COX-2
inhibitor and 8.8% of nonselective NSAID users who had a GI
bleed event compared with 8.1% of COX-2 inhibitor and 4.0%
Vol. 11, No. 7
September 2005
JMCP
TABLE 3
GI Bleed Events, Number of Cases per 1,000 Person Years
Total Population
COX-2
Inhibitor
(N = 35,007)
Sum of patients’ follow-up years
Inpatient hospitalization
or primary diagnosis for
GI bleed on 2 medical
claims
Low-Risk Population
Nonselective
NSAID
(N = 35,007)
High-Risk Population*
COX-2
Inhibitor
(N = 14,897)
Nonselective
NSAID
(N = 14,650)
COX-2
Inhibitor
(N = 20,110)
Nonselective
NSAID
(N = 20,357)
7,677
19,201
12,680
6,605
5,003
12,596
Number of cases
375
228
51
19
324
209
Number of cases per 1,000
person-years
19.5
18.0
7.7
3.8
25.7
27.2
84.0
61.5
102.0
78.0
77.0
61.0
111.5
100.9
117.3
110.6
110.6
100.0
Median time to event (days)†
Mean time to event (days)†
* High-risk was defined as a patient with (a) age older than 65 years, or (b) recent history use of either warfarin or corticosteroid, or (c) a recent hospitalization for a GI bleed.
† Median and mean time to event were measured among those patients who had a GI bleed event during the follow-up period.
of nonselective NSAID users who did not have a GI bleed event.
Proportional hazards regression was conducted to determine
the relative risk of a GI bleed event for the 2 cohorts (accounting
for different patient follow-up times and adjusting for baseline
characteristics) and to study the relationship between the event and
specific risk factors (Table 4). Among the total study population
and the high-risk population, patients who received a COX-2
inhibitor did not have a lower relative risk of having a GI bleed
event (defined as an inpatient hospitalization for GI bleed or at
least 2 medical claims with a primary diagnosis for GI bleed)
compared with patients who received a nonselective NSAID.
Other factors that were significant predictors of a GI bleed event
were older age, male gender, state of health plan (California versus Texas), higher Charlson Comorbidity Index, use of warfarin,
and a history of GI bleed. For low-risk patients, those who
received a COX-2 inhibitor had a higher relative risk of a
GI bleed event compared with patients who received
nonselective NSAIDs. Male gender, higher Charlson
Comorbidity Index, history of GI bleed, and diagnosis of
osteoarthritis were found to be additional risk factors, but age
and state of health plan were not.
Within the total study population and the high-risk population,
the test of the proportional hazards assumption (using the plot
of the scaled Shoenfeld residuals by transformed time for each
risk factor) revealed significant evidence of nonproportional
hazards for the risk factor of a preperiod history of GI bleed.
Stratification by preperiod history of GI bleed did not significantly
alter the hazard ratio for the comparison between COX-2
inhibitors and nonselective NSAIDs or for the other specific risk
factors in the model. Within the low-risk population, the test of
the proportional hazards assumption found significant evidence of
nonproportional hazards for the risk factor of arthritis indication
on a medical claim. Stratification by arthritis indication did not
significantly alter the other hazard ratios in the model.
Since a higher proportion of patients treated with a COX-2
inhibitor used a gastroprotective agent during the preperiod,
JMCP
September 2005
Vol. 11, No. 7
additional proportional hazards regression models were
performed to evaluate whether the GI bleed outcome changed after
adjusting for differences in preperiod use of a gastroprotective
agent (Table 5). In the total study population and the high-risk
population, the hazard ratios changed slightly, but the overall
study findings and level of significance were not altered.
However, in the low-risk population, patients who received a
COX-2 inhibitor no longer had a significantly greater risk of
having a GI bleed event compared with patients who received
a nonselective NSAID after adjusting for preperiod use of a
gastroprotective agent. In all 3 study populations (total population,
low-risk, and high-risk), preperiod use of gastroprotective
agents was a factor associated with a higher risk of a GI bleed
during follow-up. Similar to the previous models, stratified
models were performed because there was evidence of
nonproportional hazards for the risk factor of history of GI bleed
for the total population and the high-risk population and for
the risk factor of arthritis indication for the low-risk population.
The stratified models did not significantly alter the hazard ratios.
■■ Discussion
In this retrospective study of a population of COX-2 inhibitor
users who were matched to nonselective NSAID users with similar
baseline characteristics, a lower risk of having a GI bleed was
not observed among patients receiving COX-2 inhibitors.
Despite the fact that the early controlled clinical trials of the
COX-2 inhibitors suggested a lower risk of GI bleed, particularly
for rofecoxib,9-15 similar results were not demonstrated at the
population level within this managed care setting.
The results from the present study expand on those from a
cross-sectional time series analysis from 1994 through 2002
among patients older than 66 years in Ontario, Canada, where
a 41% increase in NSAID utilization (resulting from the
increased use of COX-2 inhibitors) was accompanied by a 10%
increase in hospitalization rates for upper GI bleed.28 Although
causation was not proven, adverse outcomes may result when
www.amcp.org
high-risk patients who would normally not be prescribed
nonselective NSAIDs are treated with COX-2 inhibitors because
these medications are perceived to be safer. In the present study,
the use of COX-2 inhibitors was not associated with higher
rates of GI bleed for the high-risk population, and there was no
measurable reduction in the risk of GI bleed for COX-2
inhibitor users compared with nonselective NSAID users.
Contrary to the findings of the present study, an earlier
observational study published in 2002 examining an
administrative health care database in Ontario, Canada, found a
higher risk of hospitalization for a GI hemorrhage with the
nonselective NSAIDs than with the COX-2 inhibitors (12.6 GI
hemorrhages per 1,000 person years for nonselective NSAIDs
versus 3.6 to 7.3 GI hemorrhages per 1,000 person years for
COX-2 inhibitors, adjusted rate ratio ranged from 1.9 to 4.4 for
nonselective NSAIDs versus COX-2 inhibitors).29 While this
study was similar in concept to the present study, the 2 studies
used different study populations and health systems, different
cohort matching techniques, different model covariates, and
different definitions of a GI bleed. The Canadian study evaluated
a high-risk population (patients aged 66 years and older), while
the present evaluated both low- and high-risk patients. In the
Canadian study, the GI bleed outcome was limited to
hospitalizations for a GI bleed, but the present study examined
hospitalizations as well as outpatient claims for a GI bleed
because many GI bleeds are treated in the outpatient setting
(hence, the absolute rate of GI bleeds was lower in the Canadian
study compared with the present study).
The Canadian study did not match nonselective NSAID
users to COX-2 inhibitor users but, instead, adjusted for
covariates using a Cox proportional hazards model. The
covariates used within the Canadian 2002 study were different
than those used in the present study. One noteworthy difference
was that the Canadian study measured past history of a GI bleed
over the prior 5-year period, while the present study limited
this look-back period to 6 months since the frequent turnover
of patients in U.S. managed care plans would not allow for
a longer evaluation without significantly reducing the study
sample size. While the differences in study design prohibit a
direct comparison of the 2 studies, the contrasting results
between the studies indicate a need for additional research to
further understand the risk of GI bleed among users of COX-2
inhibitors and nonselective NSAIDs.
Whether COX-2 inhibitors are cost effective at the population
level remains a controversial issue.5-8, 30-32 Based on data from this
managed care organization during the fourth quarter of 2004,
the average pharmacy ingredient cost per 30 days for a COX-2
inhibitor prescription was $95.70, which, if filled regularly for
a 1-year period, would have an annual drug cost of $1,148 per
patient. In comparison, the average pharmacy ingredient cost
per 30 days for a nonselective NSAID was $16.56, or $199 per
patient per year. In other words, it would be possible to treat
www.amcp.org
TABLE 4
Hazard Ratio (95% Confidence Interval)
of a GI Bleed Event*
Total
Population
(N = 70,014)
NSAID Selection
COX-2 inhibitor vs.
nonselective NSAID
Other Factors
Age
Gender
(male vs. female)
1.07 (0.90-1.26)
Low-Risk
Population
(N = 29,547)
High-Risk
Population†
(N = 40,467)
2.05 (1.21-3.48)‡
0.995 (0.84-1.19)
1.61 (1.50-1.72)‡
1.25 (0.97-1.62)
1.53 (1.39-1.68)‡
1.34 (1.13-1.58)‡
1.64 (1.02-2.63)‡
1.31 (1.09-1.56)‡
0.55 (0.17-1.77)
0.94 (0.69-1.26)
0.75 (0.30-1.87)
0.55 (0.14-2.28)
0.66 (0.48-0.91)‡
0.81 (0.48-1.36)
0.38 (0.09-1.54)
0.77 (0.47-1.25)
Washington
0.93 (0.69-1.24)
vs. California
Texas vs. California 0.68 (0.50-0.92)‡
Oregon
0.79 (0.49-1.29)
vs. California
Oklahoma
0.69 (0.43-1.09)
vs. California
Charlson
Comorbidity Index
1.11 (1.04-1.18)‡
1.26 (1.08-1.49)‡
1.10 (1.03-1.17)‡
Corticosteroid use
(yes vs. no)
1.27 (0.99-1.61)
–
1.20 (0.94-1.54)
Warfarin use
(yes vs. no)
1.75 (1.39-2.19)‡
–
1.68 (1.33-2.12)‡
History of GI bleed
(yes vs. no)
5.35 (4.00-7.15)‡ 10.30 (5.19-20.45)‡ 4.82 (3.52-6.60)‡
Rheumatoid
arthritis vs. none
Osteoarthritis
vs. none
1.05 (0.69-1.61)
1.44 (0.45-4.65)
0.98 (0.62-1.55)
1.10 (0.91-1.32)
2.01 (1.10-3.67)‡
1.03 (0.85-1.26)
* A separate regression model was performed for each of the 3 populations:
the total study population, low-risk population, and high-risk population.
† High-risk was defined as a patient with (a) age older than 65 years, or (b) recent
history use of either warfarin or corticosteroid, or (c) a recent hospitalization for a
GI bleed.
‡ Indicates a level of significance of P < 0.05.
6 patients with nonselective NSAIDs for the same drug cost
incurred for 1 patient taking a COX-2 drug.
Since patients using the COX-2 inhibitors did not demonstrate
a reduction in GI bleeding in the present study, the cost benefit
of the COX-2 inhibitors is questionable. Previous research has
shown that use of COX-2 inhibitors may not be cost effective
among patients with low or average GI risk,30 and the findings
from the current study suggest that COX-2 inhibitors may not
be cost effective even in high-risk populations. While the use of
nonselective NSAIDs for high-risk patients may not be
appropriate, the use of COX-2 inhibitors within high-risk
populations must also be questioned. The use of COX-2
inhibitors may be appropriate for some patients; however,
Vol. 11, No. 7
September 2005
JMCP
Hazard Ratio (95% Confidence Interval) of a cohort. Because patients were not randomized to treatment
GI Bleed Event, Adjusting for Preperiod Use group, there exists the potential for selection bias between the
COX-2 inhibitor and nonselective NSAID cohorts. Physicians
of a Gastroprotective Agent*
may have decided to prescribe COX-2 inhibitors instead of
Total
Low-Risk
High-Risk
nonselective NSAIDs for their higher-risk patients.
Population
Population
Population†
In many of the health plans within this managed care
(N = 70,014)
(N = 29,547)
(N = 40,467)
population, COX-2 inhibitors were subjected to prior authorization,
NSAID Selection
COX-2 inhibitor vs. 1.01 (0.85-1.19) 1.65 (0.96-2.85) 0.96 (0.80-1.14)
which required patients to meet certain clinical characteristics
nonselective NSAID
prior to receiving an authorization for coverage of a COX-2
Other Factors
inhibitor. Such criteria commonly include age, diagnosis, GI
Age
1.61 (1.51-1.73)‡
1.22 (0.94-1.57) 1.54 (1.40-1.70)‡
bleed history, and concomitant medication use. Thus, patients
Gender
1.35 (1.15-1.60)‡ 1.63 (1.02-2.62)‡ 1.32 (1.10-1.57)‡
receiving COX-2 inhibitors may have had different GI risk
(male vs. female)
factors than those receiving nonselective NSAIDs. Due to database
limitations, it was not possible to determine the proportion of
Washington
0.95 (0.71-1.28)
0.56 (0.18-1.79) 0.96 (0.71-1.30)
subjects within each cohort that required prior authorization to
vs. California
obtain a COX-2 inhibitor. However, propensity matching was
Texas vs. California 0.69 (0.51-0.92)‡
0.74 (0.30-1.86) 0.67 (0.49-0.92)‡
Oregon
0.81 (0.50-1.32)
0.54 (0.13-2.22) 0.83 (0.50-1.40)
used to minimize demographic, clinical, and geographical
vs. California
differences between the cohorts.
Oklahoma
0.68 (0.43-1.08)
0.38 (0.09-1.58) 0.76 (0.47-1.24)
The percentage of patients in the Medicare + Choice health
vs. California
plan
was statistically different for the COX-2 inhibitor group
Charlson
1.10 (1.04-1.17)‡ 1.30 (1.11-1.51)‡ 1.09 (1.02-1.17)‡
and the nonselective NSAID group (53.1% versus 55.7%). The
Comorbidity Index
practical significance of this 2.6-point absolute difference is not
Corticosteroid use
1.23 (0.97-1.57)
–
1.18 (0.92-1.52)
clear, but the Medicare + Choice health plan may have had a
(yes vs. no)
maximum annual pharmacy benefit ranging from $500 to
Warfarin use
1.78 (1.42-2.23)‡
–
1.70 (1.35-2.15)‡
(yes vs. no)
$2,000 and generic-only benefits (with or without limits),
depending on the benefit year and the geographical county of the
History of GI bleed
4.70 (3.50-6.30)‡ 6.41 (3.17-12.95)‡ 4.40 (3.20-6.04)‡
(yes vs. no)
member.
The 6-month preperiod may not have been long enough to
Rheumatoid
1.03 (0.68-1.58)
1.38 (0.43-4.45) 0.97 (0.62-1.53)
capture the patient’s entire past history. For example, if a patient
arthritis vs. none
had a GI bleed prior to the preperiod, his or her history of GI
Osteoarthritis
1.10 (0.91-1.32) 1.98 (1.08-3.60)‡ 1.03 (0.85-1.26)
bleed would not have been captured. All of the potential risk
vs. none
factors for a GI bleed could not be captured in this claims
Preperiod
1.58 (1.30-1.91)‡ 2.85 (1.70-4.80)‡ 1.43 (1.15-1.76)‡
database, including use of alcohol, tobacco, aspirin or over-thegastroprotective agent
use (yes vs. no)
counter nonselective NSAIDs. Patients may have had other risk
* A separate regression model was performed for each of the 3 populations:
factors for GI bleed (e.g., bleeding disorders) that were not
the total study population, low-risk population, and high-risk population.
considered in this analysis. As a result, some high-risk patients
† High-risk was defined as a patient with (a) age older than 65 years, or (b) recent
could have been mismatched or incorrectly classified as
history use of either warfarin or corticosteroid, or (c) a recent hospitalization for a
GI bleed.
low-risk patients because of missing data. Within the low-risk
‡ Indicates a level of significance of P < 0.05.
population, the higher rate of GI bleed among patients
COX-2 = cyclooxygenase-2; GI = gastrointestinal; NSAID = nonsteroidal anti-inflam- receiving COX-2 inhibitors versus patients receiving nonselective
matory drug.
NSAIDs may be the result of incorrect (false assignment)
classification of high-risk patients into the low-risk population.
In contrast, patients in the high-risk population were more likely
further research is warranted to determine the specific patient to be correctly classified since they were required to have a
high-risk condition (i.e., age older than 65 years, recent history
populations where the COX-2 inhibitors are cost effective.
of warfarin or corticosteroid use, or a recent hospitalization for
Limitations
a GI bleed) to be included in this population.
Since these results were obtained among health maintenance
Potential study limitations should be considered when generalizing
these results to other populations. Although patients were organizations and Medicare + Choice patients in 3 western states,
matched by propensity score to control for possible selection Texas, and Oklahoma, similar results may not be observed among
bias, the matching may not have accounted for all of the factors populations with different demographic or socioeconomic characthat could have led to adverse selection for the COX-2 inhibitor teristics (such as the Medicaid population). These results were
TABLE 5
JMCP
September 2005
Vol. 11, No. 7
www.amcp.org
subject to the definition used to identify a case of GI bleed. A
GI bleed event was defined as an inpatient hospitalization with a
diagnosis representing GI bleed or at least 2 medical claims with
a primary (first-listed) diagnosis representing a GI bleed.
Medical records could be used to validate this definition of a GI
bleed; however, examination of patient medical records was
beyond the scope of this study. Further study of the accuracy of
medical claims diagnoses in identifying actual GI bleeds would
add to the literature. Applying a less stringent or more stringent
definition for a GI bleed could have changed the study outcomes.
In order to be included in the analysis, patients were only
required to have 3 months of continuous enrollment in the health
plan following their fill of a COX-2 inhibitor or nonselective
NSAID. While some patients could have been followed for a
maximum of 1 year, others were followed for a shorter period if
they disenrolled from the health plan, switched or discontinued
their index class of medication (COX-2 inhibitor or nonselective
NSAID), or had a GI bleed. Further research is needed to understand whether these results would have been different if patients
were studied over a longer follow-up period.
While rates of concomitant use of proton pump inhibitors or
other gastrointestinal protective agents were reported, an
evaluation of the impact of gastroprotective agents on GI
outcomes was beyond the scope of this study, which was
designed to examine the relative risk of having a GI bleed
among patients using COX-2 inhibitors or nonselective NSAIDs
within this managed care population. Introducing the use of
gastroprotective agents as a variable within the analysis could
have confounded the results since the use of gastroprotective
agents can represent either a risk factor or a study outcome.
While the use of gastroprotective agents in combination with
COX-2 inhibitors or nonselective NSAIDs merits further
research, a study designed specifically to measure the
concomitant use of gastroprotective agents such as a
case-control study could help eliminate some of the confouning
associated with this measure.
■■ Conclusion
This study provides insight into the rates of GI bleeding among
a large population of managed care patients initiated on
COX-2 inhibitors or nonselective NSAIDs. Patients using a
COX-2 inhibitor did not have a reduced risk of a GI bleed
compared with patients with similar baseline characteristics
who were using nonselective NSAIDs. With the high direct-drug
cost of COX-2 inhibitors and the uncertain risk of adverse
cardiovascular events, further research is needed to reevaluate
the appropriate patient populations for cost-effective treatment
with COX-2 inhibitors.
DISCLOSURES
No outside funding supported this study. Author Karen Stockl served as
principal author of the study. Study concept and design were contributed
primarily by Stockl and author Eunice Y. Chang. Analysis and interpretation
www.amcp.org
of data were contributed by Stockl, Chang, and author Lori Cyprien. Drafting
of the manuscript was the work of Stockl and Chang, and its critical revision
was the work of Chang. Statistical expertise was contributed by Cyprien and
Chang. Stockl discloses that she owns a small amount of Merck stock; Stockl,
Cyprien, and Chang disclose no potential bias or conflict of interest relating to
this article.
REFERENCES
1. U.S. Food and Drug Administration. FDA issues public health advisory on
Vioxx as its manufacturer voluntarily withdraws the product [FDA news
release. September 30, 2004]. Available at: http://www.fda.gov/bbs/topics/news/
2004/NEW01122.html. Accessed April 29, 2005.
2. U.S. Food and Drug Administration. FDA announces important changes
and additional warnings for COX-2 selective and nonselective nonsteroidal
anti-inflammatory drugs (NSAIDs) [FDA public health advisory, April 7,
2005]. Available at: http://www.fda.gov/cder/drug/advisory/COX2.htm.
Accessed April 29, 2005.
3. Ortiz E. Market withdrawal of Vioxx: Is it time to rethink the use of COX-2
inhibitors? J Manag Care Pharm. 2004;10(6):551-54.
4. U.S. Food and Drug Administration. Prescription nonsteroidal antiinflammatory drugs (NSAIDs) [FDA alert for health care professionals, April 7,
2005]. Available at: http://www.fda.gov. Accessed April 29, 2005.
5. Cox ER, Motheral B, Mager D. Verification of a decision analytic model
assumption using real-world practice data: implications for the cost effectiveness
of cyclo-oxygenase 2 inhibitors (COX-2s). Am J Manag Care. 2003;9:785-94.
6. Stacy J, Shaw E, Arledge MD, Howell-Smith D. Pharmacoeconomic modeling
of prior-authorization intervention for COX-2 specific inhibitors in a 3-tier
copay plan. J Manag Care Pharm. 2003;9(4):327-34.
7. Bull S, Conell C, Campen DH. Relationship of clinical factors to the use of
COX-2 selective NSAIDs within an arthritis population in a large HMO.
J Manag Care Pharm. 2002;8(4):252-58.
8. Curtiss FR. Cost-effective use of COX-2 drugs and NSAIDs. J Manag Care
Pharm. 2002;8(4):295-96.
9. Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal
toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis.
N Engl J Med. 2000;343:1520-28.
10. Silverstein FE, Faich G, Goldstein JL, et al. Gastrointestinal toxicity with
celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthritis and
rheumatoid arthritis: the CLASS study: a randomized controlled trial. JAMA.
2000;284:1247-55.
11. Langman MJ, Jensen DM, Watson DJ, et al. Adverse upper gastrointestinal
effects of rofecoxib compared with NSAIDs. JAMA. 1999;282:1929-33.
12. Hawkey C, Laine L, Simon T, et al. Comparison of the effect of rofecoxib
(a cyclooxygenase 2 inhibitor), ibuprofen, and placebo on the gastroduodenal
mucosa of patients with osteoarthritis: A randomized, double-blind, placebocontrolled trial. Arthritis Rheum. 2000;43:370-77.
13. Laine L, Harper S, Simon T, et al. A randomized trial comparing the effect
of rofecoxib, a cyclooxygenase 2-specific inhibitor, with that of ibuprofen on
the gastroduodenal mucosa of patients with osteoarthritis. Gastroenterology.
1999;117:776-83.
14. Simon LE, Weaver AL, Graham DY, et al. Anti-inflammatory and upper
gastrointestinal effects of celecoxib in rheumatoid arthritis: a randomized
controlled trial. JAMA. 1999;282:1921-28.
15. Rofecoxib (Vioxx) [prescribing information]. NDA 21-042/S-018 and
NDA 21-052/S-012. Available at: http://www.fda.gov/cder/foi/label/2003/
21042se8-018,21052se8-012_vioxx_lbl.pdf. Accessed July 27, 2005.
16. Stockl K, Vanderplas A, Tafesse E, Chang E. Costs of lower-extremity
ulcers among patients with diabetes. Diabetes Care. 2004;27(9):2129-34.
Vol. 11, No. 7
September 2005
JMCP
17. White TJ, Vanderplas A, Chang E, Dezii CM, Abrams GD. The costs of
non-adherence to oral antihyperglycemic medication in individuals with diabetes
mellitus and concomitant diabetes mellitus and cardiovascular disease in a
managed care environment. Dis Manag Health Outcomes. 2004:12(3):181-88.
25. Schoenfeld D. Partial residuals for the proportional hazards regression
model. Biometrika. 1982:69:239-41.
18. Yu-Isenberg KS, Fontes CL, Wan GJ, Geissler EC, Harada AS. Acute and
continuation treatment adequacy with venlafaxine extended release compared
with fluoxetine. Pharmacotherapy. 2004;24(1):33-40.
27. Therneau T, Grambsch P. Modeling Survival Data: Extending the Cox Model.
New York: Springer-Verlag; 2000.
19. Stockl K, Vanderplas AM, Nicklasson L. A comparison of costs for four
oral antidiabetic regimens within a managed care population. Manag Care
Interface. 2003;16(7):31-36.
20. Melikian C, White TJ, Vanderplas A, Dezii CM, Chang E. Adherence to
oral antidiabetic therapy in a managed care organization: A comparison of
monotherapy, combination therapy, and fixed-dose combination therapy.
Clin Ther. 2002;24(3):460-67.
21. Celebrex NDA 20-998, approved December 31, 1998. Available at:
http://www.fda.gov/cder/approval/index.htm. Accessed July 27, 2005.
22. Rubin DB. Estimating causal effects from large data sets using propensity
scores. Ann Intern Med. 1997;127(8S):757-63.
23. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of
classifying prognostic comorbidity in longitudinal studies: development and
validation. J Chron Dis. 1987;40:373-83.
24. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for
use with ICD-9CM administrative databases. J Clin Epidemiol. 1992;45: 613-19.
JMCP
September 2005
Vol. 11, No. 7
26. Grambsch P, Therneau T. Proportional hazards tests and diagnostics based
on weighted residuals. Biometrika. 1994:81:515-26.
28. Mamdani M, Juurlink DN, Kopp A, Naglie G, Austin PC, Laupacis A.
Gastrointestinal bleeding after the introduction of COX 2 inhibitors: ecological
study. BMJ. 2004;328:1415-16.
29. Mamdani M, Rochon PA, Juurlink DN et al. Observational study of upper
gastrointestinal haemorrhage in elderly patients given selective cyclo-oxygenase2 inhibitors or conventional non-steroidal anti-inflammatory drugs. BMJ.
2002;325:624-29.
30. Spiegel BMR, Targownik L, Dulai GS, Gralnek IM. The cost-effectiveness
of cyclooxygenase-2 selective inhibitors in the management of chronic arthritis.
Ann Intern Med. 2003;138:795-806.
31. Pellissier JM, Straus WL, Watson DJ, Kong SX, Harper SE. Economic
evaluation of rofecoxib versus non-selective nonsteroidal anti-inflammatory
drugs for the treatment of osteoarthritis. Clin Ther. 2001;23:1061-79.
32. Feldman M, McMahon AT. Do cyclooxygenase-2 inhibitors provide benefits similar to those of traditional nonsteroidal anti-inflammatory drugs, with
less gastrointestinal toxicity? Ann Intern Med. 2000;132:134-43.
www.amcp.org
ORIGINAL RESEARCH
Relationship of Glycemic Control to Total Diabetes-Related Costs
for Managed Care Health Plan Members With Type 2 Diabetes
SHARASHCHANDRA SHETTY, PhD; KRISTINA SECNIK, RPh, MPH, PhD; and ALAN K. OGLESBY, MPH
ABSTRACT
OBJECTIVE: Glycosylated hemoglobin (A1c) is a well-established measure of
glycemic control, and evidence suggests that maintaining an acceptable A1c
level may be associated with lower treatment costs in adults with diabetes.
Understanding the impact on total treatment costs of staying at the target A1c
level is of great importance to managed care organizations. The goal of this study
was to determine whether type 2 diabetes patients at or below the target A1c
level of 7% had lower diabetes-related costs compared with patients above an
A1c level of 7%.
METHODS: This study was a retrospective database analysis using eligibility data,
medical and pharmacy claims data, and laboratory data from a large U.S. health
care organization. Patients were included in the study if they had 2 or more
claims for type 2 diabetes (International Classification of Diseases, Ninth
Revision, Clinical Modification [ICD-9-CM] codes 250.x0 or 250.x2) and at least
1 A1c value (first such date defined as the index date) during the 12-month period
from January 1, 2002, through December 31, 2002. Patients with 2 or more medical claims for type 1 diabetes (ICD-9-CM codes 250.x1 or 250.x3) were excluded
from the study. Study patients were divided into 2 groups, those at the target A1c
level (≤7%) and those at the above-target A1c level (>7%), and were followed for
a period of 1 year after their index date. Demographic, clinical, and cost variables
were extracted from the administrative database. Multiple linear regression
analysis was used to compare treatment costs between patients at the target
A1c level and patients above target level.
RESULTS: A total of 3,121 patients (46.0%) were identified as being at the target
A1c level, and 3,659 patients (54%) were identified as being above the target A1c
level during the study period. After controlling for confounding factors, the predicted total diabetes-related cost for the above-target group during the 1-year
follow-up period was $1,540 per patient, 32% higher than the total diabetesrelated cost ($1,171) for the at-target group (P <0.001).
CONCLUSION: Results of this analysis suggest that managed care members with
type 2 diabetes who stayed continuously at the target A1c level of 7% or less
over a 1-year follow-up period incurred lower diabetes-related costs compared
with managed care members with type 2 diabetes who were continuously over
the target A1c level of 7%.
KEYWORDS: Type 2 diabetes, A1c level, Glycosylated hemoglobin, Costs, Managed
care, Obesity
D
iabetes is a common chronic disease that is associated with considerable morbidity and mortality.1
Approximately 18 million people in the United States
are diagnosed with diabetes. The American Diabetes
Association (ADA) attributed $92 billion in direct medical
expenditures to diabetes in 2002. Inadequate glycemic control
is thought to be a cause of diabetic complications and higher
costs.2,3 Proper management of diabetes can delay complications,
reduce mortality, and reduce the costs of diabetes care. Research
has shown that aggressive glycemic control can reduce longterm complications in patients with type 1 or type 2 diabetes
and result in considerable medical cost savings.2, 4-6
Most of the previous studies have focused on the effects of
glycemic control on long-term cost savings. However, evidence
of short-term cost savings is often required before studying the
long-term implications. A few studies have suggested that better
glycemic control may result in cost savings within a short period
of time.3,7,8 Analysis of retrospective administrative claims and
laboratory data by Gilmer et al. showed that inadequate glycemic
control was associated with greater health care costs over a
3-year period.3 For every 1% increase in glycosylated hemoglobin
(A1c), Gilmer et al. found that health care costs rose 7% over
the next 3 years. Menzin et al. found, in a study using a retrospective cohort design, a reduced rate of admission for
short-term complications and reduced medical charges for
these complications in patients with better glycemic control.7
In a study using data from a staff-model health maintenance
organization, Wagner et al. suggested that a sustained reduction
in A1c level was associated with significant cost savings within
1 to 2 years of improvement.8 The studies either used change in
A1c level or created categories of A1c levels while studying the
effect of glycemic control on health care costs. In the present
study, we examined the relationship of diabetes-related costs in
patients with type 2 diabetes who stayed at the target A1c level
of ≤7% compared with patients with type 2 diabetes who stayed
above the target A1c level (had A1c levels >7%).
Authors
SHARASHCHANDRA SHETTY, PhD, is a senior researcher, i3 Magnifi, Eden
Prairie, Minnesota; KRISTINA SECNIK, RPh, MPH, PhD, is a senior scientist,
Global Health Outcomes; ALAN K. OGLESBY, MPH, is a senior health outcomes
consultant, U.S. Health Outcomes, Eli Lilly and Company, Indianapolis, Indiana.
AUTHOR CORRESPONDENCE: Sharashchandra Shetty, PhD, Senior Researcher,
i3 Magnifi, 12125 Technology Dr., MN002-0258, Eden Prairie, MN 55344.
Tel: (952) 833-8002; Fax: (952) 833-6045;
www.amcp.org
■■ Methods
This study is a retrospective, longitudinal database analysis
using eligibility data, medical and pharmacy administrative
claims data, and laboratory data from a large U.S. managed care
organization (MCO). All are commercial, preferred-provider
organization-model regional health plans of a national MCO.
The MCO had approximately 5.4 million members during 2002.
The individuals covered by this health plan are geographically
diverse across the United States. The MCO provides fully
Vol. 11, No. 7
September 2005
JMCP
Relationship of Glycemic Control to Total Diabetes-Related Costs for Managed Care Health Plan Members With Type 2 Diabetes
insured coverage for physician, hospital, and pharmacy services.
Study patients were identified during the period January 1,
2002, through December 31, 2002. The index date was the first
available A1c laboratory value recorded during the subject
identification period. Patients were included in this study if
they met the following criteria: (1) had 2 or more claims for
type 2 diabetes in either the primary or secondary position on
physician or hospital claims (International Classification of
Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes
250.x0 or 250.x2 were used to identify patients with type 2
diabetes.); (2) had at least 1 prescription for an oral hypoglycemic agent and/or insulin; (3) had at least 1 available A1c
value; (4) were commercially insured with a drug benefit; and
(5) had at least 6 months of continuous enrollment prior to the
index date and at least 12 months of continuous enrollment
following the index date. Patients with 2 or more claims for
type 1 diabetes (ICD-9-CM codes 250.x1 or 250.x3) were
excluded from the study.
Study patients were divided into those at the target A1c level
(≤7%) and those above the target A1c level (>7%). Patients with
more than 1 A1c value were required to be at target level or
above target level to be included in the at-target group or abovetarget group, respectively. Patients whose A1c values were not
continuously at target level or above target level during the
1-year follow-up period were excluded from the study.
Demographic, clinical, and cost variables were extracted
from the research database for each subject. The demographic
variables included age, gender, and health plan region. The
clinical variables included prescribing physician specialty and
the presence of select comorbid conditions. The following
comorbid conditions were identified during the preindex and
the postindex periods: hypertension, congestive heart failure,
ischemic heart disease, atherosclerosis, dyslipidemia, retinopathy,
nephropathy, neuropathy, diseases of the extremities, obesity, and
albuminuria. The practice specialty of the physician prescribing
the first hypoglycemic agent during the subject identification
period was determined, and 4 dichotomous variables indicating
the specialty of the physicians were created (general practitioner,
internist, endocrinologist, and other specialty).
To account for the difference in burden posed by
comorbidities between the 2 groups, the presence of select
comorbid conditions was examined during the study period.
Comorbid conditions were identified using ICD-9-CM codes in
any position on the physician and hospital claims and using the
National Drug Codes for prescription drugs in the pharmacy
claims. Total baseline cost during the 6-month preindex
period was also estimated to serve as a proxy for general health
status of each subject. The direct medical costs associated with
the treatment of diabetes were estimated 1 year following the
index date. The direct medical costs included all physician
office visit, outpatient visit, inpatient, emergency room, and lab
costs associated with a diagnosis of diabetes in the primary
JMCP
September 2005
Vol. 11, No. 7
position on the medical claims. Pharmacy costs were calculated
for oral hypoglycemic agents and insulin during the 1-year
follow-up period. Costs were defined as the cost to the health
plan as well as to the health plan member; i.e., cost in this study
included the total amount paid by the health plan as well as any
copayment and deductible amounts paid by the health plan
member.
Statistical Analysis
The baseline characteristics of the 2 comparison groups were
analyzed descriptively (frequencies and percentages), using
paired t tests for continuous variables and chi-square tests for
categorical variables. Multiple linear regression analysis was
used to estimate the relationship between remaining at target
A1c level and diabetes treatment cost. The primary independent
variable in the regression model was the target group indicator.
The regression analysis controlled for demographic and clinical
confounders, including age, gender, specialty of the physician
prescribing the index medication, presence of comorbid conditions,
and total baseline costs. The dependent variable in the regression
model was the total diabetes-related costs during the 1-year
follow-up period.
The distribution of the cost data was skewed. Testing for
heteroscedasticity indicated that error variances were not
constant. Logarithmic transformation of the cost data was done
prior to the analysis to make the data normal, and the regression
analysis was conducted using robust standard errors. Since all
patients in the study population had a diabetes-related cost, zero
values for cost were not a concern for this study. The adjusted log
means were transformed to a dollar scale using a smearing
estimator in order to obtain an unbiased estimate of mean
diabetes-related cost.9,10 A subanalysis was also conducted to
identify significant demographic predictors and comorbidities
that are associated with increased diabetes-related costs when
stratified by A1c level. Multiple linear regressions were
conducted separately in the at-target group and above-target
group to identify significant predictors of costs. These regression
models included the same set of independent variables
described above except for the target group indicator. All analyses
were conducted using the SAS software, version 8.2.11
■■ Results
Prior to application of the exclusion criteria, 170,566 patients
were identified with 2 or more claims for type 2 diabetes
during calendar year 2002, representing a prevalence of
approximately 3.1% in the MCO population of 5.4 million
members. After application of the exclusion criteria (Table 1),
8,991 patients were identified with at least 2 or more claims for
type 2 diabetes, at least 1 pharmacy claim for a hypoglycemic
agent, at least 1 laboratory value for A1c, and continuous
enrollment during the study period. Patients with 2 or more
claims for type 1 diabetes were excluded from the study (Table 1).
www.amcp.org
TABLE 1
Sample Selection
Criteria
Number
of Patients
Remaining (%)
Number of
Patients
Dropped (%)
Patients identified with 2 or more claims
for type 2 diabetes from January 1, 2002,
through December 31, 2002
170,566 (100)
–
Patients with 2 or more claims for type 1
diabetes from January 1, 2002,
110,042 (64.5)
60,524 (35.5)
At least 1 claim for oral hypoglycemic
agent or insulin from January 1, 2002,
68,518 (40.2)
41,524 (24.3)
At least 1 lab value for A1c from
January 1, 2002, through
December 31, 2002
31,807 (18.7)
36,711 (21.5)
Continuous enrollment 6 months
prior to index date and 1 year
following the index date
8,991 (5.3)
22,816 (13.4)
Patients not continuously at either
≤7% or >7% A1c level during the
1-year follow-up period
6,780 (4.0)
2,211 (1.3)
A1c = glycosylated hemoglobin.
A total of 2,211 patients were excluded from the study because
they were not continuously either at the target or above the target A1c level during the 1-year follow-up period. Of the 6,780
patients who were included in this study, 3,121 (46%) were
identified as being continuously at the target A1c level, and
3,659 patients (54%) were identified as being above the
target A1c level for the 1-year period following the index A1c
value.
Table 2 shows the descriptive characteristics of the 2 groups.
The at-target group (a) consisted of patients who were significantly
older (mean 53.9 years) as compared with the above-target group
(mean 52.3 years, P <0.001); (b) had a significantly lower
percentage of patients in age groups 31-40 and 41-50 and a higher
percentage of patients in the above-60-years age group (P <0.001);
(c) had a slightly lower proportion of patients from health plans in
the Midwest and a slightly higher proportion of patients from
health plans in the South (P = 0.004); (d) had a higher percentage
of patients who received only oral hypoglycemic agents (P <0.001),
a lower percentage who received only insulin, and a lower
percentage who received a combination of insulin and oral
agents (P <0.001); (e) had a higher percentage of patients with
dyslipidemia (P <0.001) and a lower percentage of patients with
retinopathy (P = 0.007), neuropathy (P = 0.035), or diseases of the
extremities (P = 0.003); and (f) had a lower percentage of patients
who had an internist prescribe the index hypoglycemic medication
(P < 0.001) but a higher percentage of patients who had an
endocrinologist prescribe the index hypoglycemic medication
(P = 0.004).
www.amcp.org
Costs during the 6-month baseline period were significantly
(27%) higher in the at-target group ($2,419) compared with the
above-target group ($1,911, P <0.001). Prior to adjustment of
costs, the at-target group during the follow-up period had
significantly lower medical costs, pharmacy costs, and total
diabetes-related costs compared with the above-target group (all
comparisons, P < 0.001, Table 2).
Table 3 presents the results of multiple regression analysis,
comparing the total diabetes-related costs of the at-target and
the above-target groups. The results of regression analysis are
interpreted in the following way: At an A1c level of 7%, the
expected cost of a subject aged 65 years with hypertension was
higher than the expected cost of a subject aged 30 years without hypertension.
The results of regression analysis revealed that the at-target
group had significantly lower total diabetes costs (P <0.001) as
compared with the above-target group after adjusting for
confounding factors. After appropriate log retransformation
using the smearing estimator, it was found that the predicted
total diabetes-related cost for the above-target group during the
1-year follow-up period was $1,540 per patient, 32% higher
than the total diabetes-related cost ($1,171) for the at-target
group.
Patients in the age groups <31 and 31-40 years had significantly lower diabetes-related costs as compared with patients
older than 60 years. Patients who had an endocrinologist
prescribe the first hypoglycemic medication during the study
period had higher diabetes-related costs. The higher costs
associated with endocrinologists may be related to referral of
more complex cases from primary care physicians to endocrinologists. Patients with a comorbid diagnosis of hypertension,
dyslipidemia, retinopathy, nephropathy, neuropathy, diseases of
the extremities, and obesity had significantly higher diabetesrelated costs.
Since some researchers have expressed concerns with the
use of logged costs related to the potential difficulties associated
with retransformation,12 we also compared the costs using a
gamma distribution with a log link (generalized linear model
[GLM]).13 This method avoids the retransformation problems
associated with log models. The magnitude of costs in the
at-target group and the above-target group using the GLM was
similar to the results obtained using the logged model, thus
providing us confidence in our results.
Table 4 displays the results of the subanalyses that identified
the significant demographic predictors and comorbidities
associated with increased diabetes costs when stratified by A1c
level. Patients in the age groups <31 and 31-40 years in the
at-target group had significantly lower total diabetes costs as
compared with patients older than 60 years. However, the
diabetes-related costs of the <31 and 31-40 years age groups in
the above-target group were not significantly different from the
costs of those older than 60 years, but patients in the age group
Vol. 11, No. 7
September 2005
JMCP
51-60 years had significantly higher total diabetes costs as
compared with patients older than 60 years. Patients with
dyslipidemia, retinopathy, nephropathy, neuropathy, and diseases
of the extremities had significantly higher diabetes costs in both
groups. The major difference in the 2 groups was that obesity
was significant in explaining the higher total diabetes costs in
the above-target group whereas it was not significant in the
at-target group.
■■ Discussion
A retrospective study using eligibility data, medical and pharmacy
claims, and laboratory data for 6,780 patients with type 2
diabetes was conducted to determine the potential economic
benefits associated with glycemic control using the A1c value of
≤7%. The results of this study found a strong association
between glycemic control and total diabetes-related costs.
Patients in the at-target group had significantly lower total
diabetes costs as compared with patients in the above-target
group after adjusting for the demographic characteristics, presence
of comorbid conditions, and baseline costs.
The findings of this study suggest that patients who were at
the target A1c level used fewer health care resources related to
diabetes as compared with patients who were above the target
level, which is reflected in the lower total diabetes costs for the
at-target group patients in the 1-year follow-up period. The
findings were consistent with some of the earlier studies that
studied the relationship between A1c levels and health care
costs. Reduced health care costs among patients with better
glycemic control may be related to symptomatic relief and
improvements in quality of life.14 Wagner et al. suggest that
better glycemic control may increase the comfort of the
primary care physician and the patient, resulting in reduction in
physician visits.8
The at-target-group patients had significantly higher baseline costs as compared with the above-target-group patients.
This finding is consistent with the results of some of the other
studies that found that patients whose A1c levels improved had
increased health care use prior to the decrease in A1c levels.8,15,16
In an effort to keep the health care costs under control, patients
and physicians may have increased the intensity of treatment to
achieve glycemic control.
The importance of the association between obesity and total
diabetes costs in the above-target group is of great interest.
Obesity was significant in explaining the higher diabetes costs
in the above-target-group patients but was not significant in the
at-target group although both groups had a similar prevalence
of obesity. It may be particularly important to focus on reducing
weight in patients who are above the target A1c level.
Given the above findings regarding the higher diabetes costs
associated with patients with A1c levels continuously above
7%, it seems important to focus on getting diabetic patients in
control. Previous research has shown that intervention programs
JMCP
September 2005
TABLE 2
Characteristic
Age† [mean ± SD]
Descriptive Characteristics
of the Study Population
At-Target Group*
n = 3,121
Above-Target Group
n = 3,659
Number (%)
Number (%)
P Values
53.9 [9.1]
52.3 [9.1]
<0.001
Age, years
<31
36 (1.2)
51 (1.4)
0.380
31-40
227 (7.3)
351 (9.6)
<0.001
41-50
739 (23.7)
1,028 (28.1)
<0.001
51-60
1,331 (42.7)
1,530 (41.8)
0.489
>60
788 (25.3)
699 (19.1)
<0.001
Gender
Male
1,792 (57.4)
2,156 (58.9)
0.210
Female
1,329 (42.6)
1,503 (41.1)
Health plan location
Northeast
157 (5.0)
195 (5.3)
0.580
Midwest
727 (23.3)
961 (26.3)
0.004
South
2,043 (65.5)
2,272 (62.1)
0.004
West
194 (6.2)
231 (6.3)
0.869
Type of treatment
Oral agents only
2,954 (94.7)
2,976 (81.3)
<0.001
Insulin only
61 (2.0)
140 (3.8)
<0.001
Oral agents + insulin
106 (3.4)
543 (14.8)
<0.001
Comorbidities
Hypertension
2,444 (78.3)
2,846 (77.8)
0.601
Congestive heart
79 (2.5)
114 (3.1)
0.149
failure
Ischemic heart
495 (15.9)
539 (14.7)
0.197
disease
Atherosclerosis
154 (4.9)
165 (4.5)
0.410
Dyslipidemia
2,442 (78.2)
2,681 (73.3)
<0.001
Retinopathy
332 (10.6)
466 (12.7)
0.007
Nephropathy
96 (3.1)
109 (3.0)
0.816
Neuropathy
272 (8.7)
374 (10.2)
0.035
Diseases of the
40 (1.3)
82 (2.2)
0.003
extremities
Obesity
368 (11.8)
403 (11.0)
0.315
Albuminuria
90 (2.9)
129 (3.5)
0.136
Physician specialty
General practitioner
1,332 (42.7)
1,511 (41.3)
0.250
Internist
986 (31.6)
1,302 (35.6)
<0.001
Endocrinologist
261 (8.4)
239 (6.5)
0.004
Other specialties
542 (17.4)
607 (16.6)
0.395
Baseline cost†
$2,419 [$3,856]
$1,911 [$3,236]
<0.001
[unadjusted]
Outcomes† $
[unadjusted]
Diabetes medical
$534 [$624]
$682 [$1,508]
<0.001
costs
Diabetes pharmacy
$663 [$683]
$924 [$778]
<0.001
costs
Total diabetes costs
$1,197 [$969]
$1,606 [$1,747]
<0.001
* Target group: patients with A1c (glycosylated hemoglobin) value ≤7%; abovetarget group: patients with A1c value >7%.
† For continuous variables, mean and standard deviations are presented by [ ].
Vol. 11, No. 7
www.amcp.org
TABLE 3
Comparison of Total Diabetes-Related Costs
of Patients in the At-Target Group and
Patients in the Above-Target Group
Characteristic
Parameter Estimate (SE)
P Values
At-target group
-0.3044 (0.0189)
<0.001
TABLE 4
Identification of Significant Predictors
of Total Diabetes-Related Costs of Patients
in the At- Target Group and Patients
in the Above-Target Group
At-Target Group*
Age, years
P
Values
Parameter
Estimate (SE)
P
Values
<31
-0.4361 (0.1704)
0.010
-0.0418 (0.0874)
0.632
31-40
-0.1880 (0.0661)
0.004
-0.0929 (0.0539)
0.085
41-50
0.0298 (0.0396)
0.451
-0.0214 (0.0380)
0.573
51-60
<31
-0.2023 (0.0896)
0.023
Characteristic
31-40
-0.1365 (0.0416)
0.001
Age, years
41-50
-0.0031 (0.0274)
0.909
51-60
0.0434 (0.0247)
0.080
Gender: male
0.0242 (0.0191)
0.206
Health plan: Northeast
0.0974 (0.0570)
0.087
Health plan: Midwest
Health plan: South
0.0425 (0.0424)
-0.0209 (0.0401)
0.316
0.602
Family practitioner
0.0418 (0.0285)
0.142
Endocrinologist
0.4266 (0.0382)
<0.001
Internist
Hypertension
Congestive heart failure
Ischemic heart disease
Atherosclerosis
0.0327 (0.0297)
0.270
0.0728 (0.0237)
0.002
-0.0706 (0.0655)
0.280
0.0284 (0.0278)
0.307
-0.0239 (0.0486)
0.622
Dyslipidemia
0.1928 (0.0234)
<0.001
Retinopathy
0.3044 (0.0274)
<0.001
Nephropathy
0.2507 (0.0541)
<0.001
Neuropathy
0.2227 (0.0312)
<0.001
Diseases of the lower extremities
0.3849 (0.0868)
<0.001
Obesity
0.0897 (0.0312)
0.004
Albuminuria
0.0612 (0.0476)
0.198
Baseline cost/100
0.0017 (0.0002)
<0.001
Adjusted R2 (coefficient of determination) = 0.1191.
Reference groups include: above-target group, >60 years, female gender, and health
plan in the Western region.
that facilitate early diagnosis of diabetes, regular monitoring
of each patient’s progress, patient education, and systematic
follow-up will be helpful in attaining glycemic goals and
ultimately will result in costs savings to the health care system.17,18
Proper screening, early diagnosis, and effective management of
diabetes will likely ensure a higher percentage of patients at
target A1c levels. The use of effective drug therapy, along with
patient education and systematic follow-up, may result in
maintaining patients at target levels and may help reduce costs
associated with diabetes treatment.
Limitations
The findings of this study must be considered within the
limitations of the data and study design. First, the observational
study design does not permit causal inference of the results.
www.amcp.org
Above-Target Group
Parameter
Estimate (SE)
0.0146 (0.0357)
0.682
0.0690 (0.0345)
0.045
Gender: male
0.0422 (0.0285)
0.138
0.0090 (0.0259)
0.727
Health plan: Northeast
0.1539 (0.0853)
0.071
0.0594 (0.0766)
0.438
Health plan: Midwest
0.0083 (0.0646)
0.897
0.0756 (0.0561)
0.178
Health plan: South
-0.0188 (0.0602)
0.754
-0.0220 (0.0539)
0.682
Family practitioner
0.0337 (0.0415)
0.417
0.0461 (0.0392)
0.240
Endocrinologist
0.4463 (0.0553)
<0.001
0.4200 (0.0521)
<0.001
Internist
0.0372 (0.0443)
0.401
0.0252 (0.0402)
0.530
Hypertension
0.0666 (0.0333)
0.046
0.0710 (0.0334)
0.033
Congestive heart failure -0.0676 (0.1061)
0.524
-0.0870 (0.0834)
0.297
Ischemic heart disease
0.0258 (0.0421)
0.539
0.0294 (0.0370)
0.426
Atherosclerosis
-0.0447 (0.0728)
0.538
Dyslipidemia
0.1515 (0.0369)
<0.001
Retinopathy
0.3077 (0.0393)
Nephropathy
0.2422 (0.0835)
Neuropathy
Diseases of the
lower extremities
0.0002 (0.0650)
0.996
0.2237 (0.0301)
<0.001
<0.001
0.3030 (0.0376)
<0.001
0.003
0.2576 (0.0704)
<0.001
0.1557 (0.0478)
0.001
0.2709 (0.0412)
<0.001
0.4228 (0.1238)
<0.001
0.3563 (0.1143)
0.001
Obesity
0.0165 (0.0474)
0.727
0.1549 (0.0409)
<0.001
Albuminuria
0.1209 (0.0772)
0.117
0.0117 (0.0611)
0.848
Baseline cost/100
0.0013 (0.0004)
0.001
0.0022 (0.0005)
<0.001
* Target group: patients with A1c (glycosylated hemoglobin) value ≤7%; above-target
group: patients with A1c value >7%.
Adjusted R2 (coefficient of determination) for at-target group=0.0762; adjusted R2 for
above-target group=0.1071.
Reference groups include: >60 years, female gender, and health plan in the Western region.
Further, it could be determined whether patients remained at
the target A1c level only if they had follow-up visits with their
physicians. We made the assumption that the proportion of
patients scheduling a follow-up visit in the above-target group
is distributed similar to the at-target group. Patients who
achieved their target A1c level are probably less likely to have a
follow-up visit scheduled as compared with patients who were
above target level. We may, therefore, have overestimated the
number of patients at target if the number of follow-up visits
decreased dramatically after the target level was achieved.
However, a post hoc analysis revealed that the number of
Vol. 11, No. 7
September 2005
JMCP
follow-up visits was higher among the at-target-group patients
as compared with the above-target-group patients, thus limiting
the possibility of overestimation of the number of patients at
target (data not presented).
The present study is limited by the inability to control for
the number of A1c tests, the number of diabetes-related ambulatory visits, and the type of oral hypoglycemic agent.
Reporting the values for these 3 variables would have provided
a more comprehensive description of the study population. The
study is also limited by the lack of control for the severity of
disease, but the baseline costs may be a crude proxy for disease
severity.
We measured only diabetes-related costs, defined as (a) the
health plan and member costs for only medical claims with a
primary diagnosis of diabetes and (b) pharmacy costs for
diabetes drugs and insulin. We did not measure total medical
costs for these 2 groups of patients with type 2 diabetes.
Patients were excluded from the study if their A1c values
were not continuously at target level or above target level during the 1-year follow-up period. Since this excluded group of
2,211 patients represented 25% of the potential study population,
the results of the study may not be generalizable to the general
population of patients with type 2 diabetes.
The study population was a sample of managed care members.
Therefore, the results of this study are applicable to the
management of type 2 diabetes in a managed care setting.
It may not be possible to generalize the results to a nonmanaged
care population. In addition, administrative claims data have
some limitations that are common to all analyses conducted
using this data source. Claims data are collected for the purpose
of payment and not for research. They may be subject to possible
coding errors. Despite the limitations, administrative claims
remain a powerful source of data for research. Claims data allow
for examination of health care utilization and associated expenditures in real-world settings.
This analysis has important policy implications. The analysis
suggests that better glycemic control is associated with significantly lower costs over a 1-year period. Further research should
be conducted to explore if glycemic control is associated with
cost savings outside the managed care environment.
■■ Conclusion
Diabetes patients who were continuously at the target A1c level
of ≤7 had significantly lower diabetes-related costs over a 1-year
follow-up period compared with diabetes patients who were
continuously above the target A1c level. Managed care efforts to
help diabetes patients attain target A1c levels may reduce total
diabetes-related medical costs.
DISCLOSURES
Funding for this research was provided by Eli Lilly and Company and was
obtained by authors Kristina Secnik and Alan K. Oglesby, who are employees
and shareholders of Eli Lilly and Company. Sharashchandra Shetty is an
JMCP
September 2005
Vol. 11, No. 7
employee of i3 Magnifi and a shareholder of UnitedHealth Group; i3 Magnifi,
a division of UnitedHealth Group, receives funding from several pharmaceutical companies for outcomes research studies. The study results in this article
were presented as a poster at the International Society for Pharmacoeconomics
and Outcomes Research 10th Annual International Meeting, Washington, DC,
on May 17, 2005.
Shetty served as principal author of the study. Study concept and design,
analysis and interpretation of data, and statistical expertise were contributed
by all authors. Drafting of the manuscript was primarily the work of Shetty,
and its critical revision was the work of Secnik and Oglesby. Administrative,
technical, and/or material support was provided by i3 Magnifi.
REFERENCES
1. American Diabetes Association. Diabetes statistics. Available at:
http://www.diabetes.org/diabetes-statistics/national-diabetes-fact-sheet.jsp.
Accessed November 3, 2004.
2. The Diabetes Control and Complications Trial Research Group. The effect
of intensive treatment of diabetes on the development and the progression of
long-term complications in insulin-dependent diabetes mellitus. N Engl J Med.
1993;329:977-86.
3. Gilmer TP, O’Connor PJ, Manning WG, Rush WA. The cost to health plans
of poor glycemic control. Diabetes Care. 1997;20:1847-53.
4. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose
control with sulphonylureas or insulin compared with conventional treatment
and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet.
1998;352:837-53.
5. Eastman RC, Javitt JC, Herman WH, et al. Model of complications of
NIDDM, II: analysis of health benefits and cost-effectiveness of treating
NIDDM with the goal of normo-glycemia. Diabetes Care. 1997;20:735-44.
6. Brown JB, Pedula KL, Bakst AW. The progressive cost of complications in
type 2 diabetes mellitus. Arch Intern Med. 1999;159:1873-80.
7. Menzin J, Langley-Hawthorne C, Friedman M, Boulanger L, Cavanaugh R.
Potential short-term economic benefits of improved glycemic control: a managed care perspective. Diabetes Care. 2001;24:51-55.
8. Wagner EH, Sandhu N, Newton KH, McCulloch DK, Ramsey SD, Grothaus
LC. Effect of improved glycemic control on health care costs and utilization.
JAMA 2001;285(2):182-89.
9. Duan N. Smearing estimate: a nonparametric retransformation method.
J Am Stat Assoc. 1983;78:605-10.
10. Rutten-van Molken MP, van Doorslaer EK, van Vilet RC. Statistical analysis
of cost outcomes in a randomized controlled clinical trial. Health Econ. 1994;
3:333-45.
11. SAS statistical software, Version 8.2, 2001; SAS Institute, Inc., Cary, NC.
12. Manning WG. The logged dependent variable, heteroscedasticity, and the
retransformation problem. J Health Econ. 1998;17(3):283-95.
13. Blough DK, Madden CW, Hornbrook MC. Modeling risk using generalized
linear models. J Health Econ. 1999;18(2):153-71.
14. Testa MA, Simonson DC. Health economic benefits and quality of life during improved glycemic control in patients with type 2 diabetes mellitus: a randomized, controlled, double-blind trial. JAMA. 1998;280:1490-96.
15. Wagner EH, Curry SJ, Grothaus L, Saunders KW, McBride CM. The
impact of smoking and quitting on health care use. Arch Intern Med. 1995;
155:1789-95.
16. O’Connor PJ, Crabtree BF, Abourizk NN. Longitudinal study of a diabetes
education and care intervention: predictors of improved glycemic control.
J Am Board Fam Pract. 1992;5:381-87.
17. The Expert Committee on the Diagnosis and Classification of Diabetes
Mellitus: Report of the Expert Committee on the Diagnosis and Classification
of Diabetes Mellitus. Diabetes Care. 1997;20:1183-97.
18. Solberg LI, Reger LA, Pearson TL, et al. Using continuous quality
improvement to improve diabetes care in populations: the IDEAL model.
Improving care for Diabetics through Empowerment Active collaboration and
Leadership. Jt Comm J Qual Improv. 1997;23(11):581-92.
www.amcp.org
ORIGINAL RESEARCH
Economic Burden of Anemia in an Insured Population
ALLEN R. NISSENSON, MD, FACP; SALLY WADE, MPH; L. TIM GOODNOUGH, MD;
KEVIN KNIGHT, MD, MPH; and ROBERT W. DUBOIS, MD, PhD
ABSTRACT
OBJECTIVE: Anemia is a common hematological disorder characterized by
reduced hemoglobin concentrations. Despite information on prevalence and
associated outcomes, little is known about the impact of anemia on health care
utilization and costs. This study examines anemia prevalence and associated
medical costs and utilization, using administrative claims for adults newly
diagnosed with anemia, including up to 12 months of follow-up.
METHODS: Patients predisposed to anemia, based on selected comorbid conditions
(chronic kidney disease, human immunodeficiency virus, rheumatoid arthritis,
inflammatory bowel disease, congestive heart failure, and solid-tumor cancers),
were identified. Costs for anemic patients and a random sample of nonanemic
patients with these conditions were compared. Associations were evaluated after
adjustment for potential confounders using a regression model. Clinical care
patterns were examined overall and by condition.
RESULTS: Anemia was observed in 3.5% (81,423) of approximately 2.3 million
health plan members in 2000; 15% of anemic patients received an identified
treatment, with transfusion being the most frequent intervention. Utilization and
costs were significantly higher for anemic patients (P < 0.001). Average annualized
per-patient costs were $14,535 for anemic patients (55% outpatient, 33% inpatient,
13% pharmacy), 54% higher than the $9,451 average cost for nonanemic
patients (45% outpatient, 36% inpatient, 19% pharmacy). After adjustment for
age, other comorbidities (e.g., chronic kidney disease and cancer), sex, and
insurance type (indemnity, preferred provider organization/point of service, or
health maintenance organization, in the Medstat MarketScan database), anemic
patients had average costs that were more than twice the adjusted costs of
nonanemic patients.
CONCLUSION: Medical costs for anemic patients are as much as twice those for
nonanemic patients with the same comorbid conditions.
KEYWORDS: Anemia, Chronic conditions, Utilization, Medical costs
Authors
ALLEN R. NISSENSON, MD, FACP, is a professor of medicine, associate dean for
special projects, and director of the dialysis program, Division of Nephrology, David
Geffen School of Medicine, University of California, Los Angeles; SALLY WADE,
MPH, is a consultant, Wade Outcomes Research and Consulting, Salt Lake City,
Utah; KEVIN KNIGHT, MD, MPH, is a consultants, Santa Monica, California;
L. TIM GOODNOUGH, MD, is a professor of pathology, Department of Pathology
and Medicine, Stanford University School of Medicine, Stanford, California;
ROBERT W. DUBOIS, MD, PhD, is senior vice president, Cerner Health Insights,
Beverly Hills, California.
AUTHOR CORRESPONDENCE: Allen R. Nissenson, MD, FACP, Director, Dialysis
Program, Division of Nephrology, David Geffen School of Medicine, UCLA, 200
Medical Plaza, Suite 565-59, Los Angeles, CA 90095-6945. Tel: (310) 825-9464;
Fax: (310) 206-2985; E-mail: [email protected]
www.amcp.org
A
nemia is an important public health concern. It occurs
commonly and is characterized by reduced concentrations
of hemoglobin due to a variety of underlying causes.1,2
Estimates of anemia prevalence vary considerably. The National
Center for Health Statistics conservatively estimates that
approximately 3.4 million individuals in the United States are
anemic.3
The National Health and Nutrition Examination Survey
(NHANES) provides an assessment of anemia prevalence based on
laboratory testing. Results from NHANES III show anemia to be
most prevalent in children through age 16, women aged 17 to 49
years, and the elderly (aged 75 years and older), especially
elderly men.4 Anemia prevalence is higher among individuals
with certain chronic conditions, including chronic kidney
disease (CKD), human immunodeficiency virus (HIV), rheumatoid
arthritis (RA), inflammatory bowel disease (IBD), congestive heart
failure (CHF), and cancer.2,5 In previous studies, anemia with
chronic disease has been identified in 36% of patients with
CKD6 and 27% of patients with RA.7 The increasing amount of
information on the prevalence of anemia among individuals
with such diseases supports the view that it is a condition
of growing concern.7-11
Anemia has been shown to be associated with increased
mortality and morbidity as well as with decreased physical
functioning and quality of life.2,12 Although anemia is often
associated with disease progression or increased disease severity,
evidence of its independent effect on these key outcomes is still
accumulating.
Despite information on anemia prevalence and associated
outcomes, little is known about the impact of anemia on health
resource utilization and costs. Although the cost of anemia care
has been examined using Medicare data for CKD,13 cancer,14,15
and heart failure,16 the literature provides no data on anemiarelated costs in other populations. Consequently, this study was
undertaken to estimate the health care costs and treatment
patterns of patients with anemia in a privately insured population.
■■ Methods
This study is based on retrospective administrative claims data
from commercially insured and Medicare plans represented in
the Medstat MarketScan database, containing the combined
administrative claims for more than 2 million health plan members
for employers from across the country. The study population
was selected only from those enrolled in plans with complete
capture of facility-based and professional services, as well as
outpatient prescription medications. The combined Medicare
and commercial populations from which the study sample was
Vol. 11, No. 7
September 2005
JMCP
TABLE 1
Diagnosis Codes (ICD-9-CM) Used to Identify 6 Study Conditions
Chronic kidney disease
Malignant hypertensive renal disease (without renal failure)
403.00
Malignant hypertensive renal disease (with renal failure)
403.01
Benign hypertensive renal disease (without renal failure)
403.10
Benign hypertensive renal disease (with renal failure)
403.11
Unspecified hypertensive renal disease (without renal failure)
403.90
Unspecified hypertensive renal disease (with renal failure)
403.91
Malignant hypertensive heart and renal disease (with renal failure)
404.02
Malignant hypertensive heart and renal disease
(with heart and renal failure)
404.03
Benign hypertensive heart and renal disease (with renal failure)
404.12
Benign hypertensive heart and renal disease
404.13
Unspecified hypertensive heart and renal disease
(with renal failure
404.92
404.93
Nephrotic syndrome
581.0-581.9
Chronic glomerulonephritis
582.0-582.9
Nephritis (NOS as acute or chronic)
583.0-583.9
Chronic renal failure
585
Renal failure, unspecified
586
Renal sclerosis, unspecified
587
Chronic pyelonephritis
(without lesion of renal medullary necrosis)
590.00
Chronic pyelonephritis (with lesion of renal medullary necrosis)
590.01
HIV infection
HIV infection
042
Rheumatoid arthritis
Felty's syndrome
Other rheumatoid arthritis with visceral or systematic involvement
Rheumatoid lung
Other
714.0
714.1
714.2
714.81
714.89
Inflammatory bowel disease
Small intestine
Large intestine
Small intestine with large intestine
Unspecified site
Ulcerative (chronic) enterocolitis
Ulcerative (chronic) ileocolitis
Ulcerative (chronic) proctitis
Ulcerative (chronic) proctosigmoiditis
Left-sided ulcerative (chronic) colitis
Universal ulcerative (chronic) colitis
Other ulcerative colitis
Ulcerative colitis, unspecified
Malignant hypertensive heart disease (with congestive heart failure)
Benign hypertensive heart disease (with congestive heart failure)
Unspecified hypertensive heart disease (with congestive heart failure)
(with congestive heart failure)
(with congestive heart failure and renal failure)
555.0
555.1
555.2
555.9
556.0
556.1
556.2
556.3
556.5
556.6
556.8
556.9
428.0
402.01
402.11
402.91
404.01
404.03
404.11
404.13
404.91
404.93
Cancer (primary only)
Cancer diagnoses in the range of 140.0-199.x, excluding 173.0-173.9. For
cancer patients, we required 2 outpatient diagnoses on separate service dates
within 6 months or 1 diagnosis on an inpatient admission record.
ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification; NOS = not otherwise specified.
drawn resided primarily in the southern, northern central, and
northeastern regions of the United States, with a smaller
representation from the western region.
Analyses were conducted to examine the prevalence of
anemia and related utilization and health plan costs in an adult
population. This paper presents results for the entire study
population and separately for groups with specific conditions
that are often associated with an increased occurrence of
anemia or in which anemia presents particular clinical
challenges. For condition-specific subgroup results, patients
were identified based on the presence of diagnosis codes from
the International Classification of Diseases, 9th Revision, Clinical
Modification, (ICD-9-CM) for 6 conditions: CKD, HIV, RA, IBD,
CHF, and solid-tumor cancer (Table 1). Patients who had
multiple diagnoses during the study period were included in all
condition-specific groups for which they qualified.
Since laboratory values such as hemoglobin levels are not
JMCP
September 2005
Vol. 11, No. 7
generally available in medical claims data, anemia was identified by
the presence of at least 1 diagnosis code for anemia (occurring
in any position on a claim), 1 procedure code (Current
Procedural Terminology, 4th Edition [CPT-4], ICD-9-CM, or
Health Care Financing Administration Common Procedure
Coding System [HCPCS]), or 1 drug code indicative of anemia
treatment (e.g., blood transfusions, injections with recombinant
erythropoietin. Diagnoses used for patient selection include
iron deficiency anemia, pernicious anemia, anemia of chronic
disease, nutritional anemia, other specified aplastic anemias,
and other unspecified anemia. Codes for acute anemias were
not included, and blood transfusion was only considered in the
absence of a diagnosis of acute anemia (ICD-9-CM 285.1)
(Table 2).
The study was divided into 2 components. The first component
assessed anemia prevalence in the year 2000. The denominator
included all adult health plan members who had continuous
www.amcp.org
Diagnosis Codes (ICD-9-CM) and
Procedure Codes (CPT-4, ICD-9-CM,
HCPCS) Used to Identify Anemia
Iron deficiency anemia
280.x: Iron deficiency anemia
Anemia in chronic illness
285.2x: Anemia in chronic illness
Pernicious anemia
281.0 Pernicious anemia
Other anemia
The “other anemia” category includes patients with evidence of anemia diagnoses other than those listed above, plus patients with evidence of anemia
treatment but no corresponding diagnosis.
281.1-281.9: Other nutritional anemias
285.9: Anemia, unspecified
284.8: Other specified aplastic anemias
Transfusion
V58.2, 99.0x (ICD-9-CM); 36430, 36440, 96400-96549 (CPT-4); P9010,
P9011, P9012, P9013, P9016, P9017, P9018, P9019, P9020, P9021, P9022,
P9023 (HCPCS)
Note: Only patients with the above transfusion codes but no acute anemia
(DX = 285.1) in both the 12-month period preceding the transfusion and during the
follow-up period were included in the anemic population.
Epoetin alfa injection
HCPCS:
Q9920, Q9921, Q9922, Q9923, Q9924, Q9925, Q9926, Q9927, Q9928,
Q9929, Q9930, Q9931, Q9932, Q9933, Q9934, Q9935, Q9936, Q9937,
Q9938, Q9939, Q9940, Q0136
NDC:
55513014401, 55513047810, 55513047801, 55513028301, 55513026701,
55513014810, 55513082301, 55513014410, 55513028310, 55513012610,
55513012601, 05551326710, 05551314810, 05551314410, 05551312610,
55513014801, 55513026710, 55513082310, 59676031200, 00062740103,
00062740201, 00062740501, 00403489718, 59676740104, 59676740000,
59676034001, 59676032001, 54868252300, 59676031201, 59676031002,
59676031001, 59676030402, 59676030401, 59676030302, 00062740003,
59676030301, 59676740503, 59676030202, 59676030201, 54868252301
CPT-4 = Current Procedural Terminology, 4th Edition; DX = Diagnosis;
HCPCS = Health Care Financing Administration Common Procedure Coding System;
ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical
Modification; NDC = National Drug Code.
medical and drug benefits coverage during 2000, and the
numerator included all members with evidence of a specified
anemia diagnosis or treatment. The second component used
administrative claims for dates of service between January 1,
1998, and June 30, 2001, to compare the health care cost and
utilization patterns between anemic and nonanemic patients
within the 6 study conditions. We captured the sequence of
anemia-related services in the year following the initial diagnosis.
For the second part of the study, we selected only those patients
who were “newly diagnosed” with anemia, defined as those with
at least 1 year of continuous medical and drug benefits coverage prior to their anemia index date (date of first anemia diagnosis or procedure in the study period [Table 2]) and no
evidence of anemia diagnoses or treatment during this 1-year
“history” period. For comparison, we selected patients who met
the same health plan enrollment requirements as the anemic
patients but had no evidence of anemia. We identified a
comparison group for the overall anemic population and also
constructed 6 condition-specific comparison groups using the
www.amcp.org
FIGURE 1
Year 2000 Anemia Prevalence
Overall and by Condition
40
35
33.5%
30
% of Patients
TABLE 2
25
21.4%
20
17.8%
13.2%
15
11.9%
10
10.2%
3.5%
5
0
CKD
Cancer
CHF
IBD
RA
HIV
Overall
Condition
Overall n=81,423 (out of overall Medstat MarketScan research database population=
2,296,832); CKD=4,834; cancer=14,023; CHF=7,234; IBD=1,349; RA=1,885;
HIV=151. Percentages for each condition are percentages of the anemic population.
N values for anemia prevalence by condition and age:
CKD: age 18-49 = 378; age 50-64 = 771; age 65+ = 1,160
Cancer: age 18-49 = 1,201; age 50-64 = 2,784; age 65+ = 2,768
CHF: age 18-49 = 216; age 50-64 = 804; age 65+ = 2,726
IBD: age 18-49 = 296; age 50-64 = 283; age 65+ = 213
RA:
age 18-49 = 326; age 50-64 = 550; age 65+ = 572
HIV: age 18-49 = 26;
age 50-64 = 7;
age 65+ = 2
CKD = chronic kidney disease; CHF = congestive heart failure; HIV = human immunodeficiency virus; IBD = inflammatory bowel disease; RA = rheumatoid arthritis.
same diagnostic criteria that we used for the anemic patients.
Since the objective of this study component was to characterize anemia care in the first year after diagnosis, follow-up
data were examined for a maximum of 12 months after the
anemia index date. Individual follow-up periods were
determined by the amount of time that each patient had
continuous benefits coverage following the anemia index date.
To avoid skewing the study toward a sicker population, we did
not require that patients remain in the health plan for the full
12 months of potential follow-up; this variable follow-up
period was taken into consideration in the analyses.
For the cost model, we developed a variable to adjust for
disease severity for each of the 6 conditions. Severity adjustment
was based on specific ICD-9 codes, HCPCS codes for durable
medical equipment, or pharmacy codes. Patients were separated
into mild, moderate, and severe categories based on specific
ICD-9 codes, HCPCS codes for durable medical equipment, or
pharmacy codes. Cancer patients who were actively receiving
chemotherapy were categorized as part of the moderate severity
category, whereas those with evidence of metastasis were
categorized as severe. CKD patients with 1 CKD hospitalization
during follow-up were categorized as moderate while those
with either more than 1 CKD hospitalization or with a kidney
transplant were categorized as severe. Use of biologic therapies
Vol. 11, No. 7
September 2005
JMCP
FIGURE 2
Year 2000 Anemia Prevalence in Females
by Condition and Age
40
■ Age 18-49
■ Age 50-64
■ Age 65+
35
% of Patients
30
25
20
15
10
5
0
CKD
Cancer
CHF
IBD
Condition
RA
HIV
CKD: age 18-49 = 378; age 50-64 = 771; age 65+ = 1,160
Cancer: age 18-49 = 1,201; age 50-64 = 2,784; age 65+ = 2,768
CHF: age 18-49 = 216; age 50-64 = 804; age 65+ = 2,726
IBD: age 18-49 = 296; age 50-64 = 283; age 65+ = 213
RA:
age 18-49 = 326; age 50-64 = 550; age 65+ = 572
HIV: age 18-49 = 26;
age 50-64 = 7;
age 65+ = 2
CHF = congestive heart failure; CKD = chronic kidney disease; HIV = human immunodeficiency virus; IBD = inflammatory bowel disease; RA = rheumatoid arthritis.
FIGURE 3
Year 2000 Anemia Prevalence in Males
by Condition and Age
40
■ Age 18-49
■ Age 50-64
■ Age 65+
35
% of Patients
30
25
20
15
10
5
0
CKD
Cancer
CHF
IBD
Condition
RA
HIV
CKD: age 18-49 = 280; age 50-64 = 763; age 65+ = 1,482
Cancer: age 18-49 = 298; age 50-64 = 1,774; age 65+ = 5,198
CHF: age 18-49 = 105; age 50-64 = 708; age 65+ = 2,675
IBD: age 18-49 = 152; age 50-64 = 211; age 65+ = 194
RA:
age 18-49 = 55; age 50-64 = 167; age 65+ = 215
HIV: age 18-49 = 64; age 50-64 = 43;
age 65+ = 9
CHF = congestive heart failure; CKD = chronic kidney disease; HIV = human immunodeficiency virus; IBD = inflammatory bowel disease; RA = rheumatoid arthritis.
JMCP
September 2005
Vol. 11, No. 7
and selected nonbiologic therapies indicated moderate RA
while joint surgery indicated severe RA. For IBD, more
advanced medications, surgery, and multiple hospitalizations
were indicators of severe disease. For CHF, the severity
increased as the number of concomitant CHF medications and
hospitalizations increased.
The Charlson Comorbidity Index (CCI) was used to quantify
burden of illness in the study population based on claims
incurred during the 6 months before each patient’s index date.17
In order to assess treatment patterns, we identified specific
procedures and medications that are commonly used in the
management of anemia. These included blood transfusions,
erythropoietin injections, B12 injections, iron injections, and
use of testosterone, nandrolone, folate, or folic acid. It is important
to note that the study database captured information only on
outpatient prescription medications and did not include
utilization of any inpatient or over-the-counter medications.
Anemia treatment regimens were assessed for up to 1 year
following patients’ anemia index dates; follow-up periods ended
either at the conclusion of the study period or when patients left
the health plan. All relevant therapies provided on or after the
anemia index date were counted.
The costs evaluated in this study were the payments made
by the health plan, after subtraction of member cost-share, as
reported on the final adjudicated version of each claim. No
adjustments were made to standardize costs across the study
period. Unfortunately, the structure and level of detail of administrative claims precluded us from simply summing up
payments in order to determine the cost of anemia. As is
common in cost studies such as this one, we created an
algorithm to estimate the direct health plan payments attributable
to anemia.14-16 For inpatient and outpatient (nonpharmacy)
services, claims with either a primary or secondary diagnosis
resulted in the attribution of a portion of the costs to anemia. In
general, if anemia was listed as the primary diagnosis, 50% of the
costs on the claim were attributed to anemia. If anemia was
listed only as a secondary diagnosis, 25% of the costs were
attributed to anemia.
Allocations for individual claims ranged from 0% to 100%,
depending on whether the anemia diagnosis was primary or
secondary, how many additional diagnoses were on the claims,
and whether anemia-specific services (e.g., erythropoietin injection)
appeared on the claim. In addition, costs for anemia-specific
procedure codes were attributed to anemia even when anemia
was not listed explicitly as a diagnosis (e.g., transfusion).
For outpatient pharmacy claims, all erythropoietin costs were
attributed to anemia.
Pairwise comparisons were performed for each variable
according to the nature of the data involved: continuous
variables were compared using t tests or nonparametric equivalents,
and categorical variables were compared using chi-square tests.
A multivariate analysis was also conducted to estimate cost
www.amcp.org
TABLE 3
Number of Patients and Average Follow-Up Months by Condition and Anemia Status
Based on Claims Incurred January 1, 1998, to June 30, 2001
Newly Diagnosed Anemic Patients
Overall population
N (Average
Follow-up Months)
Average Age
118,332 (8.9)
56.9
7,545 (9.0)
65.8
Nonanemic Comparison Patients
% Female
N (Average
Follow-up Months)
Average Age
% Female
66.3
35,948 (9.3)
61.6
53.0
49.3
5,814 (9.3)
62.7
44.4
Condition-Specific Populations
Human immunodeficiency virus
354 (8.7)
45.9
31.9
232 (9.1)
44.5
29.7
3,852 (9.1)
61.2
76.1
3,303 (9.2)
59.8
71.6
2,538 (9.3)
55.9
61.5
2,139 (9.3)
53.8
55.1
14,985 (8,7)
72.5
53.5
11,886 (9.2)
71.3
49.5
Cancer
22,030 (8.7)
65.7
51.6
17,542 (9.3)
65.3
53.0
differences between anemic and nonanemic patients, adjusting
for factors likely to influence health care utilization and expenditures. An exponential model was fit using a generalized linear
modeling technique, with patient age and gender, coverage type
(e.g., preferred provider organization, indemnity), predisposing
condition (i.e., the 6 study conditions), and disease severity for
each predisposing condition as covariates and a binary indicator
variable for presence or absence of anemia. Due to the skewed
nature of distributions of payment data, a gamma variance function
was chosen using the Park test, and bootstrap standard errors
were estimated. All analyses were conducted using SAS software
version 8.02 (Cary, NC) and STATA version 7.0 (College
Station, TX).
■■ Results
Based on data for the 2,296,832 adult health plan members
with continuous benefits coverage during 2000, the overall
anemia prevalence was 3.5% (81,423) (Figure 1). Although
statistical comparisons cannot be made because the condition
groups were not mutually exclusive, it is clear that the prevalence of anemia varied significantly by condition, with CKD
defining the upper end. Within each of the 6 study conditions,
the relationship between anemia and age among females was
not consistent. However, among males, anemia prevalence
increased with age (Figures 2 and 3).
Overall, 118,332 anemic patients and a random sample of
35,948 nonanemic patients were identified for inclusion in
the cost and utilization component of the study (Table 3).
(Case-matching was determined to be unnecessary in order to
provide reasonably precise adjusted measures of association
since the size of the study population was large). The number
of patients in the condition-specific subgroups ranged from
354 anemic patients and 232 nonanemic patients in the HIV
subgroup to 22,030 anemic patients and 17,542 nonanemic
www.amcp.org
patients in the cancer subgroup (Table 3). Females made up the
majority of both the overall anemic and control populations (66%
and 53%, respectively), and the proportion of females in the anemic
population was statistically higher (P<0.001). With the exception of
the HIV population, females were more common in the anemic
populations for all study conditions (P < 0.01).
Overall, nonanemic patients were nearly 5 years older on
average compared with the anemic patients (61.6 years vs. 56.9
years, P < 0.001). The opposite was true in the condition-specific
populations: anemic patients were older, on average, than
nonanemic patients in all of the condition-specific populations
except for HIV (P < 0.002). With respect to the CCI, the overall
population of anemic patients did not differ statistically from
nonanemic comparison subjects (P = 0.22). However, in each of
the 6 condition-specific populations, anemic patients had a
statistically higher burden of illness as evidenced by higher CCI
scores (P < 0.001). The difference in CCI scores was greatest in
patients with CKD (1.6 for anemic vs. 0.98 for nonanemic) and
those with cancer (1.6 for anemic and 0.56 for nonanemic).
Anemia Management
The average follow-up period for patients in this study was
approximately 9 months.
Overall Population
In the overall population, the majority of anemic patients
(86.5%) did not receive any of the therapies evaluated in this
study (Figure 4). Among therapies evaluated, transfusion was
the most commonly used, with nearly 1 out of every 10 anemic
patients (9.3%) receiving at least 1 blood transfusion during the
follow-up period. These patients averaged 1.1 transfusions per
month. For most transfused patients, transfusion was the only
therapy used; approximately 1 in 5 also received erythropoietin,
which was nearly always given after the transfusion.
Vol. 11, No. 7
September 2005
JMCP
FIGURE 4
lowest average number of transfusions per patient per month
(0.8). Therapy involving both transfusion and erythropoietin
was most common among anemic HIV patients (27.0% of
transfused patients) and least common among anemic RA
patients (15.5% of transfused patients). Nearly 1 in 5 transfused
CKD patients (19.1%) and cancer patients (19.2%) also
received erythropoietin.
Among condition-specific anemic populations, erythropoietin
was used most commonly by HIV patients (4.5%) and least
commonly by RA patients (2.8%). The majority of erythropoietin
use in these populations was in conjunction with (usually
following) blood transfusion.
Anemia Management
Strategies by Condition
16
14
% of Patients
12
10
8
6
4
2
0
All
CKD
Treatment
Cancer
All
CHF
Condition
IBD
CKD Cancer CHF
RA
HIV
IBD
RA
HIV
■■
Any
118,322 7,545 22,030 14,985 2,539 3,852 354
■■
Transfusion
11,017
738
2,130
1,528
257
349
37
■■
EPO
3,743
271
739
527
86
109
16
■■
B12 injections
3,146
213
599
438
61
94
6
■■
Iron injections
238
11
40
22
1
6
2
CHF = congestive heart failure; CKD = chronic kidney disease; EPO = erythropoietin;
HIV = human immunodeficiency virus; IBD = inflammatory bowel disease;
RA = rheumatoid arthritis.
TABLE 4
Average Annualized Utilization of Key
Services per Patient Based on Claims Incurred
January 1, 1998, to June 30, 2001
Total Population
Service
Anemic
N = 118,332
(Average
Follow-up
Months = 8.9)
Nonanemic
N = 35,948
(Average
Follow-up
Months = 9.3)
P Value
(2-Tailed
Student’s
t Test)
Outpatient visits
6.1
5.3
<0.001
Emergency room visits
0.2
0.1
<0.001
Inpatient admissions
0.4
0.3
<0.001
Hospital days
3.1
2.0
<0.001
Laboratory tests
5.3
2.6
<0.001
Condition-Specific Populations
As in the overall population, the majority of patients (85.2%86.9%) in each of the 6 condition-specific populations had no
documented anemia treatment. In these populations, the
pattern of anemia treatment was similar to that in the overall
population. Anemic HIV patients had the highest transfusion
use (10.5% with at least 1 transfusion), but they also had the
JMCP
September 2005
Vol. 11, No. 7
Utilization and Costs
Utilization of selected key services was significantly higher for
anemic patients (P < 0.001 in all cases), as presented in Table 4.
Similarly, per-patient payments for health care services (Tables
5 and 6) were higher for anemic patients than for nonanemic
patients, both overall and within each of the 6 study conditions
(P < 0.001). With the exception of outpatient pharmacy-based
prescription drugs, anemic patients exhibited higher costs than
did nonanemic patients for all types of care, including inpatient,
outpatient, emergency room, and outpatient laboratory
(P < 0.001). With the exception of outpatient facility care for
HIV patients, this pattern of higher costs (including, in this
case, higher outpatient prescription costs) among anemic
patients persisted in the condition-specific populations (P < 0.03).
Among anemic patients, average total annualized costs were
$14,535 per patient. Outpatient care, including physician office
visits, accounted for more than half ($7,927, 54.5%) of the
average total costs. Inpatient care accounted for nearly one third
($4,775, 33%) of the average total costs. Payments for pharmacybased outpatient prescriptions averaged $1,833. In the nonanemic
population, average total annualized costs were $9,450 per
patient. Outpatient care accounted for 45% ($4,262) of the total
average costs, while inpatient care accounted for 36% ($3,375).
Outpatient pharmacy payments averaged $1,813 (19%). Cost
differences between anemic and nonanemic patients were
statistically significant (P <0.001) for all types of care except
outpatient pharmacy (P = 0.24).
Table 7 presents adjusted and unadjusted differences in
average annualized per-patient costs. Cost differences persisted
after adjusting for differences in patient gender and age, coverage
type (e.g., preferred provider organization, indemnity), predisposing condition (i.e., the 6 study conditions), and disease
severity for each predisposing condition. The average annualized
total cost per anemic patient was more than twice the average for
nonanemic patients. Both outpatient and inpatient costs were
more than twice as high for anemic patients as for nonanemic
patients.
Services that we could attribute to anemia, based on our
algorithm (i.e., claims for anemia treatment or claims containing a
www.amcp.org
diagnosis of anemia), accounted for only 5% to 11% of the cost
differential between anemic and nonanemic patients. Most of
the difference was accounted for by services without an anemia
diagnosis code or another unambiguous relationship to anemia.
Average annualized anemia-attributed payments, using our
algorithm, were $563 per patient. Outpatient care accounted
for the largest share of anemia-attributed expenditures (which
included a proportion of costs for services with either an
anemia diagnosis or anemia-specific procedure such as
erythropoietin administration), averaging $412 per patient
annually (73% of total anemia-attributed costs). The costs of
inpatient care attributable to anemia, based on our algorithm,
were slightly lower than those for professionally administered
outpatient medications, $90 and $99 per patient annually.
Anemia-attributed outpatient prescription costs (i.e., costs for
anemia-related medications) averaged $61 per patient annually.
■■ Discussion
For this study, we examined anemia prevalence, current treatment patterns, and associated costs of care in a privately insured
population in order to determine the impact of anemia on the
use of health care resources. In general, anemia occurred with
noticeable frequency even in a relatively healthy privately
insured population and resulted in higher health care utilization
and costs.
Overall, 3.5% of the study population was anemic at some
point during the study year. It is somewhat challenging to
compare this estimated anemia prevalence with estimates from
previous studies, given that no standard definition of anemia is
currently used and that reported and actual prevalence vary
widely depending on the nature of the population studied. It is
likely that the prevalence estimates of anemia in this study
underestimate the true prevalence since, in order to be considered
anemic, a patient was required to have a diagnosis of anemia
recorded on a claim during the study period or to have received
one of the specified anemia therapies.
These results underscore the fact that anemia is common
enough to merit attention, even outside the context of those
conditions with which it has historically been associated. Nearly
4% of the overall study population had anemia that was serious
enough to be recorded as a diagnosis on a medical claim, to
receive an anemia-related prescription medication, or to require
an anemia-related procedure.
In general, these results highlight the importance of understanding the demographic and clinical risk factors that increase
the likelihood that a particular individual will be anemic. From
both public health and provider perspectives, such profiles of
“at-risk” populations are critical for improving anemia screening,
detection, and treatment. This study suggests that anemia
merits particular attention in routine clinical care for women
and the elderly.
Anemic patients used significantly more health care services
www.amcp.org
TABLE 5
Average Annualized Health Care
Payments by Type of Care for Anemic
and Nonanemic Patients Based on Claims
Incurred January 1, 1998, to June 30, 2001
Type of Care
Payments ($)
Payments ($) per per Nonanemic
Anemic Patient
Patient
N = 118,332
N = 35,948
P Value
(Average
(Average
(2-Tailed
Follow-up
Follow-up
Student’s
Months = 8.9)
Months = 9.3)
t Test)
Total
14,535
9,451
<0.001
4,775
3,375
<0.001
137
$101
<0.001
Outpatient care
7,927
4,262
<0.001
Pharmacy-based outpatient
medications
1,833
1,813
0.238
Professionally administered
outpatient medications
157
0.13
<0.001
Inpatient care
Emergency room
and had higher costs ($14,535 vs. $9,451, P <0.001), even
compared with patients with the same underlying condition
who were not anemic. Since it is often assumed that anemia
may simply be a marker for the severity of a key underlying
disease (e.g., RA) and that disease severity would therefore be
the primary driver for any observed cost differences, our multivariate analysis was designed to adjust for the presence of key
conditions and the severity of those conditions as well as for
other factors that could influence costs (i.e., patient age, gender,
and coverage type). Our results indicate that costs of anemic
patients were more than twice those of nonanemic patients even
after adjusting for these other potential confounders.
The majority of the anemic patients (85%) did not receive
any of the therapies assessed. This is quite striking since these
patients were primarily identified through the presence of
explicit anemia diagnoses on the medical claims and, therefore,
those with mild anemia may be underrepresented. One possible
explanation is the use of oral iron, which was not captured in
the study database because it is an over-the-counter medication.
Nonetheless, the finding that only 15% of patients received any
apparent treatment raises serious concern that anemia is
inadequately managed. It is possible that physicians in general
do not attribute much clinical importance to anemia, especially
in patients who do not have medical conditions that may be
exacerbated by anemia, and/or for whom anemia management
is part of the standard of care. If that is the case, then efforts to
increase awareness of anemia’s risk factors and consequences
are needed, along with practical clinical treatment guidelines.
It may not be surprising that anemia in the overall health
plan population appears to be either undertreated or minimally
treated. It is surprising, however, that despite explicit guidelines
Vol. 11, No. 7
September 2005
JMCP
TABLE 6
Average Annualized Health Care Payments for Anemic and Nonanemic Patients
Anemic Patients
Nonanemic Patients
N
Average Payment
($) per Patient
118,332
14,535
35,948
9,451
5,084*
563 (11%)
7,545
41,292
5,814
12,535
28,757*
2,324 (8%)
354
37,424
232
13,579
23,845*
1,300 (5%)
3,852
17,186
3,303
7,777
9,409*
502 (5%)
2,538
19,113
2,139
7,678
11,435*
676 (6%)
14,985
29,703
11,886
12,459
17,244*
1,141 (7%)
Cancer
22,030
34,009
17,542
9,034
24,975*
1,480 (6%)
Condition
All
HIV
Average Payment
($) per Patient
N
Payment ($)
Difference
AnemiaAttributed
Payment ($)
per Patient
(% Difference)
* P < 0.001 based on 2-tailed Student’s t test.
HIV = human immunodeficiency virus.
TABLE 7
Unadjusted and Adjusted Differences* in
Average Annualized Health Care Payments
for Anemic and Nonanemic Patients
Average
Total
Payments ($)
per Patient†
Average
Inpatient
Payments ($)
per Patient
Average
Outpatient
Payments ($)
per Patient
Unadjusted costs for
anemic patients
(N = 118,332)
14,535
4,775
7,927
Unadjusted costs for
nonanemic patients
(N = 35,948)
9,451
3,375
4,262
Adjusted costs for
nonanemic patients
7,106
1,996
3,297
Unadjusted cost differences
5,084
1,400
3,665
Adjusted cost differences‡
7,429
2,779
4,630
* Costs for the nonanemic patients were standardized using the covariate levels in the
anemic population. Covariates included age, sex, coverage type (preferred provider
organization), presence and severity of 6 study conditions, Charlson Comorbidity
Index.
†Total payments include inpatient, outpatient, and outpatient pharmacy costs.
‡ P values for differences were all <0.001 based on 2-tailed Student’s t test.
emphasizing aggressive anemia management in CKD18 and cancer,19
such patients are not receiving adequate or appropriate care.19,20
These results also suggest that when anemia treatments are
employed, their usage is remarkably similar across the 6 study
conditions. The one exception is HIV, where the use of
erythropoietin is more commonly observed than in the other
conditions. This pattern suggests that the presence of underlying conditions may not play a significant role in clinical
judgments about which anemia treatment is most appropriate
JMCP
September 2005
Vol. 11, No. 7
or even about how aggressively to manage anemia.
Some of the observed treatment patterns also highlight
specific quality-of-care considerations. Despite growing
concern about the risks associated with transfusions and a wide
array of initiatives to promote blood conservation, transfusions
represented the predominant treatment among patients with
newly diagnosed anemia in the study database. There is a clear
gap between current practice relative to blood conservation and
recommendations for transfusion alternatives. For example,
current recommendations for use of blood products list iron,
folate, B12, and erythropoietin therapy as specific therapies that
should be administered instead of blood transfusions if the
patient’s condition permits time for these agents to be utilized.21
Since the current study focused only on use of injectable drugs,
further analysis is necessary to understand the full extent to
which these recommended first-line pharmaceutical therapies
are used.
Limitations
Administrative claims data are one of the richest sources of
information on health care utilization and cost and have historically
served as the foundation for many areas of health services research.
Like any data source, claims data present limitations: the most
important is that the level of detail available is limited to that
required for claims adjudication and internal and external
health plan reporting. This limitation may lead to underidentification of anemia in the study population since the anemia
diagnosis codes and anemia-related procedures and pharmaceutical therapies used for patient selection are only proxy
indicators of anemia. The lack of actual hemoglobin levels for
the patients in the study population also precludes the
assessment of anemia severity. While claims data do present a
reasonable amount of clinical information, no simple standard
methodology was available to stratify patients by severity for
www.amcp.org
the 6 study conditions. Finally, although the costs presented in
this paper may differ from those experienced by another health
plan due to differing fee schedules, the more critical finding
is the statistically higher health care costs associated with
anemia.
Although the data collected are not sufficient to explain why
health care costs are higher among anemic patients, we would
like to suggest a few possible explanations. First, despite our
best efforts to control for disease severity and comorbidity
burden in this analysis, it is possible that anemic patients are
simply sicker than the controls and therefore use more health
care services. It is also possible that the presence of anemia may
contribute to a higher rate of detection of comorbidities in these
patients or that anemic patients with comorbidities may be
more likely to have their anemia detected, resulting in an
association that is not causal.
These adjusted cost comparisons should also be considered
in light of our examination of anemia-attributed costs
(e.g., costs due to specific anemia visits, tests, and therapies).
Although these anemia-attributed costs certainly contributed to
overall health care costs in the study population, they
represented only a small percentage of the total costs per
patient. These results suggest that anemia may be responsible
for excess costs in areas that cannot be captured by our
algorithm or, alternatively, that the association is a marker for
disease severity associated with increased costs (i.e., that the
association is not causal).
■■ Conclusions
The results of this study demonstrate that the elevated clinical
burden that anemia imposes at the patient level in turn increases
the resource burden at the health plan level. As anemia gains
greater recognition as both an important clinical and public
health issue, careful consideration should be given to determining
the most cost-effective approaches to anemia screening in highrisk populations and efforts to improve anemia diagnosis and
treatment. Given the challenges inherent in isolating the true
costs of anemia, future research should examine the economic
impact of more aggressive anemia treatment to determine if
expected short-term cost increases incurred by earlier treatment
would be offset by savings from fewer admissions, shorter
lengths of stay, and less use of other expensive services in the
treatment of anemia-related outcomes. Target research should
also examine the extent to which patient care is consistent with
current guidelines.
DISCLOSURES
Funding for this research was provided by the National Anemia Action
Council, Inc. (NAAC); additional funding was also provided by the Richard
Rosenthal Dialysis Fund. Corporate sponsors of NAAC include Amgen,
Watson Pharmaceuticals, American Regent Laboratory, and Fibrogen. Author
Allen R. Nissenson discloses that he has received research support from
Amgen, OrthoBiotech, Roche, and Watson pharmaceutical companies and
American Regent Laboratory and has been a speaker for ARL. L. Tim
Goodnough is on the speaker’s bureau of Amgen, Novo Nordisk, Watson,
American Regent Laboratory, and OrthoBiotech; Sally Wade and Kevin Knight
provide consulting services to the health care industry; Robert W. Dubois is an
employee Cerner Health Insights, which provides consulting services to the
health care industry. The authors disclose no potential bias or conflict of
interest relating to this article.
Nissenson served as principal author of the study. Study concept and
design were contributed by all authors. Analysis and interpretation of data
were contributed by Wade and Knight. Drafting of the manuscript was primarily the work of Wade and Dubois, and its critical revision was the work of all
authors. Statistical expertise was contributed by Knight.
REFERENCES
1. Dallman PR, Yip R, Johnson C. Prevalence and causes of anemia in the
United States, 1976 to 1980. Am J Clin Nutr. 1984;39:437-45.
2. Goodnough LT, Dubois RW, Nissenson AR. Anemia: not just an innocent
bystander? Arch Intern Med. 2003;163:1400-04.
3. Adams PF, Hendershot GE, Marano MA. Current estimates from the
National Health Interview Survey, 1996. Vital Health Stat 10. 1999:1-203.
4. Guralnik JM, Eisenstaedt RS, Ferrucci L, Klein HG, Woodman RC.
Prevalence of anemia in persons 65 years and older in the United States:
evidence for a high rate of unexplained anemia. Blood. 2004;104:2263-68.
5. Goodnough LT, Dubois RW, Nissenson AR. Anemia: not just an innocent
bystander? [correction]. Arch Intern Med. 2003;163:1820.
6. Kausz AT, Steinberg EP, Nissenson AR, Pereira BJG. Prevalence and management of anemia among patients with chronic kidney disease in a health maintenance organization. Dis Manag Health Outcomes. 2002;10:505-13.
7. Baer AN, Dessypris EN, Krantz SB. The pathogenesis of anemia in rheumatoid arthritis: a clinical and laboratory analysis. Semin Arthritis Rheum. 1990;
19:209-23.
8. Knight K, Wade S, Balducci L. Prevalence and outcomes of anemia in cancer: a systematic review of the literature. Am J Med. 2004;116(suppl 7A):11S26S.
9. Belperio PS, Rhew DC. Prevalence and outcomes of anemia in individuals
with human immunodeficiency virus: a systematic review of the literature.
Am J Med. 2004;116(suppl 7A):27S-43S.
10. Wilson A, Reyes E, Ofman J. Prevalence and outcomes of anemia in
inflammatory bowel disease: a systematic review of the literature. Am J Med.
2004;116(suppl 7A):44S-49S.
11. Wilson A, Yu HT, Goodnough LT, Nissenson AR. Prevalence and outcomes
of anemia in rheumatoid arthritis: a systematic review of the literature. Am J
Med. 2004;116(suppl 7A):50S-57S.
12. Weiss G, Goodnough LT. Anemia of chronic disease. N Engl J Med. 2005;
352:1011-23.
13. Collins AJ, Li S, Ebben J, Ma JZ, Manning W. Hematocrit levels and associated Medicare expenditures. Am J Kidney Dis. 2000;36:282-93.
ACKNOWLEDGMENTS
The authors wish to thank Lisa Kaspin, PhD, Cerner Health Insights, Beverly
Hills, CA, for her assistance in preparing this article and Onur Baser, PhD, and
Xue Song, PhD, of Medstat, Ann Arbor, MI, for providing data and technical
programming for this study.
www.amcp.org
14. Berndt E, Crown W, Kallich J, et al. The impact of anaemia and its treatment on employee disability and medical costs. Pharmacoeconomics. 2005;
23:183-92.
15. Lyman GH, Berndt ER, Kallich JD, et al. The economic burden of anemia
in cancer patients receiving chemotherapy. Value Health. 2005;8:149-56.
Vol. 11, No. 7
September 2005
JMCP
16. Nordyke RJ, Kim JJ, Goldberg GA, et al. Impact of anemia on hospitalization time, charges, and mortality in patients with heart failure. Value Health.
2004;7:464-71.
20. Nissenson AR, Collins AJ, Hurley J, et al. Opportunities for improving the
care of patients with chronic renal insufficiency: current practice patterns.
J Am Soc Nephrol. 2001;12:1713-20.
17. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for
use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613-19.
21. American Association of Blood Banks, America’s Blood Centers, and the
American Red Cross. Circular of information for the use of human blood and
blood components. Available at: http://www.aabb.org/All_About_Blood/
COI/coi0702.pdf. Accessed June 15, 2005.
18. NKF-DOQI clinical practice guidelines for the treatment of anemia of
chronic renal failure. National Kidney Foundation-Dialysis Outcomes Quality
Initiative. Am J Kidney Dis. 1997;30:S192-S240.
19. Groopman JE, Itri LM. Chemotherapy-induced anemia in adults: incidence and treatment. J Natl Cancer Inst. 1999;91:1616-34.
JMCP
September 2005
Vol. 11, No. 7
www.amcp.org
C O N T E M P O R A RY S U B J E C T
Pharmacist Response to Alerts Generated From
Medicaid Pharmacy Claims in a Long-term Care Setting:
TROY K. TRYGSTAD, PharmD, MBA; DALE CHRISTENSEN, RPh, PhD;
JENNIFER GARMISE, PharmD; ROBERT SULLIVAN, MD, MPH; and STEVEN E. WEGNER, MD, JD
ABSTRACT
OBJECTIVE: In response to burgeoning drug costs, North Carolina (NC) Medicaid
encouraged pharmacists and prescribers to develop collaborative programs to reduce
drug expenditures. One of these programs, the North Carolina Polypharmacy Initiative,
was a focused drug therapy management intervention aimed at reducing polypharmacy
in nursing homes. This intervention targeted patients with more than 18 prescription
fills in 90 days, beginning in November 2002. These patients were believed to have a
high likelihood of experiencing potential drug therapy problems (PDTPs). Consultant
pharmacists were asked to utilize profiles displaying alerts generated from pharmacy
claims to guide interventions in addition to usual-care drug regimen reviews. The
pharmacists documented their reviews, recommendations, and resulting changes in
drug therapy. Our objectives were to determine (1) the persistence of PDTP alerts
following interventions by consultant pharmacists and (2) the impact of these interventions on patient drug costs from a payer perspective.
METHODS: A before-after study with comparison group design was used. Medicaid
prescription claims data were compared for the 90-day periods prior to the intervention (June-August 2002) and following the intervention (March-June 2003). The
90-day postintervention period allowed for 2 to 3 follow-up prescriptions and reduced
the drop-out rate. The 5 categories of potential problem alerts included potentially
inappropriate medications (Beers criteria), substitution opportunity for a lower-cost
drug, 16 drugs or drug classes with specific quality improvement opportunities
(Clinical Initiatives list), therapeutic duplication, and length of drug therapy evaluation.
RESULTS: A total of 253 nursing homes, involving 110 consultant pharmacists and
6,344 patients, were in the intervention arm, with 5,160 patients (81.3%) remaining
at the end of the follow-up period. At baseline, study-group patients used an average
of 9.7 prescriptions per month, costing the NC Medicaid program $517 per patient
per month (PPPM). There were 6,360 recommendations offered for 3,400 patients, or
an average of 1.87 recommendations per patient. Physicians concurred with 59.8%
(3,801 of 6,360) of all recommendations to change drug therapy, about half involving
a switch to a lower-cost drug. Two of 5 alert categories had significant (P <0.01)
reductions in alert persistence: -10.8% for the study group versus -0.7% for the
comparison group for the Clinical Initiatives list and -29.7% for the study group
versus -14.1% in the comparison group for the drug substitution opportunity. Median
drug costs per patient in the study group decreased by $12.14 (-0.92%), from
$1,329.46 to $1,317.32, and increased in the comparison group by $44.98 (3.35%),
from $1,341.25 to $1,386.23, creating a relative cost reduction of $57.12 per patient
in the 3-month follow-up period, or $19.04 PPPM.
CONCLUSION: A supplemental program of medication reviews for nursing home
patients targeted by high drug utilization resulted in a reduction in the persistence of PDTP alerts and was cost beneficial based solely on drug cost savings.
This intervention may be a model for future medication therapy management
services provided by prescription drug plans under Medicare Part D for patients
in long-term-care settings and possibly ambulatory patients.
KEYWORDS: Nursing homes, Pharmaceutical care, Medication therapy management, Drug use review, Polypharmacy, Drug regimen review
www.amcp.org
T
he passage of the Medicare Prescription Drug,
Improvement, and Modernization Act of 2003 (MMA)
and the ensuing rollout of the outpatient drug benefit in
January 2006 have focused attention on ensuring elderly
patient access and cost-effective prescribing and use of drugs.
Those responsible for Part D program administration within the
Centers for Medicare and Medicaid Services (CMS) and prescription drug program sponsors share the formidable task of
managing both the cost and quality of drug regimens for more
than 40 million Medicare beneficiaries. Medicare will become
the largest single payer of drug benefits in the United States,
with a projected $70 billion in expenditures in 2006.1
The elderly have more chronic illnesses and use more
prescription drugs than any other age segment, increasing the likelihood of adverse drug events, many of which are avoidable.2-4
In an attempt to ameliorate the cost burden and ensure rationale
and optimal drug use, Congress took the novel approach of
requiring prescription drug plans (PDPs) and Medicare Advantage
PDPs to offer a Medication Therapy Management Program
(MTMP) as part of their drug benefit. Despite considerable variations in strategy and implementation, prior MTMP-like programs
have demonstrated significant cost savings and reductions in drug
therapy problems for other targeted patient populations.5-7
Authors
TROY K. TRYGSTAD, PharmD, MBA, is a contract data analyst, AccessCare, Inc.,
Morrisville, North Carolina, and a PhD candidate (American Foundation for
Pharmaceutical Education fellow), Division of Pharmaceutical Policy and
Evaluative Sciences, School of Pharmacy, University of North Carolina, Chapel Hill;
DALE CHRISTENSEN, RPh, PhD, is a professor, Division of Pharmaceutical Policy
and Evaluative Sciences, School of Pharmacy, University of North Carolina, Chapel
Hill; JENNIFER GARMISE, PharmD, is a pharmacy projects manager,
AccessCare, Inc., Morrisville, North Carolina; ROBERT SULLIVAN, MD, MPH, is
medical director, Community Care of North Carolina, and associate professor,
Department of Community and Family Medicine, Department of Medicine, Center
for the Study of Aging and Human Development, Duke University Medical Center,
Durham, North Carolina; STEVEN E. WEGNER, MD, JD, is president and medical director, AccessCare, Inc., Morrisville, North Carolina, and an adjunct assistant
professor, Department of Pediatrics, University of North Carolina, Chapel Hill.
AUTHOR CORRESPONDENCE: Troy K. Trygstad, PharmD, MBA, Division of
Pharmaceutical Policy and Evaluative Sciences, School of Pharmacy, C.B. 7360,
University of North Carolina, Chapel Hill, NC 27599-7360. Tel: (919) 843-3775;
Fax: (919) 966-8466; E-mail: [email protected]
Vol. 11, No. 7
September 2005
JMCP
Pharmacist Response to Alerts Generated From Medicaid Pharmacy Claims in a Long-term Care Setting:
FIGURE 1
Inclusion Criteria and Cohort Development
131
Nonparticipating Nursing Homes
253
Participating Nursing Homes
>12,000
North Carolina Nursing Home Residents
With at Least 18 Prescription Fills in 90 Days
BeforePeriod
Screen
1,204
Patients Discharged
or Deceased
9,208
Patient Drug Profiles Sent
Intervention
Period
1,660
Profiles Not
Returned
6,344
Profile Reviews
Completed
AfterPeriod
Screen
2,202
Patients Remain in
Comparison Group
5,160
Patients Remain Following
Intervention Period
3,400
Patients With
Recommendations
2,305
Patients With Accepted
Recommendations
Defining the nature and scope of MTMP services within
Medicare Part D continues to be a dynamic and ongoing
endeavor. A consortium of pharmacy trade and professional
associations published a working definition in July 2004.8 This
definition was expanded by the American Pharmacists
Association and the National Association of Chain Drug Stores
in April 2005.9 However, CMS’s final rules pertaining to MTMP
services remain broadly defined, leaving the operational details
to PDP sponsors.10
MMA was not the first federal legislation to require pharmacist
involvement in the drug-use process. Beginning in the 1970s,
federal regulations imposed a requirement that monthly drug
regimen reviews (DRRs) be conducted in long-term-care facilities
by consultant pharmacists.11 Subsequent Omnibus Reconciliation
Act legislation (OBRA ’87) required that this review be
accomplished in collaboration with the attending physician.
These regulations contained explicit requirements for reviewing
therapy for targeted drugs and drug classes deemed to be over-
JMCP
September 2005
Vol. 11, No. 7
used in long-term-care settings. While such reviews have resulted
in improved care since first mandated,12 there is room for
improvement, and a more holistic approach based upon the
optimization of both the type and use of all drugs taken by
Medicare Part D recipients seems prudent.13
Medicaid recipients are also subject to drug reviews through
OBRA ’90 regulations that require ongoing statewide drug
utilization review (DUR) activities. These programs typically
focus on drug use by ambulatory Medicaid recipients. The
legislation compelled states to establish committees and
systems to review patterns of drug use believed to be problematic
but did not go as far as MMA to allow for explicit compensation of
pharmacists as providers of care.
MMA legislation effectively shifts the burden of drug costs
incurred by elderly Medicaid recipients from the state-federal
program to the federal government. Prior to the passage of
MMA, states were burdened with Medicaid drug expenditures
that were ballooning at unsustainable rates despite the federal
sharing of Medicaid costs. North Carolina (NC) Medicaid spent
more than $1.2 billion on drugs in 2003, with the elderly
accounting for 11% of recipients but 32% of all prescription
drug costs.14,15 In response to these trends, NC Medicaid introduced a program that combined the state-level, top-down
administration characteristic of DUR activities with patient-level,
pharmacist-driven activities typical of DRRs. This program was
titled the North Carolina Polypharmacy (NCPP) Initiative.
Following a successful pilot study, the NCPP Initiative was
launched in 253 nursing homes in North Carolina with emphasis
on elderly Medicaid recipients. In addition to mandated DRRs,
the initiative provided a targeted drug therapy management
consultation provided by a pharmacist with the treating physician.
In these targeted drug therapy management consultations,
pharmacists were to (1) review a drug profile generated from
Medicaid pharmacy claims with potential drug therapy problem
(PDTP) alerts and medical records of Medicaid patients in nursing
homes, (2) determine if a drug therapy problem existed,
(3) recommend a change if needed, and (4) perform a follow-up
to determine if the change was implemented.
The NCPP Initiative was organized as a collaborative activity
that incorporated a physician primary care practice network
(AccessCare of North Carolina), a pharmacy consultant coalition,
and a network of nursing home medical directors. The nature
of the NCPP Initiative and its organization was described in an
earlier paper that reported the type and frequency of pharmacist
interventions and estimated the cost impact of drug therapy
changes by type of PDTP.16 Intervention documents submitted
by pharmacists were used as a single data source. For the 6,344
patients with reviews, pharmacists responded to approximately
20,000 drugs with alerts by making 6,520 recommendations,
resulting in changes in drug therapy 58% of the time.16 These
changes were projected to save NC Medicaid $30.33 per patient
per month (PPPM).16
www.amcp.org
FIGURE 2
Patient Profile Generated From Medicaid Pharmacy Claims—
With Potential Drug Therapy Problem Alerts
In the present article, we reconcile the projected drug cost
impact of pharmacist intervention activities with actual Medicaid
claims data spanning a 6-month period. We describe the nature of
PDTP alerts, drugs involved, recommendations, and actions taken
after physician consultation. We also assess changes in drug
therapy from a qualitative and economic perspective using a
before-after study design with a comparison group.
Our working hypothesis was that a systematic program of
pharmacist-directed DUR that supplements requisite OBRA ’87
DRRs in nursing homes would produce drug therapy changes
that maintain or improve the quality of care while decreasing
drug costs. The specific objectives of the current study were to
determine (1) the persistence of PDTP alerts following interventions by consultant pharmacists and (2) the impact of these
interventions on patient drug costs from a payer perspective.
This study received approval from the Institutional Review
Board at the University of North Carolina at Chapel Hill.
www.amcp.org
■■ Methods
Setting and Participants
Phase 1 of the NCPP Initiative was conducted by 110 pharmacists
in 253 nursing homes, representing approximately 70% of all
nursing homes in North Carolina (Figure 1). Participation in the
intervention was solicited through the North Carolina Long
Term Care Pharmacy Alliance, a representative group of
pharmacists serving nursing homes throughout the state.
Exempted were 13 homes that contracted with a single
pharmacy provider and were involved in a separate, ongoing
intervention project. All Medicaid residents of the participating
facilities who had 18 or more prescription fills in the 90-day
period prior to the start of the study were eligible for an
on-site profile review by a consultant pharmacist. This time
horizon was chosen to capture, on average, 3 monthly supplies
of medications while limiting the dropout rate as much as
possible.
Vol. 11, No. 7
September 2005
JMCP
Pharmacist Responsibilities
Participating pharmacists were introduced to the project, toolkit,
and documentation form during two 1-hour group meetings
and one 1-hour conference call. Other professional interactions
took place throughout the course of the project, including
informational meetings with geriatric associations, nursing
home medical directors, and network physicians, as well as the
use of telephone follow-ups. The toolkit contained instructions
for documenting interventions and explained the screening
criteria used to select (flag) drugs for attention.
Each consultant pharmacist was provided with drug profiles
computer-generated from Medicaid pharmacy claims that
displayed flags for patients and suggestions for modifications of
drugs and classes of drugs. Pharmacists were asked to record
both the result of the review (i.e., the intervention) and the
result of the consultation with the prescribing physician
(i.e., the outcome) on a documentation form (Figure 2).
Recording the result of the intervention required awaiting the
prescriber’s response to the recommendation. Pharmacists were
required to conduct these assessments during their regularly
scheduled visits to each home. Consultant pharmacists
employed their usual methods of communicating with physicians
(fax, phone, or written notation in the medical record) to make
recommendations and to learn the outcome of the change recommendation. We categorized the drug therapy flags as (1) unnecessary drug therapy, (2) more cost-effective drug available,
(3) wrong dose/delivery, (4) potential for adverse drug reaction,
(5) needs additional therapy, and (6) other problem. We coded
intervention results as (1) dose/delivery changed, (2) drug
added, (3) drug changed (from one to another), (4) drug discontinued, (5) no change, and (6) other intervention.
If an intervention resulted in a drug therapy change of any
type, the new drug, dose, and quantity were noted. Drug, dose,
and quantity were also reported for each new drug added for
previously untreated indications. Pharmacists were compensated
$12.50 for each comprehensive profile review for which results
were clearly documented on the forms provided (i.e., the
patient profile). This compensation amount was based on our
estimate of the additional time required for these focused
reviews above and beyond normal review activities and a
customary rate of pay of $50 per hour. Pharmacists were
compensated regardless of problem determination and/or the
offering of a recommendation.
Drug Profiles and PDTP Alert Criteria
Patient drug profiles were generated from Medicaid claims data
and contained, for each listed drug, a space for all alert
categories, marked with the appropriate flag/alert if a PDTP was
determined by matching claims data with drug lists generated
from alert categories. The profiles were a compilation of all
drugs for which a claim was paid in the 90 days prior to generation, regardless of the presence of an alert. The first alert criterion
JMCP
September 2005
Vol. 11, No. 7
was receipt of a drug widely considered to be inappropriate for
use in the elderly (Beers drug list).17 In order to engender
participation and maximize the quality of the PDTP alerts,
program administrators also elicited input from local physicians
and consultant pharmacists. Thus, the second criterion was
receipt of a drug on the Community Care of North Carolina
Prescription Advantage List (PAL), which encourages substitution
of less expensive drugs within a therapeutic class. This voluntary
preferred drug list was conceived and is maintained by a
committee of practicing physicians in North Carolina specifically
for NC Medicaid. There are 3 categories of PAL drug alerts.
PAL-3 drugs are considered to incur “significant cost” to the
Medicaid program (e.g., Nexium, Prilosec, Zestril, Prinivil, as of
November 2002), while PAL-2 drugs offered “no clear
cost advantage” (e.g., Prevacid, Aciphex, Accupril, Monopril,
Lotensin, Altace, as of November 2002), and PAL-1 drugs offer
“significant cost savings” to the Medicaid program
(e.g., Protonix, lisinopril, enalapril, captopril, as of November
2002). The third criterion was the appearance of a drug on a
“Clinical Initiatives” list. The Clinical Initiatives list was
developed by consultant pharmacists participating in the NCPP
Initiative and included 16 drugs and/or drug classes (e.g., COX-2
inhibitors, statin drugs, sleep aids, low-sedating antihistamines)
that had the potential for quality improvement and cost savings.
Program administrators offered 2 additional alerts: therapeutic
duplication and a “consider length of therapy” alert that was
derived from classes of drugs considered appropriate only for
short-term use (e.g., antibiotics, injectable enoxaparin).
Research Design
We first evaluated pharmacist action and reporting by reconciling
the response to alerts with downstream prescribing activity
using the Medicaid dispensed prescription claims database.
Using a before-after, study-comparison-group design, we compared
prescription use during the 3 months before intervention (JuneAugust 2002) with a period of equal length at the end of Phase 1
(March-June 2003). Second, we assessed whether or not PDTP
alerts were reduced during the follow-up period compared with
usual-care controls (nonrandomized comparison group). Third,
we describe the economic consequences of pharmacist activities
in terms of changes in drug cost using pharmacy paid claims
data.
Study-group patients were Medicaid recipients residing in
participating nursing homes who received a completed profile
review by a consultant pharmacist. The comparison group
consisted of patients in nursing homes not responding to the
invitation for inclusion in Phase 1 of the intervention. Inclusion
of patients in comparison-group homes was determined
by criteria identical to study-group patients (i.e., more than 18 prescription fills in 90 days, Figure 1). Several of the nursing homes
in the comparison group became participants in later phases of
the project, but only after the 6-month study window in this
www.amcp.org
TABLE 1
Baseline Characteristics of Continuously Enrolled Patients by Treatment Group
Characteristic
Study Group
(n = 5,160)
Study Group
(With Recommendation*)
(n = 3,400)
Study Group†
(With Acceptance)
(n = 2,305)
Comparison Group
(n = 2,202)
Sex, no. (%)
Male
Female
1,289 (24.98)
3,871 (75.02)
820 (24.12)
2,580 (75.88)
533 (23.99)
1,752 (76.01)
484 (21.98)
1,718 (78.02)
Race, no. (%)
White
Other
3,533‡ (68.47)
1,627 (31.53)
2,325‡ (68.38)
1,075 (31.62)
1,588‡ (68.89)
717 (31.11)
1,667 (75.70)
535 (24.30)
77.57 ± 12.72
(80.0)
77.63 ± 12.42
(80.0)
77.67 ± 12.44
(80.0)
78.65 ± 12.46
(81.0)
29.04 ± 9.92
(27.0)
29.86 ± 10.27
(28.0)
30.19 ± 10.53
(28.0)
28.18 ± 10.74
(26.0)
$1,549.89 ± 1,652.49
($1,329.46)
$586.91 ± 919.17
($1,392.14)
$1,610.02‡ ± 926.77
($1,427.13)
$1,543.67 ± 921.98
($1,341.25)
Age, years, mean ± SD
(median)
No. of prescription fills, 3 month period, mean ± SD
(median)
Amount of paid claim ($), 3-month period, mean ± SD
(median)
Note: Difference of proportions tests were used to determine differences in sex and race. T-testing was used to determine differences in age, number of prescription fills,
and amount of paid claims.
* Study group (with recommendation) = those patients having a recommendation resulting from pharmacist consultation, regardless of outcome.
† Study group (with acceptance) = those patients having a recommendation and a change in therapy as a result of a recommendation provided by a pharmacist.
‡ Denotes significantly different from comparison group at P <0.01.
analysis. For study-group patients, we linked prescription drug
use elicited through claims data to pharmacist-reported interventions (or lack thereof) on patient profiles. We examined
2 study subgroups: (1) patients whose drug use received pharmacist
recommendations and (2) patients for whom recommendations
were accepted.
Studies in the long-term-care arena are often burdened by a
high attrition rate. Using a combination of claims data and pharmacist report, we estimated an annual nursing home resident
attrition rate of 36% due to death or discharge in North
Carolina. Since we were not able to verify dropout from
prescription claims with certainty, only residents having claims
in the last 35 days of the 90-day follow-up period were included
in both the study and comparison groups.
Statistical testing was performed using SAS statistical software,
version 8.2 (1999-2001, SAS Institute Inc., Cary, NC). We used
nonparametric statistical testing to account for possible skewness in the data.
■■ Results
Prescription profiles were generated and sent to consultant
pharmacists for 9,208 patients. Pharmacists returned 7,548
(82%) of all profiles sent to them (Figure 1). After excluding
1,204 patients (13%) who were discharged or deceased, 6,344
patients were subjected to profile reviews. This number diminished
to 5,160 patients who remained in the Medicaid population
throughout the follow-up period, constituting an 18% dropout
rate over 6 months due to death or discharge. This is consistent
with historical dropout rates for Medicaid recipients. Remaining
patients had received an average of 9.7 prescription fills (median
www.amcp.org
9) per month during the 3-month period prior to profile generation.
Exclusive of manufacturer rebates, the average PPPM drug cost
to NC Medicaid was $517, with a median of $443.
The comparison group consisted of 2,202 patients selected
in the same manner as study-group patients (having 18 or more
prescription fills in a 90-day period). We compared study and
comparison groups based on age, gender, race, baseline
prescription use, and dropout rates (Table 1). The groups
differed with respect to race, with a lower proportion of whites
in study nursing homes versus comparison-group homes (69%
vs. 76%, respectively, P <0.01). At baseline, drug usage and
costs were similar for study and comparison-group patients
with one exception: the study subgroup with changes resulting
from recommendations had higher baseline prescription costs.
Dropout rates from the original cohorts were also similar across
the groups (at 18% to 19%).
Among study group patients, the most common PDTP alert
was for a drug with a potential therapeutic duplication with an
average of 5.11 alerts (Table 2). Therapeutic duplication alerts
were common because a single potential duplication triggered at
least 2 alerts. Clinical Initiative alerts averaged 2.77 alerts per
patient. This was followed by PAL-2 or PAL-3 drugs (1.58 per
patient) and Beers list drugs (0.78 per patient). A total of 6,360
interventions were offered for 3,400 patients in the study group,
an average of 1.87 per patient with intervention. Based on
pharmacist reporting, physicians concurred with 59.8% (3,801
of 6,360) of all interventions to change drug therapy (Table 3).
Pharmacist suggestion for a more cost-effective drug was the
most popular recommendation (3,327) with the greatest
frequency of success (2,088, 62.8%). A recommendation for a
Vol. 11, No. 7
September 2005
JMCP
TABLE 2
Comparison of Potential Drug Problem
Alert Rates Before and After a Single
Retrospective Intervention
No. of Alerts Per Patient
Before
After
(3 months) (3 months)
Alert Type
Difference (%)
Beers List§
Study
0.78
0.70
-0.08 (-10.8)
Study (w/recommendation*)
0.82
0.72
-0.10 (-12.2)
Study (w/acceptance†)
0.83
0.71
-0.12 (-14.5)
Comparison
0.83
0.74
-0.09 (-10.8)
PAL List (2 or 3)||
Study
1.58
1.11
-0.47‡ (-29.7)
1.76
1.16
-0.60‡ (-34.1)
1.82
1.10
-0.72‡ (-39.6)
Comparison
1.63
1.40
-0.23 (-14.1)
Clinical Initiatives List¶
Study
2.77
2.47
-0.30‡ (-10.8)
3.00
2.67
-0.33‡ (-11.0)
3.09
2.68
-0.41‡ (-13.3)
Comparison
2.73
2.71
-0.02 (-0.7)
Consider Duration Flag#
Study
0.16
0.15
-0.01 (-6.3)
0.15
0.15
0.00 (0.0)
0.14
0.14
0.00 (0.0)
Comparison
0.18
0.15
-0.03 (16.7)
Therapeutic Duplication**
Study
5.11
4.63
-0.48 (-9.4)
5.15
4.78
-0.37 (-7.2)
5.22
4.75
-0.47 (-9.0)
Comparison
5.00
4.56
-0.44 (-8.8)
Note: The Wilcoxon 2-sample test was used to assess differences in alert rates between
the comparison group and study.
Sample sizes:
Study group: n = 5,160
Study group with recommendations: n = 3,400
Study group with accepted recommendations: n = 2,305
Comparison group: n = 2,202
* Study group (with recommendation) = those patients having a recommendation
resulting from pharmacist consultation, regardless of outcome.
† Study group (with acceptance) = those patients having a recommendation and a change
in therapy as a result of a recommendation provided by a pharmacist.
‡ Denotes significantly different from comparison group at P <0.01.
§ The Beers List is a list of drugs generally considered to be inappropriate in the elderly.17
|| PAL = Prescription Advantage List, a categorization of drug alerts proposed by practicing
physicians in North Carolina. PAL 3 drugs are considered to “incur significant cost.”
PAL 2 drugs are considered to offer “no clear cost advantage.” PAL 1 drugs are
considered to offer “significant cost savings.” The rates of PAL 2 and 3 drug alerts are
shown in this table.
¶ Clinical Initiatives List refers to potential drug therapy problem alerts proposed by
consultant pharmacists in North Carolina.
# Consider Duration alerts were derived from classes of drugs considered appropriate
only for short-term use.
** Therapeutic Duplication alerts were generated based upon duplications within hierarchical drug class codings.
TABLE 3
Frequency of Recommendation by Type
With Resultant Success* (n = 3,400)
Recommendation Type
Frequency
Success, No. (%)
Wrong dose or strength
545
444 (81.5)
More cost-effective drug available
3,327
2,088 (62.8)
Drug has potential for ADRs
632
328 (51.9)
Needs additional therapy
167
69 (41.3)
Other (not specified)
432
146 (33.8)
Total
6,360
3,801 (59.8)
* Recommendations were considered successful when a change in therapy occurred
subsequent to a recommendation by the clinical pharmacist.
ADRs = adverse drug reactions.
JMCP
September 2005
Vol. 11, No. 7
different dose garnered the highest rate of success (444 of 545,
81.5%). For Clinical Initiatives and PAL-2 or PAL-3 flags,
pharmacists made interventions for change in 46.2% (2,271 of
4,916) of patients; physicians endorsed 60.2% (1,939 of 3,222)
of the recommendations. Beers drugs and “consider length” (of
drug therapy) categories garnered considerably fewer recommendations.
We next examined persistence of computer-generated alerts
in the drugs received by patients before and after intervention.
Our working hypothesis was that, if the intervention program
was successful, there should be a decrease in the number of
PDTP alerts on subsequent patient drug profiles using the same
computer-screening process employed in the before-intervention
period. We found statistically significant declines in the number
of alerts per patient for both PAL and Clinical Initiatives flags
(P < 0.01) for all study groups (-29.7% and -10.8%, respectively)
compared with the comparison group (-14.1% and -0.7%,
respectively) using the Wilcoxon 2-sample test (Table 2).
As expected, even greater declines in alert rates were observed
in the study subgroup that received intervention (-34.1% and
-11.0%) and in the subgroup that had drug therapy changes as
a result of dispensing pharmacist recommendations (-39.6%
and -13.3%). When compared with baseline drug use, all flag
categories in all study groups had statistically significant reductions (P <0.01; Wilcoxon signed rank test), with the exception
of the “consider length” (of drug therapy) flag.
Finally, we examined before-after changes in the amount
paid for prescriptions (Table 4). Median drug costs per patient
in the intervention group decreased by $12.14 (-0.92%) from
$1,329.46 to $1,317.32 and increased in the comparison group
by $44.98 (3.35%) from $1,341.25 to $1,386.23, creating a
relative cost reduction of $57.12 per patient in the 3-month
follow-up period, or $19.04 PPPM. Even larger reductions in
drug costs were observed in the study subgroups with (1) documented profile reviews and with recommendations for change,
where a median decline of $25.83 per patient was observed and
(2) in the subgroup for which drug therapy changes occurred as
a result of the recommendations, where a decline of $61.68 per
patient was observed.
■■ Discussion
The results indicate that the addition of PDTP alerts to usualcare DRRs was associated with more changes in drug therapy
and a reduction in computer-generated drug therapy alerts
during the follow-up period. Among drug problem alert
categories, we found statistically significant differences between
the study group and the comparison group in alert persistence
for Clinical Initiatives and PAL drugs. These 2 categories were
constructed by physician and pharmacist leaders, suggesting
that practitioner involvement with a centralized DUR process
aids in program response. Beers list and therapeutic duplication
alerts decreased in all study groups and in the comparison
www.amcp.org
group, but persistence was not statistically different between
study and comparison groups. This finding is consistent with
the role of DRRs as outlined in OBRA ’87.11 These types of drugs
and drug problems are explicitly mentioned as part of the
guidelines for conducting customary mandated DRRs.
Residents in comparison homes were not subject to drug
profile reviews with PDTP alerts generated from pharmacy
claims as part of the NCPP Initiative. However, residents in both
study and comparison homes were subject to requirements
based on OBRA ’87 and screening guidelines for the overuse of
particular prescription drugs. This may explain the reduction in
both groups. A JMCP article published in April 2005 demonstrated significant reductions in the use of Beers list drugs
associated with an intervention involving letters to prescribers,
pharmacist phone consultations, and written literature disseminated in a predominantly ambulatory population of Medicare +
Choice (now Medicare Advantage) recipients.18 It would seem
prudent, given previous success, to attempt to replicate the
NCCP Initiative in an ambulatory Medicare setting. Notably,
few recommendations were made pursuant to the “consider
length” (of therapy) flag category in all study and comparison
groups. This length of therapy category generated only 205
alerts in total and did not contain drugs such as benzodiazepines
or psychotropic medications customarily scrutinized for length of
therapy during regularly scheduled DRRs.
This analysis of prescription claims data supports previous
findings with regard to drug cost savings resulting from the
NCPP Initiative as well as its pilot project. The NCPP Pilot
Project was found to have generated an approximate 4% reduction
in drug costs.19 A previously published article utilizing primary
data from pharmacist reports found that the NCPP Initiative
produced savings of $30.33 PPPM savings in the month immediately following the intervention.16 The resulting cost minimization ratio was determined to be 12:1.16
In the present study, we utilized Medicaid claims data to
reconcile documented pharmacist interventions and to determine
the downstream effects of those interventions. We also added a
comparison group to further strengthen its internal validity.
Using the results from Medicaid claims data in conjunction with
comparison group findings, we observed a savings of $19.04
(P=0.06) PPPM for all patients receiving profile reviews, $23.60
for patients receiving interventions (P <0.01), and $35.55
(P <0.01) for patients having at least 1 accepted intervention.
The 3-month PPPM difference between the study group and
comparison group of $57.12 remains substantial and justifies
the implementation of the Polypharmacy Initiative on the basis
of drug cost savings alone.
Previous projections based upon the first month immediately
following the interventions did not allow us to consider the
persistence of the intervention effect. An intervention may not
have been carried out for reasons unknown to the consultant
pharmacist. The intervention decision may have been reversed
www.amcp.org
TABLE 4
Total Amount Paid for Prescriptions
in the Before and After Periods
Before
After
Period
Period
(3 Months) (3 Months)
(Median) (Median) Difference (%) P Value
Study group (n = 5,160)
$1,329.46
$1,317.32 -$12.14 (-0.92)
0.06
(w/recommendation*)
$1,392.14
$1,366.31 -$25.83 (-1.86)
<0.01
(w/acceptance†)
$1,427.13
$1,365.45 -$61.68 (-4.32)
<0.01
Comparison group (n = 2,202) $1,341.25
$1,386.23
$44.98 (3.35)
n/a
Note: the Wilcoxon 2-sample test was used to assess differences in total amount of paid
claims between the comparison and study groups.
* Study group (with recommendation) = those patients having a recommendation
resulting from pharmacist consultation, regardless of outcome.
† Study group (with acceptance) = those patients having a recommendation and
a change in therapy as a result of a recommendation provided by a pharmacist.
n/a = not applicable.
by the physician after the pharmacist documented acceptance
in the report. Pharmacists may also have underreported new
drugs found on the nursing home medical record but not
appearing on the drug profile generated from Medicaid
pharmacy claims due to lag time from profile receipt to regularly
scheduled DRR activities. We noted an average difference of
$15 per month between claims analysis and pharmacist-reported
drug cost data ($516.63 in claims analysis versus $502.96 in
pharmacist-reported data) in baseline costs between these studies.
This difference illustrates the importance of reconciling
pharmacist intervention reporting with administrative claims.
Using both data sources, as we did in the present study, is
advantageous since we can tie observed medication-level interventions derived from pharmacist reporting with actual costs
incurred from claims data to validate pharmacist action.
The NCPP Initiative combines population-level, drug-specific
surveillance of DUR programs with patient-level, comprehensive
reviews characteristic of DRR activities. Alerts were generated
by the payer, in this case NC Medicaid, and were provided to
prescribing physicians to encourage change in targeted drugs
and drug classes. In line with usual care in long-term-care
settings, pharmacists were free to review and recommend
therapy changes for any drug in a patient’s profile for any
problem they discovered. Beginning in 2006, Medicare PDP
sponsors will take on a DUR role with differing approaches to
MTMP under the MMA. Standard DUR approaches have offered
little evidence, to date, of effectively improving patient outcomes for state Medicaid recipients despite the large budget
outlays to these programs.20-23 However, targeted, populationspecific interventions such as the NCPP Initiative have shown
some success.24-26 Focused reviews based upon patient-specific
Vol. 11, No. 7
September 2005
JMCP
profiles generated from administrative pharmacy claims, in
combination with collaborative activities that individualize
care,27 such as DRRs, may be a better strategy for PDPs to adopt
through the MTMP service requirement.
This strategy is not limited to the long-term-care setting and
may in fact be more effective in an ambulatory setting where
less frequent review of drug use profiles takes place. The strategy
is generally applicable to any group of beneficiaries that use
online adjudication for processing pharmacy claims.
Limitations
It was not possible to draw a true random sample of patients,
nursing homes, or pharmacist consultants due to the intermingling
of providers. Our comparison group was not, by design, a
randomized sample of patients. Due to clustering effects, it is
difficult to construct a truly randomized patient-level sample
within a nursing home because physicians often provide care to
patients in more than one nursing home. Additionally, groups of
pharmacists are often clustered through consulting organizations
serving multiple nursing homes, and multiple nursing homes
often operate under a common ownership structure.
Fortunately, baseline demographic characteristics and prescription
drug costs between the study group and the comparison group
were remarkably similar ($516.63 in the study group versus
$514.56 in the comparison group).
The study group, its subgroups, and the comparison group
did not differ statistically with respect to gender, age, or number
of prescriptions filled at baseline. There was a statistically
significant difference with respect to race, with the study group
and its subgroups having a greater proportion of nonwhite
participants. We do not suspect that this difference confounded
the results. Whatever unmeasured population differences existed,
we believe our use of before-after comparisons within groups and
the relatively large sample sizes enhance the validity of study
results. We assume that contamination effects arising from
sharing of pharmacist consultant firms between study and
comparison facilities was trivial. While some pharmacist
consulting firms served several different nursing homes, no
individual pharmacist provided consulting services to both
study and comparison group homes. Pharmacist turnover was not
a problem since the time period was relatively short. To the extent
that contamination effects were present, they would serve to
diminish observed between-group differences. We do not know
the effect of repeated interventions, the effects of continually
evolving PDTP alerts criteria,28 or intervention persistence
beyond 6 months.
We cannot confidently project the long-term impact of these
interventions. Our 3-month follow-up period reflected a
balance in our approach. On the one hand, we wanted at least
two 1-month follow-up periods to ensure that drug therapy
changes were reflected in claims data and persisted. On the
other hand, a longer follow-up period of 6 to 12 months would
JMCP
September 2005
have incurred problems of patient attrition within the nursing
homes, given the statewide average attrition rate of 36% per
year. Yet another factor was the strong desire by the sponsor to
finish the analysis of Phase 1 as soon as possible for public
policy planning and budgeting purposes.
Using administrative claims data to measure differences in
drug costs is not without limitations. Drugs may have been
filled without submission of a claim, or nursing homes may
have paid for products such as over-the-counter medications
out of a separate budget. However, this study takes a payer
perspective, and paid claims are the most meaningful measurement
from this perspective. Administrative claims are also poor standalone proxies for measuring changes in quality, particularly in
such areas as adverse effects or health status. On the other hand,
the very large sample sizes involved in our study suggest that
our findings are real and replicable.
As with any nonrandomized observational study, regression
toward the mean must be considered. We chose our comparison
group in the same way we chose study group patients; hence,
both should have equally incurred this regression effect, and it
is, in essence, neutralized for purposes of differential analysis.
Using a payer perspective, we assessed the impact of all drug
claims not just those drugs flagged in profiles from preintervention
screening. It is likely that our broader focus diluted our findings
toward the null. Yet we found important drug cost differences
on a PPPM basis.
■■ Conclusions
A program of supplemental pharmacist review targeting
patients with high drug use and the potential for multiple drug
therapy problems was successful in generating changes in drug
therapy. We believe that involving pharmacists and physicians
in the creation of PDTP alerts was crucial to widespread adoption.
The changes in drug therapy that resulted from a single
(compensated) pharmacist retrospective review significantly
reduced the number of PDTP alerts at follow-up. Currently,
regulations governing DRRs do not explicitly focus on costeffectiveness or cost reductions of pharmaceuticals received by
patients, nor do they explicitly compensate reviewers for such
services. Results from this study suggest that a program to
encourage and compensate pharmacists for conducting focused
reviews of drug therapy regimens for targeted high-risk patients
as a supplement to usual mandated review activities can lower
drug therapy costs and maintain or enhance the quality of drug
therapy. Interventions by pharmacists were economically beneficial when labor costs and savings in drug costs are considered.
Elements of this program can be applied to both ambulatory
and long-term-care settings.
ACKNOWLEDGMENTS
The authors thank the North Carolina Long Term Care Pharmacy Alliance;
North Carolina Office of Research, Demonstrations and Rural Health
Development; North Carolina Medical Directors Association; North Carolina
Vol. 11, No. 7
www.amcp.org
Health Care Facilities Association; and the University of North Carolina
School of Pharmacy at Chapel Hill for their endorsement of this project. The
authors also thank CoraLynn B. Trewet, MS, PharmD, University of Iowa, Des
Moines, and Brandy Newman, PharmD, King Soopers Pharmacy, Aurora, CO,
for their assistance with data compilation and manuscript guidance.
DISCLOSURES
Financial and technical support of this project was provided by AccessCare,
Inc., a provider network within Community Care of North Carolina, and the
American Foundation for Pharmaceutical Education and was obtained by
authors Dale Christensen and Steven E. Wegner (AccessCare) and Troy K.
Trygstad (AFPE). Trygstad, Wegner, and author Jennifer Garmise are employed
by AccessCare. The authors disclose no potential bias or conflict of interest
relating to this article.
Author Trygstad served as principal author of the study. Study concept and
design were contributed by Trygstad, Christensen, Wegner, and author Robert
Sullivan. Analysis and interpretation of data were contributed by Trygstad,
Christensen, and Sullivan. Drafting of the manuscript was primarily the work
of Trygstad, and its critical revision was the work of all authors. Statistical
expertise was contributed by Trygstad and Christensen.
REFERENCES
1. Heffler S, Smith S, Keehan S, Clemens MK, Won G, Zezza M. Health spending
projections for 2002-2012 [Web exclusive, February 7, 2003]. Health Aff. Available
at: http://www.healthaffairs.org. Accessed Feburary 24, 2005.
2. Hanlon JT, Artz MB. Drug-related problems and pharmaceutical care: what
are they, do they matter, and what’s next? Med Care. 2001;39:109-12.
3. Hanlon JT, Schmader K, Gray S. Adverse drug reactions. In: Delafuente JC,
Stewart RB, eds. Therapeutics in the Elderly. 3rd ed. Cincinnati, OH: Harvey
Whitney Books; 2000:289-314.
4. Hanlon JT, Schmader KE, Ruby CM, Weinberger M. Suboptimal prescribing
in elderly inpatients and outpatients. J Am Geriatr Soc. 2001;49(2):200-09.
5. McPeak ME. State-by-state look at Medicaid retail pharmacy delivery: trials
and triumphs. J Manag Care Pharm. 1998;4(6):546-47, 551-52, 554.
6. Carter BL, Chrischilles EA, Scholz D, Hayase N, Bell N. Extent of services
provided by pharmacists in the Iowa Medicaid Pharmaceutical Case
Management program. J Am Pharm Assoc. 2003;43(1):24-33.
7. Christensen DB, Neil N, Fassett WE, Smith DH, Holmes G, Stergachis A.
Frequency and characteristics of cognitive services performed by pharmacists
in response to a financial incentive. J Am Pharm Assoc. 2000;41:609-17.
8. Medication Therapy Management Services Definition and Program Criteria.
Academy of Managed Care Pharmacy, American Association of Colleges of
Pharmacy, American College of Apothecaries, American College of Clinical
Pharmacy, American Society of Consultant Pharmacists, American Pharmacists
Association, American Society of Health Systems Pharmacists, National
Association of Boards of Pharmacy, National Association of Chain Drug Stores,
National Community Pharmacists Association, National Council of State
Pharmacy Association Executives. Approved July 27, 2004. Available at:
http://www.aphanet.org/AM/Template.cfm?Section=APhA_Resources_Medicare
&Template=/CM/ContentDisplay.cfm&ContentID=1681. Accessed June 15,
2005. Also available at: www.aacp.org/Docs/MainNavigation/Resources/
6308_MTMServicesDefinitionandProgramCriteria27-Jul-04.pdf. Accessed
June 15, 2005.
9. American Pharmacists Association, National Association of Chain Drug
Stores. Medication therapy management in community pharmacy practice:
core elements of an MTM service. April 29, 2005. Available at:
http://www.aphanet.org/AM/Template.cfm?Template=/CM/ContentDisplay.cfm
&ContentID=3303. Accessed June 15, 2005.
www.amcp.org
10. The Centers for Medicare and Medicaid Services. Medication therapy
management: excerpt from the Medicare Prescription Drug Benefit Final Rule
42 CFR Parts 400, 403, 411, 417, 423. Available at: http://www.cms.hhs.gov/
medicarereform/pharmacy/mtm.pdf. Accessed June 15, 2005.
11. Clark TR, ed. Nursing Home Survey Procedures and Interpretive Guidelines.
Alexandria, VA.: American Society of Consultant Pharmacists; 1999.
12. Bootman JL, Harrison DL, Cox E. The health care cost of drug-related
morbidity and mortality in nursing facilities. Arch Intern Med. 1997;157:208996.
13. Grymonpre RE, Mitenko PA, Sitar DS, et al. Drug-associated hospital
admissions in older medical patients. J Am Geriatr Soc. 1988;36:1092-98.
14. North Carolina Medicaid, Table 10, state fiscal year 2003, Medicaid
service expenditures by recipient group. Available at: http://www.dhhs.
state.nc.us/dma/ncms.htm. Accessed August 13, 2005.
15. North Carolina Medicaid, Table 11, state fiscal year 2003, Medicaid
service expenditures by recipient group. Available at: http://www.dhhs.
state.nc.us/dma/ncms.htm. Accessed August 13, 2005.
16. Christensen DB, Trygstad TK, Sullivan R, et al. A pharmacy management
intervention for optimizing drug therapy for nursing home patients. Am J
Geriatr Pharmacother. 2004;2(4):248-56.
17. Beers MH. Explicit criteria for determining potentially inappropriate
medications use by the elderly. Arch Intern Med. 1997;157:1531-36.
18. Kaufman MB, Brodin KA, Sarafian A. Effect of prescriber education on
the use of medications contraindicated in older adults in a managed Medicare
population. J Manag Care Pharm. 2005;11(3):211-19.
19. Christensen DB, Trygstad TK. Assessment of the Polypharmacy Initiative
in Nursing Homes: a preliminary analysis. Report to Carmen Hooker Odom,
Secretary, North Carolina Department of Health and Human Services. July
2002.
20. A Soumerai SB, Lipton HL. Computer-based drug-utilization review—
risk, benefit, or boondoggle? N Engl J Med. 1995;332(24):1641-45.
21. Hennessy S, Bilker WB, Zhou L, et al. Retrospective drug utilization
review, prescribing errors, and clinical outcomes. JAMA. 2003;290(11):1494-99.
22. Stuart B, Fahlman C. Outcomes of prospective drug-use review of betaagonist inhaler use in an elderly Medicaid population. Clin Ther. 1999;21(12):
2094-112.
23. Kidder D, Bae J. Evaluation results from prospective drug utilization
review: Medicaid demonstrations. Health Care Financ Rev. 1999;20(3):107-18.
24. Smith DH, Christensen DB, Stergachis A, Holmes G. A randomized
controlled trial of a drug use review intervention for sedative hypnotic
medications. Med Care. 1998;36(7):1013-21.
25. Walker S, Willey CW. Impact on drug costs and utilization of a clinical
pharmacist in a multisite primary care medical group. J Manag Care Pharm.
2004;10(4):345-54.
26. Zuckerman IH, Weiss SR, McNally D, Layne B, Mullins CD, Wang J.
Impact of an educational intervention for secondary prevention of myocardial
infarction on Medicaid drug use and cost. Am J Manag Care. 2004;10(7 pt 2):
493-500.
27. Crownover BK, Unwin BK. Implementation of the Beers criteria: sticks
and stones—or throw me a bone. J Manag Care Pharm. 2005;11(5):416-17.
28. Curtiss FR. Evidence-based medicine: which drugs are truly contraindicated for use in older adults? J Manag Care Pharm. 2005;11(3):259-60.
Vol. 11, No. 7
September 2005
JMCP
EDITORIAL
■■ Fish Oil for Heart Disease—Happy to Be in Second Place
From 1970-1977, the Minnesota Vikings played in the National
Football League’s championship game—the Super Bowl—four
times. They lost every time. The Denver Broncos repeated the
feat until finally winning in their fifth attempt. Both teams were
mocked and pitied for their futility; however, there were worse
fates than second place. The rest of their conference teams had
spent millions of dollars and countless hours of work for the
same opportunity and fell short. Obviously, first place was
the goal, but coming in second was still better than 26 other
competitors.
In the highly competitive field of lipid treatment, many
groups strive to stand out from their peers. Niacin, fibrates, and
resins were the early favorites but were overtaken by statins.
Statins are the usual first-line recommended medical treatment
by multiple expert panels for high-risk primary and secondary
prevention of atherosclerotic cardiovascular disease (ASCVD).1-4
The large volume of morbidity and mortality data makes them
the undisputed champions for treating dyslipidemia.5-8
Ezetimibe has been touted as add-on therapy for additional
low-density lipoprotein cholesterol (LDL-C) lowering; however,
we are still waiting for data to show that the LDL-C effect of this
agent results in any benefit for myocardial infarction, stroke,
cardiovascular mortality, or overall mortality.9 Hormone replacement therapy also showed value for LDL-C lowering in
observational studies; however, we now know that stroke and
myocardial infarction risk actually increase despite the improvement in lipid values.10 Obviously, other known factors are at
work, such as antiinflammatory and antiplatelet effects to
explain why some agents that lower LDL-C help morbidity and
mortality, while others are neutral or actually harmful.
Into the fray, a new challenger is now gaining notoriety—
omega fatty acids, also known as fish oil supplements.
(Terminology is complex regarding marine versus plant sources
and subcomponents of omega fatty acids. An extensive description is beyond the scope of this editorial, and excellent reviews
are available.11) Many clinicians have rightfully been leery of
unproven claims from the herbal sector. As food supplements,
they carry the classic label: “(T)his product is not intended to
diagnose, treat, cure, or prevent any disease.” A few notable
exceptions exist, such as the Rotta preparation of glucosamine
for joint pain, but the majority of agents lack substantive
support of effectiveness.12
Evidence has been building lately for omega fatty acids,
especially for secondary ASCVD prevention. In a randomized
controlled trial of patients with recent myocardial infarction,
2,836 individuals were given omega fatty acids (dose 850-882
mg) while 2,828 were given placebo. After 3.5 years, the treatment group had a lower rate for the combined end point of
overall mortality, nonfatal stroke, and recurrent infarction relative to placebo (12.3% vs. 14.6%, number needed to treat = 44,
P = 0.048).13 A meta-analysis in 2002 pooled 11 trials involving
almost 16,000 patients for primary and secondary prevention,
performed over the previous 30 years.14 With an average followup of 20 months and dosing ranging from 0.3-6.0 g/day, results
showed omega fatty acids significantly decreased the risk of fatal
myocardial infarction (risk ratio = 0.7; 95% confidence interval
[CI], 0.6-0.80), sudden death (risk ratio=0.6; 95% CI, 0.6-0.9),
and overall mortality (risk ratio = 0.8; 95% CI, 0.7-0.9).
Pulling the evidence into tighter focus was a meta-analysis in
2005 that compared true end points, such as overall mortality,
among the various classes of lipid-lowering agents.15 In more
than 100,000 patients enrolled in 35 statin trials, the risk ratio
for overall mortality was 0.87 for statins compared with placebo
(95% CI, 0.81-0.94), with a 20% average reduction in total
cholesterol. Surprisingly, for overall mortality in 15 studies
involving 20,000+ patients with preexisting heart disease treated
with omega fatty acids, the risk ratio was 0.77 (95% CI, 0.630.94), despite a negligible 2% average reduction in total cholesterol.
Statins lowered cholesterol 10-fold more than omega fatty acids
yet did not outperform on mortality measures in patients with
ASCVD. Fibrates, niacin, resins, and diet therapy all failed to
show significance in lowering overall mortality. For primary
prevention with statins, it is necessary to treat 228 persons
(95% CI, 123-2,958) for 3.3 years. In patients with known
heart disease, 50 patients (95% CI, 38-78) would have to
be treated with a statin to prevent one additional death, and
44 patients (95% CI, 31-84) would need to be treated with fish
oil to prevent one additional death, each over an average 4.4
years of therapy (excluding one low-quality study).15
Not all reviews of omega fatty acids have been positive. An
earlier Cochrane meta-analysis that was last reviewed in 2004
failed to find a reduction in overall mortality for omega
fatty acids. No safety concerns were found, but the authors
recommended additional research.16 An evidence report from
the Agency for Healthcare Research and Quality evaluated
123 trials for the effect of omega fatty acids on intermediate
markers for ASCVD. Consistent with prior studies, they found
a reduction of triglycerides ranging from 10% to 33%, but
negligible effects on LDL-C or high-density lipoprotein cholesterol.17 The report did not assess the effect on overall mortality.
Despite the differences among reviews, the U.S. Food and
Drug Administration determined in November 2004 that
sufficient data existed to approve the first prescription omega
fatty acid agent in the United States; the agent was first
launched in Austria in July 2002.18,19 Containing 1 g per capsule,
the agent is indicated for adjunct therapy to diet at
4 g daily to reduce very high (≥ 500 mg/dL) triglyceride levels
in adult patients. Of note, the prescription agent will vary
markedly from over-the-counter preparations. For example, a
1 g capsule of the prescription agent will have roughly triple the
amount of fish oil as a popular 1 g capsule available commercially
at present (eicosapentaenoic acid 465 mg vs. 180 mg; docosahexaenoic acid 375 mg vs. 120 mg).20 It is expected that the
584 Journal of Managed Care Pharmacy JMCP September 2005 Vol. 11, No. 7 www.amcp.org
Editorial
manufacturer will push for approval of 1 g daily dosing in postmyocardial infarction patients for secondary prevention of
ASCVD.
Although more studies are needed, omega fatty acids are
now standing out from the crowd. They will likely never replace
statins as first-line therapy for dyslipidemia, but they have more
data to support a benefit for reducing mortality in patients with
preexisting heart disease than do the other antilipid agents.
Central to their success in population management of coronary
heart disease and mortality risk will be low discontinuation
rates. As discussed previously in JMCP, 18-month discontinuation rates greater than 50% are common for all antilipid drug
classes.21 Despite various hurdles, omega fatty acids are now
ready to take over second place in the lineup for treatment of
hypertriglyceridemia and secondary prevention of ASCVD.
Brian K. Crownover, MD, FAAFP,
Lt. Col., MC, USAF
Associate Editor
[email protected]
7. Ross DS, Allen IE, Connelly JE, et al. Clinical outcomes in statin treatment
trials. A meta-analysis. Arch Intern Med. 1999;159:1793-1802.
8. MRC/BHF Heart Protection Study of cholesterol-lowering with simvastatin
in 5,963 people with diabetes: a randomised placebo-controlled trial. Lancet.
2003;361:2005-16.
9. Ballantyne CM, Houri J, Notarbartolo A, et al. Effect of ezetimibe coadministered with atorvastatin in 628 patients with primary hypercholesterolemia.
Circulation. 2003;107:2409-15.
10. Writing Group for the Women's Health Initiative Investigators. Risks and
benefits of estrogen plus progestin in healthy postmenopausal women.
Principal results from the Women's Health Initiative randomized controlled
trial. JAMA. 2002;288:321-33.
11. Kris-Etherton PM, Taylor DS, Yu-Poth S, et al. Polyunsaturated fatty acids
in the food chain in the United States. Am J Clin Nutr. 2000;71(1)179S-188S.
12. Towheed TE, Maxwell L, Anastassiades TP, et al. Glucosamine therapy for
treating osteoarthritis. Cochrane Database Syst Rev. 2005; issue 2; art. no.:
CD002946.pub2. DOI: 10.1002/14651858.CD002946.pub2.
13. GISSI-Prevenzione investigators. Dietary supplementation with n-3
polyunsaturated fatty acids and vitamin E after myocardial infarction: results
of the GISSI-Prevenzione trial. Lancet. 1999;354:447-55
14. Bucher HC, Hengstler P, Schindler C, Meier G. N-3 polyunsaturated fatty
acids in coronary heart disease: a meta-analysis of randomized controlled
trials. Am J Med. 2002;112(4):298-304.
15. Studer M, Briel M, Leimenstoll B, Glass TR, Bucher HC. Effect of different
antilipidemic agents and diets on mortality. A systematic review. Arch Intern
Med. 2005;165:725-30.
DISCLOSURE
The author is a board-certified family physician assigned to Eglin AFB Florida,
where he serves as residency program director. The opinions and assertions
contained herein are the private views of the author and are not to be
construed as official or as reflecting the views of any organization, including
the U.S. Air Force medical department or the U.S. Air Force. He discloses no
potential bias or conflict of interest relating to this editorial.
REFERENCES
1. Veterans Health Administration, Department of Defense. VHA/DoD clinical
practice guideline for the management of dyslipidemia in primary care.
Washington, DC: Veterans Health Administration, Department of Defense;
December 2001.
16. Hooper L, Thompson RL, Harrison RA, et al. Omega 3 fatty acids for
prevention and treatment of cardiovascular disease. Cochrane Database Syst
Rev. 2004, issue 4; art. no.: CD003177.pub2. DOI: 10.1002/14651858.
CD003177.pub2.
17. Balk E, Chung M, Lichtenstein A, et al. Effects of omega-3 fatty acids on
cardiovascular risk factors and intermediate markers of cardiovascular disease.
Summary, Evidence Report/Technology Assessment: Number 93. Rockville,
MD: Agency for Healthcare Research and Quality; March 2004. AHRQ
Publication no. 04-E010-1. Available at: http://www.ahrq.gov/clinic/
epcsums/o3cardrisksum.htm. Accessed July 2, 2005.
18. Omacor [product label]. Washington, DC: U.S. Food and Drug
Administration; 2005. Available at: http://www.fda.gov/cder/foi/label/2004/
21654lbl.pdf Accessed July 2, 2005.
2. Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical
trials for the National Cholesterol Education Program Adult Treatment Panel
III guidelines. Circulation. 2004;110(2):227-39.
19. Omacor [product information]. Marietta, GA: Solvay Pharmaceuticals;
2005. Available at: http://www.solvaypharmaceuticals.com/html/products/
Cardiology/omacor.html#Worldwide%20avail. Accessed July 2, 2005.
3. Haffner SM. Dyslipidemia management in adults with diabetes. Diabetes
Care. 2004;27(suppl 1):S68-S71.
20. GNC Fish Body Oils 1000 [product information]. Pittsburgh, PA: General
Nutrition Centers, Inc.; 2005. Available at: http://www.gnc.com/
productDetails.aspx?id=887911&lang=en. Accessed July 2, 2005.
4. Institute for Clinical Systems Improvement. Lipid management in adults.
Bloomington, MN: Institute for Clinical Systems Improvement; July 2004.
5. Scandinavian Simvastatin Survival Study Group. Randomized trial of
cholesterol lowering in 4,444 patients with coronary heart disease: the
Scandinavian simvastatin survival study (4S). Lancet. 1994;344:1383-89.
21. Abugosh SM, Kogut SJ, Andrade SE, Larrat EP, Gurwitz JH. Persistence
with lipid-lowering therapy: influence of the type of lipid-lowering agent and
drug benefit plan option in elderly patients. J Manag Care Pharm. 2004;10(5):
404-11.
6. The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID)
Study Group. Prevention of cardiovascular events and death with pravastatin
in patients with coronary heart disease and a broad range of initial cholesterol
levels. N Engl J Med. 1998;339:1349-57.
www.amcp.org
Vol. 11, No. 7
September 2005
JMCP
EDITORIAL
■■ Medication Therapy Management Services
for Long-term Care Patients: No Road Maps for
Those Trying to Find Their Way
As required by the Medicare Modernization Act,1 organizations
offering Medicare Part D prescription drug plans (PDPs) are
obligated to provide medication therapy management programs
(MTMPs) as part of the benefit. While medication therapy
management (MTM) has been performed by pharmacists in a
variety of settings for years, a consensus definition of MTM was
not developed until last year.2 As defined in the Act, MTMPs are
“furnished by a pharmacist [and] designed to assure . . . that covered part D drugs are appropriately used to optimize therapeutic
outcomes through improved medication use, and to reduce the
risk of adverse events, including adverse drug interactions.”1
While the Act defines the beneficiaries who are targeted for
MTM (those with multiple chronic conditions, taking multiple
medications, and likely to have high drug expenses) and basic
elements of the program (promotion of enhanced enrollee
understanding of the appropriate use of medications, adverse
event risk reduction, increased medication adherence, and
detection of adverse drug events and patterns of prescription
drug overuse or underuse), the government has taken a decidedly
hands-off approach to regulating MTMPs. Citing insufficient
standards and performance measures to evaluate MTMPs and
the inability to determine what requirements would enhance
MTMPs and improve patient outcomes, the Centers for Medicare
and Medicaid Services (CMS) has given what it considers maximum flexibility to plans so they may develop MTMPs that can
achieve the goal of improving therapeutic outcomes.
Organizations that bid on the Part D drug plans have
remained silent about their MTMP plans, but some guidance is
available as to what can be expected for MTM services at certain
pharmacy settings. In April 2005, the American Pharmacists
Association and the National Association of Chain Drug Stores
Foundation released a model framework for implementing
MTM services in a community pharmacy setting.3 The framework (a) provides 9 factors that community pharmacists can
consider when targeting patients for MTM services, (b) identifies
5 core components of MTM programs in community pharmacy
(medication therapy review, a personal medication record,
a medication action plan, intervention and referral, and
documentation and follow-up) and describes pharmacist
responsibilities for each, and (c) provides sample personal
medication records and medication action plans for community
pharmacists to use in their programs. Regarding senior care, the
American Society of Consultant Pharmacists (ASCP) released an
issue paper on MTM services for ambulatory Medicare beneficiaries
in April 2004, reviewing the goals of MTM services in the
elderly and the types of MTM services that a pharmacist could
provide to ambulatory seniors.4 An MTMP for low-income
ambulatory seniors was described by Stebbins et al. in a recent
issue of JMCP.5
However, there is little guidance for developing MTM services
for those residing in long-term care (LTC) facilities. At this time,
the only point that is clear is that MTM services need to be
distinct from the monthly drug regimen review (DRR) process
mandated for all LTC residents. While the DRR process may
save as much as $3.6 billion per year in medication-related
problems, these problems still generate an estimated $4 billion
in annual costs in the LTC setting.6 Inappropriate prescribing
remains a problem for the nation’s 1.8 million LTC and assistedliving residents and is a significant factor contributing to the
risks for morbidity and mortality.7-9
While the deadline has passed for submitting bids for Part D
PDPs for 2006, organizations preparing bids for PDPs for 2007
and beyond should look for evidence of success in MTM services
in the LTC arena when formulating their bids. In this issue,
Trygstad et al. describe results of the North Carolina
Polypharmacy Initiative, a targeted drug therapy management
consultation where a pharmacist reviews medical records and
Medicaid pharmacy claim drug profiles to determine if a potential
drug therapy problem alert requires action and, if so, to make
interventions and determine if a therapy change was enacted.10
As measured by the study objectives, the program was successful
because it was associated with a reduction in potential drug
therapy problem alerts and median per-patient-per-month
prescription drug costs. The program further demonstrated the
utility of consultant pharmacists in the LTC setting since physicians
implemented nearly 60% of pharmacist recommendations.
However, the study by Trygstad et al. did not look at health
outcomes and therefore did not determine if the drug therapy
changes would maintain or improve the quality of care that
residents received. It is unlikely that a consultant pharmacist
would make a recommendation that would result in lower
quality of care for an LTC resident, but the study by Trygstad
et al. measured only intermediate (process of care) and drug
cost outcomes. Research is still necessary to demonstrate that a
pharmacy intervention program designed for multiple disease
states in LTC leads to improvements in resident health
outcomes. Such research would provide powerful evidence of
an intervention’s utility. It is easy to sympathize with the
researchers; by intervening in many different disease states, it is
very difficult to quantify the impact of the program on LTC
resident health outcomes. Limiting the study to retrospective
pharmacy claims data makes even more remote the possibility
of determining health outcomes.
In the MTMP described in the study, the pharmacist had the
ability to recommend that drug therapy be withdrawn, modified,
or added, as each case required. Changes to drug therapy
impact more than prescription drug expenses; these changes
can have an effect on immediate and longer-term medical
expenses as well. However, the study only presented prescription
drug costs and did not report the impact of the intervention on
medical costs or overall health care costs. The true cost for a
Editorial
health system of an intervention or series of interventions is
best assessed by examining overall costs.
How Part D providers choose to examine the costs of
MTMPs may very well depend upon the type of program they
are offering. Medicare Advantage prescription drug (MA-PD)
plans are at risk for overall health costs, so their level of interest in
the impact of a MTMP on overall costs is likely to be greater
than a PDP, which is not at risk for overall health costs. One
must wonder if the Part D program will suffer from the silo
approach to health care finance that managed care plans have
struggled to overcome. CMS’s proposal11 for linking deidentified
Part D prescription claim data to hospital, physician, and other
medical utilization data in a public database may allow health
policy analysts to compare prescription drug expenses and
overall health care expenses for PDPs and MA-PDs.
More questions than answers presently remain pertaining to
MTM services in LTC facilities. While the forthcoming report
from the Department of Health and Human Services on
standards of practice for pharmacy services (including clinical
services) provided to patients in LTC settings will be a useful
tool for PDP developers to consider when designing or refining
MTMPs, the experiences that the LTC industry encounter starting
on January 1, 2006 (“Part D-Day”), will be even more valuable.
The sharing of MTM best practices by PDPs and LTC facilities
can reduce the trial-and-error approach to MTM services and
can improve the quality of services delivered to LTC residents.
Joshua J. Spooner, PharmD, MS
Director, Clinical and Outcomes Services
Advanced Concepts Institute
University of the Sciences in Philadelphia
Philadelphia, Pennsylvania
[email protected]
DISCLOSURE
The author discloses no potential bias or conflict of interest relating to this
editorial.
REFERENCES
1. Public Law 108-173. The Medicare Prescription Drug, Improvement, and
Modernization Act of 2003. December 8, 2003. Available at: http://www.cms.
hhs.gov/medicarereform/. Accessed August 12, 2005.
2. Medication therapy management services definition and program criteria.
July 27, 2004. Available at: http://www.amcp.org/data/jmcp/Letters-179186.pdf and http://www.aacp.org/Docs/MainNavigation/Resources/
6308_MTMServicesDefinitionandProgramCriteria27-Jul-04.pdf. Accessed
August 12, 2005.
3. The American Pharmacists Association and the National Association of
Chain Drug Stores Foundation. Medication therapy management in community pharmacy practice: core elements of an MTM service. Version 1.0. April
29, 2005. Available at: http://www.aphanet.org/AM/Template.cfm?Template=/
CM/ContentDisplay.cfm&ContentID=3303. Accessed August 12, 2005.
4. The American Society of Consultant Pharmacists. Medication therapy
management services for ambulatory Medicare beneficiaries. April 28, 2004.
Available at: http://www.ascp.com/medicarerx/docs/ASCPMTMS.pdf.
Accessed August 12, 2005.
5. Stebbins MR, Kaufman DJ, Levens Lipton H. The PRICE clinic for lowincome elderly: a managed care model for implementing pharmacist-directed
services. J Manag Care Pharm. 2005;11(4):333-41.
6. Bootman JL, Harrison DL, Cox E. The health care cost of drug-related morbidity and mortality in nursing facilities. Arch Intern Med. 1997;157:2089-96.
7. Gurwitz JH, Field TS, Avorn J, et al. Incidence and preventability of
adverse drug events in the nursing home setting. Am J Med. 2000;109:87-94.
8. Sloane PD, Gruber-Baldini AL, Zimmerman S, et al. Medication undertreatment in assisted living settings. Arch Intern Med. 2004;164:1621-25.
9. Gerety MB, Cornell JE, Plichta DT, Eimer M. Adverse events related to
drugs and drug withdrawal in nursing home residents. J Am Geriatr Soc.
1993;41:1326-32.
10. Trygstad TK, Christensen D, Garmise J, Sullivan R, Wegner SE.
Pharmacist response to alerts generated from Medicaid pharmacy claims in
a long-term care setting: results from the North Carolina Polypharmacy
Initiative. J Manag Care Pharm. 2005;11(7):575-83.
11. Centers for Medicare and Medicaid Services. Medicare prescription drug
data strategy: improving evidence for patient care through the Medicare prescription drug benefit. Available at: http://www.cms.hhs.gov/medicarereform/
CMSPaper-DataStrategyforMedicareDrugBenefitOverview.pdf. Accessed August
8, 2005.
www.amcp.org
Vol. 11, No. 7
September 2005
JMCP
EDITORIAL
■■ Evaluating the Relationship Between Diabetes Treatment
and Health Care Costs—Measuring the Atom With a Yardstick?
In this issue of JMCP, Shetty et al. evaluated the relationship
between poor glycemic control, defined as a glycosylated hemoglobin (HbA1c, now commonly abbreviated as simply A1c)
>7%, and diabetes-related medical and pharmacy costs.1 They
used 18-months of data from medical and pharmacy administrative claims as well as laboratory data. Other studies have been
conducted with similar data sources that also characterized the
relationship of poor glycemic control to health care costs,
hospitalizations, and resources used.2-4 A principal limitation of
each of these analyses was the inability of the investigators to
determine disease duration. In their analysis, Shetty et al.
attempted to adjust for a number of potential confounders,
including age, gender, physician specialty, comorbid conditions,
and total baseline costs. However, they looked at patients for
only a 6-month preindex and a 12-month follow-up period, so
they could not estimate whether patients were newly diagnosed
diabetics or whether they had disease of long-standing duration.
The average diabetic many years after diagnosis is very different
from a patient with new-onset diabetes5; they are more dependent
on insulin, more difficult to control, and have more insulin
resistance. Thus, if investigators had been able to control for the
length of time each patient had the disease, the relationship
between A1c level and costs would have been better elucidated.
Of course, the duration of disease is difficult to determine in
these types of analyses. To measure resource utilization directly,
investigators require health plan data for patients who are
continuously enrolled over the entire study period in order to
ensure that all resources used by the patient are captured.
Because patients frequently change health plans over the course
of their lifetimes, the follow-up period required to determine
disease duration would be prohibitive for this type of analysis.
The ideal database for this type of analysis would include
patient data for an entire lifetime, but such an administrative
claims database does not exist anywhere in the United States.
Alternatively, small-scale analyses could be conducted using
medical and pharmacy data from small, rural communities
where populations are relatively stable and receive health care
from a closed set of hospitals and clinics. However, even then,
a large proportion of patients would likely have to be excluded
because of moving into or out of the geographic area. In addition,
patients would not be representative of the larger population as
a whole, and sample sizes might be too small to ensure adequate
power.
Another limitation to these types of analyses is that
International Classification of Diseases, Ninth Revision (ICD-9)
coding is replete with errors. Shetty et al. used a range of ICD-9
codes indicative of type 2 diabetes and excluded patients with
two or more claims for type 1, presumably because investigators
realized that disease misclassification was likely present in the
database. Interestingly, the investigators reported that 35.5% of
the patients with two or more ICD-9 codes for type 2 diabetes
also had two or more ICD-9 codes for type 1. As a result, more
than one third of the source population had to be excluded
because of an inability to determine which disease the patient
actually had. This illustrates the error inherent in using a database such as this, and one has to ask whether excluding the
patients with conflicting diagnoses really eliminated all of the
misclassification bias. Any residual non-differential misclassification of disease in this database would bias the results of the
analysis to the null hypothesis.6 Thus, we might expect an even
stronger association between a lack of glycemic control and
increased medical and pharmacy costs.
The accuracy of administrative databases has been questioned previously. Peabody et al. attempted to characterize the
extent of error in administrative data sources by sending scripted
actors to simulate the symptoms of 4 diseases, including
diabetes.7 Although a study designed to measure how accurately
patients who are feigning illness are diagnosed with that illness
has its own inherent problems, investigators reported that
administrative data sets contained the correct primary diagnosis
only about 57% of the time. They broke down the places in the
chain of events following the patient encounters where errors
were made and found that the clinicians made incorrect
diagnoses 13% of the time, encounter forms were missing or
incomplete 8% of the time, and the data were incorrectly
entered from encounter forms 22% of the time. Even if the
clinicians in this study made incorrect diagnoses because the
patients did not really have the disease, the administrative database was still wrong 30% of the time.
Interestingly, Peabody et al. found that the accuracy of the
database for comorbid or secondary diagnoses among diabetes
patients, such as hypertension and hypercholesterolemia, was
even lower. They reported finding accurate secondary diagnoses
only 36% of the time, and most of the error (42%) occurred
when the data were incorrectly or incompletely entered from
the encounter forms. The investigators attributed the poor quality
of recording of comorbid condition to specific variations in
administrative systems at each site; not all of the sites required
the entry clerks to document secondary diagnoses. In light of
the questionable quality of these data sources with respect to
comorbidities, one has to wonder about the true impact of these
conditions on the outcomes reported by Shetty et al. The investigators adjusted for 11 comorbidities in their analysis, but only
6 of them were significantly correlated with cost (hypertension,
dyslipidemia, retinopathy, nephropathy, neuropathy, and
diseases of the lower extremities). If comorbidity tracking was
as inaccurate in their dataset as it was in that evaluated by
Peabody et al., the impact on the study outcomes could have
been substantial.
An additional problem with these data sources is the degree
of resolution. Encounter forms used in the clinic setting usually
have a list of the top 20 to 30 potential diagnoses clinicians are
Editorial
likely to see. These diagnoses tend to correspond to the most
general classification of a disease, including the NEC (not elsewhere classified) or NOS (not otherwise specified) categories. A
report by a Canadian investigator into the accuracy of ICD-9
reporting of complications in gall-bladder surgery found that
when non-specific codes were interpreted as indicating clinically
important complications, the rate of these complications
appeared much higher.8 The interpretation of nonspecific diagnosis
codes must be tempered with the understanding that these
categories may be used excessively, despite the existence of
more-specific codes that are more accurate. This results in poor
resolution for investigators using these data to answer outcomes
research questions, and yet, they are often asked pointed,
specific questions that require telescopic resolution in order to
answer adequately. Outcomes research is often an attempt to
measure minute details with somewhat inaccurate instruments—
analogous to measuring atomic particles with a yardstick.
Frequently, investigators attempt to answer these questions
anyway, and it is left up to the reader to determine the
magnitude of the limitations of the reported results.
Data sources for outcomes researchers are very crude tools.
Understanding the inaccuracies and limitations of administrative
datasets is critical. To discount all outcomes research because of
these limitations is to be overcautious and is unwarranted.
These databases are valuable, particularly because they allow us
to readily and inexpensively evaluate large numbers of patients
across geographically diverse areas and because they are often
closed systems, capturing all patient encounters over the enrollment period. Few other data sources have such characteristics.
However, readers should bear in mind the limitations when
interpreting and evaluating these studies and when basing
patient care decisions on them. Perhaps a utopian future exists
where national health tracking databases include patient data
for a lifetime and administrative systems document patient data
with high resolution. Until then, our only choice is to keep
sharpening our instruments and learning what we can from the
available sources.
DISCLOSURE
The author is program coordinator, Drug Regimen Review Center, and
research assistant professor, Department of Pharmacotherapy, University of
Utah College of Pharmacy, Salt Lake City, Utah. She discloses no potential bias
or conflict of interest relating to this editorial.
REFERENCES
1. Shetty S, Secnik K, Oglesby AK. Relationship of glycemic control to total
diabetes-related costs for managed care health plan members with type 2
diabetes. J Manag Care Pharm. 2005;11(7):559-64.
2. Gilmer TP, O’Conor PJ, Manning WG, Rush WA. The cost to health plans
of poor glycemic control. Diabetes Care. 1997;20:1847-53.
3. Menzin J, Langley-Hawthorne C, Friedman M, Boulanger L, Cavanaugh R.
Potential short-term economic benefits of improved glycemic control: a managed care perspective. Diabetes Care. 2001;24(1):51-55.
4. Wagner EH, Sandhu N, Newton KM, McCulloch DK, Ramsey SD, Grothaus
LC. Effect of improved glycemic control on health care costs and utilization.
JAMA. 2001;285:182-89.
5. Fox CS, Sullivan L, D’Agostino RB Sr, Wilson PW. Framingham Heart
Study. The significant effect of diabetes duration on coronary heart disease
mortality: the Framingham Heart Study. Diabetes Care. 2004;27(3):704-08.
6. Rothman KJ, Greenland S. Precision and validity in epidemiologic studies.
In: Rothman KJ, Greenland S, eds. Modern Epidemiology. Philadelphia, PA:
Lippincott Williams & Wilkins; 1998:115-34.
7. Peabody JW, Luck J, Jain S, Bertenthal D, Glassman P. Assessing the accuracy
of administrative data in health information systems. Med Care. 2004;42:
1066-72.
8. Taylor B. Common bile duct injury during laparoscopic cholecystectomy in
Ontario: does ICD-9 coding indicate true incidence? Can Med Assn J. 1998;
158:481-85.
Joanne LaFleur, PharmD
Associate Editor
[email protected]
www.amcp.org
Vol. 11, No. 7
September 2005
JMCP
EDITORIAL SUBJECTS—IN THIS ISSUE AND IN PREVIOUS ISSUES
■■ COX-2 Inhibitors: Little or No GI Protection,
Increased Risk of Cardiovascular Events, High Cost,
and Other Class-less Effects
On September 30, 2004, the U.S. Food and Drug
Administration (FDA) requested the market withdrawal of
rofecoxib (Vioxx) due to safety concerns associated with an
apparent increased risk of cardiovascular events, particularly
heart attack and stroke.1 On February 1, 2005, about 9,000
physicians associated with Kaiser Permanente in northern and
southern California agreed to stop writing prescriptions for
valdecoxib (Bextra).2 The Kaiser notice regarding the moratorium
cited “significant concerns raised about the cardiovascular safety
of Bextra…” and “compelling” evidence about cardiovascular
safety risks and “modest benefit” for valdecoxib. On April 7,
2005, the FDA asked the manufacturer of valdecoxib to
voluntarily withdraw the drug from the market.3 Steve Galson,
acting director of the FDA Center for Drug Evaluation & Research,
said that there was “no added advantage, and a special risk,” in
the higher rate of adverse skin reactions with Bextra, and “the
cardiovascular risks of these drugs are what we consider a class
effect.”4 On August 1, 2005, the FDA approved celecoxib
(Celebrex), the sole remaining cyclooxygenase (COX)-2
inhibitor in the U.S. market, for an additional (sixth) indication
for the relief of signs and symptoms associated with ankylosing
spondylitis, but with the label warning of increased risk of
serious cardiovascular thrombotic events, myocardial infarction,
and stroke.5
The class effect of COX-2 inhibitors on increased cardiovascular risk remains controversial. Using administrative claims
data from Kaiser Permanente in California, Graham et al.
evaluated 2,302,029 person-years of follow-up for users of nonselective nonsteroidal anti-inflammatory drugs (NSAIDs) and
celecoxib and rofecoxib.6 There were 8,143 cases of serious
coronary heart disease, of which 2,210 (27.1%) were fatal. The
multivariate adjusted odds ratio for the incidence of serious
coronary heart disease for rofecoxib (all doses) was 1.59 (95%
confidence interval [CI], 1.10-2.32; P = 0.015) compared with
celecoxib. The authors concluded that rofecoxib use increases
the risk of serious coronary heart disease compared with
celecoxib use, and naproxen use does not protect against
serious coronary heart disease.
The evidence produced by the scrutiny of administrative
claims data and results from clinical trials suggests that the
COX-2 inhibitors do not have equivalent (class) effects in either
protection from adverse gastrointestinal (GI) effects or in their
adverse cardiovascular effects. It is important to remember that
the FDA (a) allowed the manufacturer to add to the label for
rofecoxib (Vioxx) the results of the VIGOR (Vioxx
Gastrointestinal Outcomes Research) trial that suggested a GI
protective effect compared with naproxen, (b) never permitted
a claim of less GI harm for any COX-2 inhibitor except
rofecoxib, and (c) rejected the request from the manufacturer to
JMCP
September 2005
Vol. 11, No. 7
drop the warning of possible GI adverse effects from the label
for celecoxib.7 Comparison of the adverse event data across all
NSAIDs and COX-2 inhibitors shows an incidence of myocardial
infarction of <1% for all NSAIDs except tolmetin (3% to 9%),
rofecoxib (3.5% to 10%), valdecoxib (1.6% to 2.1%), celecoxib
(<2%), and meloxicam (Mobic, <2%).8
The results of the CLASS (Celecoxib Long-term Arthritis
Safety Study) trial were published in September 2000 by
Silverstein et al., including the finding that the annualized
incidence rates of upper GI ulcer complications were 0.76% for
celecoxib versus 1.45% for ibuprofen or diclofenac (P = 0.09).9
Based largely upon the results of the CLASS trial, the Celebrex
label was revised in June 2002 to state, “Differences in the
incidence of complicated ulcers between Celebrex and the
combined group of ibuprofen and diclofenac were not statistically
significant.” The FDA written statement elaborated that
Celebrex “did not show a safety advantage in upper GI events
for Celebrex compared to either ibuprofen or diclofenac.”10
Under the Precautions, the Celebrex label states, “Because of its
lack of platelet effects, Celebrex is not a substitute for aspirin for
cardiovascular prophylaxis.” A wealth of information regarding
the advisories, the request for market withdrawal of valdecoxib
on April 7, 2005, and the history of label changes for the
COX-2 inhibitors can be found on the FDA Web site.11
On April 7, 2005, the FDA asked manufacturers of all
marketed prescription NSAIDs, including celecoxib, to revise
the labeling (package insert) for their products to include a
boxed warning, highlighting (a) the potential for increased risk
of cardiovascular events and (b) the serious, potential
life-threatening GI bleeding associated with their use.11 The
revised celecoxib label also included the warning about longterm use.12 The additional warning for celecoxib was based, in
part, on results of the Adenoma Prevention with Celecoxib
(APC) trial in which, over a range of 2.8 to 3.1 years of follow-up,
the composite end point of death from cardiovascular causes
(myocardial infarction, stroke, or heart failure) was reached in
7 of 679 patients in the placebo group (1.0%), compared with
16 of 685 patients who received 200 mg celecoxib twice daily
(2.3%; hazard ratio [HR], 2.3; 95% CI, 0.9-5.5), and 23 of 671
patients who received 400 mg celecoxib twice daily (3.4%; HR
3.4; 95% CI, 1.4-7.8).13 The FDA Alert on April 7, 2005, for
celecoxib specifically states, “Celebrex has been associated with
an increased risk of serious adverse cardiovascular (CV) events
in a long-term placebo controlled trial. Based on the currently
available data, FDA has concluded that an increased risk of serious
adverse CV events appears to be a class effect of NSAIDs
(excluding aspirin).”14
Aside from the safety risks associated with the use of COX-2
inhibitors, they are no more efficacious than nonselective
NSAIDs. Simon et al. found that celecoxib and naproxen are
equally effective in treating the symptoms of rheumatoid arthritis
(RA).15 Evidence-based evaluations such as that produced by
www.amcp.org
Editorial Subjects—In This Issue and in Previous Issues
the National Institute for Clinical Excellence conclude that
COX-2 selective inhibitors “are more effective than placebo and
of equivalent efficacy to other NSAIDs in their ability to reduce
pain (resting and active) and to improve physical and global
function in both OA [osteoarthritis] and RA [rheumatoid arthritis]
patients.”16
So, the COX-2 inhibitors appear to share an increased
cardiovascular risk with the NSAIDs (except aspirin), which the
FDA defines as a “class effect,” and they are no more efficacious
than nonselective NSAIDs. The COX-2 inhibitors would suffer
a third blow if they proved to be no more protective for GI bleed
events compared with the nonselective NSAIDs. In real-world
use, Stockl, Cyprien, and Chang, in this issue of JMCP, found no
difference in the risk of GI bleed for the COX-2 inhibitors
compared with NSAIDs in their study of 50 months of administrative claims data from a large managed care organization
(MCO).17 There were 375 cases of GI bleed among 19,201
follow-up years for COX-2 inhibitor users (19.5 cases per
1,000 person-years) versus 228 cases of GI bleed among 12,680
follow-up years for NSAID users (18.0 cases per 1,000 personyears). The risk of GI bleed was not significantly different for
COX-2 inhibitor users compared with nonselective NSAID
users (relative risk [RR] 1.07; 95% CI, 0.90-1.26). Even among
high-risk patients, a significant reduction in the risk of a GI
bleed was not found among COX-2 inhibitor users (RR 0.995;
95% CI, 0.84-1.19).
Examination of the study methods and the characteristics of
the MCO that was the subject of the research by Stockl et al. is
warranted. First, this MCO operated a prior authorization (PA)
requirement for COX-2 inhibitors that affected an estimated
two thirds of the enrolled population. Therefore, one would
expect the COX-2 inhibitor users in this MCO to have a higher
risk of GI bleed compared with a population of health plan
beneficiaries not subject to PA requirements for COX-2
inhibitors. However, this is not evident in the data presented
by Stockl et al. in their Table 3, in which 57.4% of the
COX-2 inhibitor users were categorized at high-risk of GI bleed
compared with 58.2% of the NSAID users. The influence of this
PA program on the study results is not obvious, but it seems
plausible that the population of COX-2 inhibitor users in this
MCO would be more likely to experience a GI bleed than
the general population of COX-2 inhibitor users. Therefore, the
finding of no difference in the rate of GI bleed for COX-2
inhibitor users versus NSAID users seems stronger as a result of
the coincident operation of a PA program for COX-2 inhibitors
in this MCO that may have led to selection bias in the group of
COX-2 inhibitor users.
Second, the prevalence of COX-2 inhibitor and NSAID use is
of interest. In this MCO, 1,038,437 members received one or
more prescriptions for either an NSAID or COX-2 inhibitor
over the 44-month identification period of the study. Assuming
5% turnover among MCO members per year, the prevalence of
www.amcp.org
use of NSAID or COX-2 inhibitors was 31.7% (1,308,437 users
divided by 3.28 million MCO members) during this 44-month
study period. The proportion of COX-2 inhibitor users to total
NSAID/COX-2 inhibitor users was 6.2% after exclusion for discontinuous enrollment, age younger than 18 years, and prior
use of either a COX-2 inhibitor or NSAID in the 6-month
preperiod.
Third, Stockl et al. did not use the type of health plan
(Medicare + Choice [now Medicare Advantage] or commercial)
in their propensity matching, with the result that a smaller
proportion (53.1%) of COX-2 inhibitor users were enrolled in
Medicare + Choice compared with NSAID users (55.7%). While
this 2.6% absolute difference (5% relative difference) is statistically
significant, its practical significance is uncertain, particularly
because there was no difference in the mean age for COX-2
inhibitor users (63.4 years ± 16.3) versus NSAID users (63.3
years ± 16.3). However, most of the Medicare + Choice plans
had annual maximum benefits of $1,000 or less; e.g., in the
Texas health plans of this MCO, the annual maximum benefit
was in the range of $500 to $600, and in 2002, brand drugs
such as the COX-2 inhibitors were excluded from coverage
entirely.18
The potential value of the use of COX-2 inhibitors withers
further under economic evaluation. For administrative claims
data for dates of service in 2003-2004, Gleason et al. found that
denial of coverage of COX-2 inhibitors by a PA program was
associated with savings in pharmacy costs with no increase in
total medical costs.19 The prevalence of use of COX-2 inhibitors
was 4.2% in the pre-PA period and 1% in the post-PA period.
The direct COX-2 inhibitor cost savings were $1.35 per
member per month (PMPM) or $1.00 PMPM after subtraction
of administrative costs and lost drug manufacturer rebates.
In addition to the direct cost savings, there was no evidence of
GI problems after coverage of COX-2 inhibitors ended in 737
patients who were using COX-2 inhibitors before institution of
a PA program. In the 3 months after PA program implementation
(for January 1, 2003, through March 31, 2003), 620 (84.1%) of
737 members had no claims for a COX-2 inhibitor, and during
this period, their pharmacy costs initially declined by 40%
(P <0.001) and their medical costs declined by 18.7%
(P<0.001) and remained significantly lower. Among a subgroup
of 156 members (21.2%) who tried to fill a COX-2 inhibitor
prescription but were denied coverage, pharmacy costs initially
declined by 48.1% (P <0.001) and medical costs declined by
10.3% (P <0.001). These findings are not surprising in light of
the fact that 68% of patients who attempted to obtain a COX-2
inhibitor were denied coverage because they did not meet the PA
criteria for patients at elevated risk for a GI bleed, nearly identical
to the 65% of COX-2 inhibitor users determined not to be high
risk in the work reported in 2003 (for 2000-2001 data) by Cox,
Motheral, Frisse, et al.20
Spiegel et al. performed a cost-utility analysis and found that
Vol. 11, No. 7
September 2005
JMCP
use of a COX-2 inhibitor instead of a nonselective NSAID in
average-risk patients had an incremental cost of $275,809 to
gain 1 additional quality-adjusted life-year (QALY).21 Adding
the potential for adverse cardiovascular events to the pharmacoeconomic analysis pushed the incremental cost to nearly
$400,000 per QALY. The pharmacoeconomic considerations
around the use of COX-2 inhibitors versus other NSAIDs for the
relief of joint pain, OA, RA, and other pain management needs
have contributed to the amount of attention that the COX-2
inhibitors have received in prior issues of JMCP. As pointed out
by Ortiz, no drug class has received more attention in articles
published in JMCP.22
In calendar year 2001, celecoxib and rofecoxib were the
number 5 and number 8 drugs by total expenditure dispensed
in community pharmacies,23 propelled, in part, by the large
spending on direct-to-consumer advertising (DTCA). In the first
quarter of 2001, celecoxib was the second most heavily advertised drug in the United States, and rofecoxib was the third most
heavily advertised drug.24 For all of 2001, the midpoint of the
study period in Stockl et al., roefecoxib and celecoxib were the
top 2 drugs promoted by DTCA, accounting for $135 million
and $130.4 million, respectively.25 Celecoxib was the second
fastest selling drug in U.S. history. In its first 13 weeks on the
market in early 1999, celecoxib had 2.5 million new prescriptions.26 Bull, Conell, and Campen found, in their analysis of data
from 1999, that physician education and use of a clinical
practice guideline in a large health maintenance organization
were associated with a 5-fold difference in the utilization of
COX-2 inhibitors among patients determined to be in the highest-risk decile versus patients in the lowest-risk decile, 8.3%
versus 1.5%, respectively, and the authors speculated
that DTCA probably influenced patient requests for COX-2
inhibitors.27
So, the available evidence indicates that COX-2 inhibitors
(a) are expensive, (b) have no pain relief benefit over NSAIDs,
(c) offer little if any GI protection in the general population of
NSAID users, and (d) may cause increased risk of cardiovascular events. Just how much value does the sole COX-2 inhibitor
remaining on the U.S. market have in patients with a risk of a
GI bleed? Chan et al. found, in a randomized controlled trial,
that 7 of 144 patients (4.9%) who received 200 mg of celecoxib
twice daily plus placebo experienced recurrent GI bleeding
versus 9 of 143 patients (6.3%) who received 75 mg diclofenac
twice daily plus 20 mg omeprazole daily for 6 months. All
patients had arthritis treated by NSAIDs, had suffered a GI bleed,
and were randomly assigned after ulcer healing and testing negative for Helicobacter pylori infection. This absolute difference of
1.5% was not statistically significant.28 It is also important to note
that this study used an endoscopic end point rather than clinical ulcer disease. The minimum 3 mm size of the endoscopic ulcer
would appear to be practically significant, but there is no convincing evidence that endoscopic ulcers are a reasonable surrogate for
JMCP
September 2005
Vol. 11, No. 7
clinical ulcer disease or for ulcer complications.29 So, the study
by Chan et al. suggests that omeprazole 20 mg per day plus
diclofenac 75 mg twice daily is comparable to celecoxib 200 mg
twice daily in the incidence of recurrent GI bleed in patients at
high risk for GI bleed, and this study attempted to account for
the 2 potentially confounding variables in studies of COX-2
inhibitors: (a) aspirin use, which carries a risk of complications
of 1.5% to 2.5% per year,30 and (b) status of infection with
Helicobacter pylori.31 Another way to view the data is that, for the
many patients currently on omeprazole, celecoxib offers no
additional GI bleed protection compared with diclofenac.
If GI bleeding is a significant risk in a given patient, and we
suspend our concerns for cardiovascular death from COX-2
inhibitor use, how do the costs compare? At current discounted
drug prices, the cost of celecoxib 200 mg twice daily is $5.29
per day of therapy versus $1.30 per day for diclofenac 75 mg
twice daily plus omeprazole 20 mg (over-the-counter) once
daily.32 Therefore, the diclofenac + omeprazole combination has
a drug cost 75% lower than the COX-2 inhibitor alternative
therapy in patients at high risk of a GI bleed. Even non-math
majors can appreciate that, for patients at high risk of a GI bleed,
it is possible to treat 4 patients with diclofenac + omeprazole for
the same drug cost to treat 1 patient with celecoxib.
Editor-in-Chief
[email protected]
REFERENCES
1. U.S. Food and Drug Administration. FDA issues public health advisory on
Vioxx as its manufacturer voluntarily withdraws the product. P04-95,
September 30, 2004. Available at: http://www.fda.gov/bbs/topics/news/2004/
NEW01122.html. Accessed August 1, 2005.
2. Bextra suspended by Kaiser; Celebrex Alzheimer’s data under scrutiny.
Green Sheet. February 14, 2005:1, 2.
3. U.S. Food and Drug Administration. Alert for healthcare professionals—
valdecoxib (marketed as Bextra)—April 7, 2005. Available at:
http://www.fda.gov/cder/drug/InfoSheets/HCP/valdecoxibHCP.pdf. Accessed
July 27, 2005.
4. Mathews AW, Hensley S. FDA stiffens painkiller warnings, pushing Pfizer
to suspend Bextra. Wall Street Journal. April 8, 2005:A1, A4.
5. FDA approves new use for Pfizer’s Celebrex; finalizes prescribing information, including expected warnings, for all six approved uses. Available at:
http://www.pfizer.com/pfizer/are/news_releases/2005pr/mn_2005_0801.jsp
Accessed August 2, 2005.
6. Graham DJ, Capen D, Hui R, et al. Risk of acute myocardial infarction and
sudden cardiac death in patients treated with cyclo-oxygenase 2 selective and
non-selective non-steroidal anti-inflammatory drugs: nested case-control study.
Lancet. 2005;365(9458):475-81.
7. Harris G, Carroll J. Panel approves Merck’s claims for arthritis drug. Wall
Street Journal. February 9, 2001:B10.
8. Non-steroidal anti-inflammatory agents, NSAIDs adverse reactions (%).
Facts and Comparisons, CliniSphere Version (ISBN 1-57439-036-8). St. Louis,
MO: Wolters Kluwer Health, Inc.; May 2005.
9. Silverstein FE, Faich G, Goldstein JL, et al. GI toxicity with celecoxib vs
nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. Celecoxib Long-term
Arthritis Safety Study. JAMA. 2000;284:1247-55.
www.amcp.org
10. Harris G. Study finds Celebrex as likely to cause ulcers as older drugs.
Wall Street Journal. June 10, 2002:B3.
22. Ortiz E. Market withdrawal of Vioxx: is it time to rethink the use of
COX-2 inhibitors? J Manag Care Pharm. 2004;10(6):551-54.
11. U.S. Food and Drug Administration. COX-2 selective (includes Bextra,
Celebrex, and Vioxx) and non-selective non-steroidal anti-inflammatory drugs
(NSAIDs). Available at: http://www.fda.gov/cder/drug/infopage/cox2/
default.htm. Accessed July 16, 2005.
23. Curtiss FR. Cost-effective use of COX-2 drugs and NSAIDs. J Manag Care
Pharm. 2002;8(4):295-96.
12. U.S. Food and Drug Administration. FDA statement on halting of a clinical
trial of the COX-2 inhibitor Celebrex. Available at: http://www.fda.gov/bbs/
topics/news/2004/NEW01144.html. Accessed August 1, 2005.
13. Solomon SD, McMurray JJV, Pfeffer MA, et al. for the Adenoma
Prevention with Celecoxib (APC) study investigators. Cardiovascular risk
associated with celecoxib in a clinical trial for colorectal adenoma prevention.
N Engl J Med. 2005;352:1071-80.
14. U.S. Food and Drug Administration. Alert for healthcare professionals—
celecoxib (marketed as Celebrex)—April 7, 2005. Available at:
http://www.fda.gov/cder/drug/infopage/celebrex/celebrex-hcp.pdf.
Accessed July 27, 2005.
15. Simon LS, Weaver AL, Graham DY, et al. Anti-inflammatory and upper
gastrointestinal effects of celecoxib in rheumatoid arthritis. A randomized
controlled trial. JAMA. 1999;282:1921-28.
16. National Institute for Clinical Excellence. Guidance on the use of cyclooxygenase (Cox) II selective inhibitors, celecoxib, rofecoxib, meloxicam and
etodolac for osteoarthritis and rheumatoid arthritis. Technology Appraisal
Guidance–no. 27. Available at: http://www.nice.org.uk/pdf/ coxiifullguidance.pdf. Accessed August 2, 2005.
17. Stockl K, Cyprien L, Chang E. Gastrointestinal bleeding rates among managed care patients newly started on COX-2 inhibitors or nonselective NSAIDs.
24. Burton TM, Harris G. Note of caution—study raises specter of cardiovascular risk for hot arthritis pills. Vioxx and Celebrex marketers dispute the
research, sought to downplay it. Wall Street Journal. August 22, 2001:A1.
25. Burton TM. Backlash is brewing among companies who believe flashy ads
drive up costs. Wall Street Journal. March 13, 2001:B1.
26. Sharpe R. Several deaths show a link to Celebrex. Wall Street Journal. April
20, 1999:B6.
27. Bull SA, Conell C, Dampen DH. Relationship of clinical factors to the use
of COX-2 selective NSAIDs within an arthritis population in a large HMO.
28. Chan FK, Hung LC, Suen BY, et al. Celecoxib versus diclofenac and
omeprazole in reducing the risk of recurrent ulcer bleeding in patients with
arthritis. N Engl J Med. 2002;347:2104-10.
29. Sung JY, Lau JY, Chan FK, Graham DY, How often are endoscopic ulcers
in NSAID trials diagnosed as actual ulcers by experienced endoscopists?
[abstract]. Gastroenterology 2001;120(suppl)1A-597.
30. Derry S, Loke YK. Risk of gastrointestinal haemorrhage with long-term
use of aspirin: meta-analysis. BMJ. 2000;321:1183-87.
31. Graham DY, NSAIDs, Helicobacter pylori, and Pandora’s box. N Engl J Med.
2002;347:2162-63.
32. Prices obtained from www.drugstore.com; e.g., Celebrex price available at:
http://www.drugstore.com/pharmacy/prices/drugprice.asp?ndc=00025152531
&trx=1Z5006. Accessed July 31, 2005.
18. Personal communication with the pharmacy manager for the Texas health
plans of PacifiCare on August 2, 2005.
19. Gleason PP, Williams C, Hrdy S, et al. Medical and pharmacy expenditures after implementation of a cyclooxygenase-2 inhibitor prior authorization
program. Pharmacotherapy. 2005;25(7):924-34.
20. Cox ER, Motherall B, Frisse M, et al. Prescribing COX-2s for patients new
to cyclo-oxygenase inhibition therapy. Am J Manag Care. 2003;9:735-42.
21. Spiegel BM, Targownik L, Dulai G, Gralnek IM. The cost-effectiveness of
cyclooxygenase-2 selective inhibitors in the management of chronic arthritis.
Ann Intern Med. 2003;138(10):795-806.
www.amcp.org
Letters to the Editor
JMCP welcomes letters that serve to clarify subjects published in
previous issues of the Journal or regarding subject matter of interest to
managed care pharmacists. Letters in JMCP are not peer reviewed but are
subjected to editorial review. Letters should be prepared in a word
processing program, preferably Microsoft Word, and submitted electronically at jmcp.msubmit.net. See www.amcp.org for details.
Vol. 11, No. 7
September 2005
JMCP
LETTER
■■ Standardization Is Necessary in the Methods
to Assess the Value of Electronic Prescribing Systems
In the May 2005 issue of JMCP, we reported that an electronic
prescribing system with integrated clinical decision support was
effective at improving prescribing behavior and lowering
prescription costs.1 In contrast, Ross et al. reported in the June
2005 issue of JMCP that a different electronic prescribing
system (without clinical decision support capabilities) had no
impact on generic drug utilization or formulary compliance.2
When designing our study, we also considered assessing
changes in generic prescribing rates as a surrogate marker for
cost savings but, instead, we decided to focus on the actual
differences in prescription costs. Since publishing our results,
we have been asked many times about the impact of our system
on generic drug utilization and have been involved in subsequent study design discussions in which customers wanted to
focus primarily on this outcome measure. Given the ongoing
interest in evaluating the impact of electronic prescribing
systems, we decided to assess the changes in generic drug
utilization observed in our study by conducting a post hoc
analysis of the new prescriptions dataset. These results (1) highlight a potential limitation of prior studies assessing electronic
prescribing systems, (2) confirm that an electronic prescribing
system with integrated clinical decision support can increase
generic prescribing rates, and (3) demonstrate how focusing
only on savings related to increasing generic drug utilization
may substantially underestimate the total savings associated
with electronic prescribing.
Some studies attempting to quantify the impact of electronic
prescribing systems have compared outcomes for physicians
using the system with a group of traditional prescribers over a
single time frame.2,3 For example, Ross et al. retrospectively
compared physicians using an electronic prescribing system
with a group of closely matched controls and found no difference
in generic drug utilization during a 12-month evaluation period
(37.3% versus 36.9%). While these studies can provide useful
information, the results are somewhat difficult to interpret without
also knowing if the groups had similar prescribing practices
at baseline (i.e., prior to implementation of the electronic
prescribing system).
Potential pitfalls related to comparing groups from a single
time period can be demonstrated by examining the generic drug
utilization data from our study. During our 12-month follow-up
period, from June 2002 through May 2003, the generic
dispensing rates were 55.2% in the electronic prescribing group
and 48.7% in the control group. If we had looked only at this
time period, we would have incorrectly assumed that the
electronic prescribing system resulted in a 6.5% difference in
generic drug utilization.
When baseline generic drug utilization is considered, the
results show that the generic dispensing rate increased from
50.8% to 55.2% (8.7% relative increase) in the intervention
TABLE 1
Savings on Brand-Name and Generic
Medications During the 12-Month
Follow-up Period (June 2002-May 2003)1
New Prescriptions Data
Script Volume Average Cost ($) Expenditures ($)
Generic medications
Intervention group
14,712 (55.2%)
16.79
247,014
Control (expected) values* 13,444 (51.3%)
16.35
219,809
Difference
(1,268)
(0.44)
(27,205)
845,594
Brand-name medications
Intervention group
11,962 (44.8%)
70.69
Control (expected) values* 12,786 (48.7%)
76.55
978,768
Difference
824
5.86
133,174
Script Count
Savings/(Loss)
Associated With
Each Script
Total Savings/
(Loss)
Decreased utilization of
brand-name medications
associated with switching
to a generic alternative
824
76.55
63,077
Increased utilization of
generic medications
associated with switching
from a brand-name
medication
(824)
(16.79)
(13,835)
Increased utilization of
generic medications
beyond those related
to brand-to-generic
conversions
(444)
(16.79)
(7,455)
Decreased cost of brandname medications
11,962
5.86
70,097
Increased cost of generic
medications
13,444
(0.44)
(5,915)
Contribution to Savings
Utilization†
Cost
Total savings
105,969
* Control values in Table 1 use the changes observed in the control group to represent
what would have been expected in the intervention group. Because the intervention group
(n = 19) and the control group (n = 19) were closely matched prior to implementation of
the electronic prescribing system, the impact of the system can be determined by comparing the intervention group’s expected values with their actual values.
† Based on changes observed in the control group, the intervention group had 824
fewer brand-name prescriptions than expected and 1,268 more generic drug
prescriptions than expected. The savings calculations in the table assume that 824
of the excess generic prescriptions were related to brand-to-generic conversions;
however, it is possible that some brand-name medications were discontinued rather
than switched to a generic alternative. This assumption does not affect the total
savings amount but, if incorrect, would further decrease the proportion of savings
associated with brand-to-generic switches.
group and from 48.2% to 48.7% (1.0% relative increase) in the
control group. To correct for the difference in baseline generic
dispensing rates, the data can be adjusted using the methodology
Letter
proposed by the Centers for Medicare and Medicaid Services
physician group demonstration project.4 This methodology
assumes that the change observed in the control group
represents normal variation (i.e., what would have been expected
in the intervention group if they had not started using the
electronic prescribing system). In this case, the 1% relative
change observed in the control group is applied to the
intervention group’s baseline value of 50.8% to achieve an
expected generic dispensing rate of 51.3% (i.e., 101% of the
50.8% baseline generic dispensing rate). The difference
between the intervention group’s expected generic dispensing
rate (51.3%) and their actual generic dispensing rate (55.2%)
represents a net absolute increase of 3.9 points.
Although these results confirm that the system was effective
at increasing generic prescribing, as confirmed by the generic
dispensing ratio, it is important to recognize that the savings
related to generic drug utilization account for less than half of
the total savings. As demonstrated in Table 1, the 3.9% higher
generic dispensing rate in the intervention group translates into
824 fewer brand-name prescriptions being dispensed during
the study period. While each avoided brand-name prescription
saved an average of $76.55 (total savings: $63,077), these
savings were partially offset by an increased utilization of generic
medications (total cost: $21,290); therefore, the final savings
associated with increased generic drug utilization was $41,787.
To estimate the final savings, however, differences in the
average cost of brand-name and generic medications should
also be examined. Our data suggest that the system was not only
effective at decreasing utilization of brand-name prescriptions,
but it also had a significant impact on the cost of each remaining
brand-name prescription. The average cost of brand-name
prescriptions was $5.86 (7.7%) lower in the intervention group;
therefore, interventions that focused on improving prescribing
behavior within these drug categories saved an additional
$70,097. These savings were most likely related to electronic
messages that focused on improving brand-to-brand conversions
or helping clinicians choose the most cost-effective dosing
regimens. These cost-related savings on brand-name medications
were partially offset by a $0.44 higher cost for each generic
prescription (total cost: $5,915); therefore, the total savings
attributed to price differences for both brand and generic drug
claims was $64,182.
Combining the savings estimates from interventions affecting
utilization ($41,787) and price ($64,182) results in total
savings of $105,969 (8.8%), or an average of $465 per prescriber
per month. Because all calculations in the post hoc analysis
were based on data adjusted for differences in per-member-permonth (PMPM) utilization, these savings are almost identical to
the savings that we originally estimated based on the PMPM
costs for new prescriptions.1
We hope that these results will encourage future researchers
to assess actual savings related to differences in the cost and
www.amcp.org
utilization of brand and generic medications rather than simply
to focus on changes in generic utilization. We also encourage
future researchers to identify closely matched controls based on
data gathered prior to the implementation of the electronic
prescribing system because we believe that comparing 2 groups
of clinicians during an isolated time frame may result in spurious conclusions.
S. Troy McMullin, PharmD
Clinical Decision Support
Purkinje
10426 Baur Blvd.
St. Louis, MO 63110
[email protected]
DISCLOSURE
The author is an employee of Purkinje, distributor of the electronic prescribing system that was the subject of this study.
REFERENCES
1. McMullin ST, Lonergan TP, Rynearson CS. 12-month drug cost savings
related to use of an electronic prescribing system with integrated decision
support in primary care. J Manag Care Pharm. 2005;11(4):422-32.
2. Ross SM, Papshev D, Murphy EL, Sternberg DJ, Taylor J, Barg R. Effects of
electronic prescribing on formulary compliance and generic drug utilization in
the ambulatory care setting: a retrospective analysis of administrative claims
data. J Manag Care Pharm. 2005;11(5):410-15.
3. Cap, Gemini, Ernst & Young. Allscripts: Touchscript medication management system financial analysis on pharmacy risk pools. Available at
http://www.allscripts.com. Accessed August 31, 2004.
4. Medicare program: solicitation for proposals for the physician group
demonstration project. Fed Regist. 2002;67:61116-29.
Vol. 11, No. 7
September 2005
JMCP
ABSTRACTS
Abstracts From Professional Poster Presentations
at AMCP’s 2005 Educational Conference
T
he following poster presentations have been prepared for
the Academy of Managed Care Pharmacy’s 2005
Educational Conference, October 5-8, 2005, in
Nashville, Tennessee. Poster presentations are selected by the
Program Planning Committee from proposals that are submitted
to AMCP. Authors of posters are responsible for the accuracy and
completeness of data presented in the posters and in the abstracts
published here.
For more information about the studies described below,
please contact the corresponding authors, indicated by an asterisk
(*), whose addresses are listed in full. The names of individuals
who are scheduled to present at the meeting are underlined.
■■ ANALYSIS OF THE CLINICAL AND
ECONOMIC IMPACT OF A STEP-THERAPY SEQUENCE
ON NONSEDATING ANTIHISTAMINE FAILURE RATES
Seifert RD*, Sturm LL. University of Minnesota, College of Pharmacy,
366 Kirby Plaza, 1208 Kirby Dr., Duluth, MN 55812-3095
INTRODUCTION: The purpose of this study is to determine baseline
failure rates for secondary nonsedating antihistamines (NSAs)
and the clinical and economic outcomes of a step-therapy
sequence after patients have received prior treatment with overthe-counter (OTC) loratadine.
METHODS: To establish a baseline, a retrospective study of
patients who failed OTC loratadine was conducted, reviewing
physician prescribing habits for subsequent therapy with
secondary NSAs or other allergy therapies. Subsequently, a steptherapy sequence was implemented in which patients who
failed on OTC loratadine were required to try and fail intranasal
corticosteroids and intranasal antihistamines before receiving
secondary NSA therapy. These prescribing methods were compared for clinical outcomes and economic impact in a managed
care setting.
RESULTS: Prior retrospective analysis demonstrated that 1,713
members received at least one prescription for OTC loratadine
from January 1, 2004, through March 31, 2005. The goal of the
step therapy was to reduce the number of prior authorizations
(PAs) by requiring failure with an alternative therapeutic class
of agents. Of those members, 349 required PA to obtain
approval for use of an alternative NSA. The estimated potential
savings, with a 50% reduction in failure rates, would amount to
approximately $30,000 per year in this population.
CONCLUSIONS: Several published clinical studies have demonstrated that there is a high secondary failure rate when using
JMCP
September 2005
Vol. 11, No. 7
one NSA after another NSA has already been tried. Implementing
step therapy (failure of OTC loratadine and additional failure of
a nasal antihistamine [Astelin] and a formulary corticosteroid)
before applying secondary NSA therapy would provide potential
cost savings in a managed care population.
■■ ANEMIA TREATMENT COSTS IN CANCER PATIENTS
TREATED WITH EPOETIN ALFA OR DARBEPOETIN ALFA
IN A MANAGED CARE POPULATION
Harris HM, DelAguila M, Lopez JM*, McKenzie RS, Piech CT. Ortho
Biotech Clinical Affairs Outcomes, LLC, 430 Route 22 East,
Bridgewater, NJ 08807
INTRODUCTION: Anemia management with epoetin alfa (EPO) or
darbepoetin alfa (DARB) is a common occurrence in cancer
patients in managed care populations. This study compares
anemia-related costs in cancer patients receiving EPO and
DARB.
METHODS: Administrative medical, lab, and pharmacy claims
from more than 30 diverse managed care plans were examined.
Cancer patients aged 18 years and older with at least 1 medical
claim for EPO or DARB from January 2003 through February
2004 were identified. Patients continuously eligible for benefits
for at least 3 months prior to and 4 months after the date of the
initial claim were included. Only patients new to therapy (no
EPO or DARB claims during the 3 months prior to initiating
therapy) and those with at least 1 additional medical claim for
the index agent were included in the study. Costs were
evaluated on a per-patient-month (PPM) basis, for patients with
at least 28 days of therapy.
RESULTS: 6,584 cancer patients (EPO 4,535; DARB 2,049) were
identified. The EPO group was older, had a higher proportion
of men, and had higher baseline rates of cardiac, respiratory,
digestive, and neurologic comorbidities. A higher proportion of
EPO patients was initiated in the hospital outpatient setting
than DARB patients (EPO 21%, DARB 6%). Treatment duration
was similar between the groups (EPO 77 days, DARB 78 days),
with a higher number of injections of erythropoietic agent in
the EPO group (EPO 8.4, DARB 6.5). Transfusion rates were
similar between groups (EPO 6.6%, DARB 5.9%). Compared
with DARB patients, EPO patients had lower erythropoietic
agent drug cost (EPO $2,057 vs. DARB $2,462, P < .0001).
CONCLUSIONS: Despite lower rates of baseline comorbidities,
similar treatment duration and fewer injections, the DARB
group incurred a 20% greater drug cost for amenia treatment
compared with EPO in this large managed care population.
www.amcp.org
Abstracts From Professional Poster Presentations at AMCP’s 2005 Educational Conference
■■ ASSESSING THE IMPACT OF GLYCEMIC CONTROL
ON HEALTH RESOURCE UTILIZATION AND COSTS
OVER A 1-YEAR PERIOD
Gricar J*, Welz JA, Menditto L, Koo Y. 264 E. Broadway, #C402,
New York, NY 10002
■■ BUDGET IMPACT MODEL FOR ERLOTINIB (TARCEVA)
IN ADVANCED NON–SMALL-CELL LUNG CANCER
Ramsey SD*, Clarke L, Kamath TV, Veenstra DL. Fred Hutchinson
Cancer Research Center, 1100 Fairview Ave. N, M2-B230, Seattle,
WA 98109
INTRODUCTION: The relationship between glycemic control and
health resource utilization and costs was evaluated in a cohort
of managed care members with diabetes over a 1-year period.
METHODS: A retrospective analysis was conducted on a nationally representative administrative claims database of managed
care enrollees aged 18 to 64 years between 2000 and 2002.
Patients with an International Classification of Diseases, Ninth
Revision, Clinical Modification (ICD-9-CM) code diagnosis of
diabetes who were continuously enrolled for at least 365 days of
follow-up were included. Glycemic control was measured by
using the average score of ≥2 glycosylated hemoglobin (A1c)
tests. Selected comorbidities were controlled for using multiple
logistic regression techniques. Resource utilization (hospitalizations, emergency department (ED) visits, office visits, pharmacy,
and laboratory tests) was assessed in the entire study cohort as
well as in subpopulations with health care encounters for
7 diabetes-related conditions (cardiovascular disease, hypoglycemia, nephropathy, neuropathy, retinopathy, embolic
stroke, and transient ischemic attack).
RESULTS: A total of 20,914 subjects (mean age 52 years) met the
inclusion criteria for analysis. Overall, a positive statistical
relationship was observed between increasing A1c level and
total direct medical cost over a 1-year period. In patients with an
A1c level of <7, the average total direct medical cost was
$1,418.69; corresponding costs in patients with A1c levels 7 to
<8.5%, 8.5% to 10%, and >10% were $1,867.53, $2,382.26,
and $2,904.33, respectively. The relationship extended to rates
of hospitalizations, ED visits, and office visits, and the number
of prescriptions for diabetes-related drugs. Similar trends were
evident in the subpopulations, with a positive statistical
relationship observed between increasing A1c level and both
total direct medical cost and hospitalization rate in all 7 subpopulations.
CONCLUSIONS: Tighter glycemic control is associated with lower
medical costs during a 1-year period. By promoting ways to
improve glycemic control in their members with diabetes, such
as optimizing drug therapy and high-risk member diabetes case
management, health insurers may benefit from averted medical
costs, even in the short term.
OBJECTIVE: To develop a tool that can be used by health care
plans to assess the budgetary impact of covering erlotinib for
treating patients, aged 18-65 years, with stage IIIB/IV
non–small-cell lung cancer (NSCLC) who have failed at least
1 prior chemotherapy regimen.
METHODS: An Excel-based decision model was developed to
assess costs and outcomes for a formulary that includes U.S.
Food and Drug Administration-approved and National
Comprehensive Cancer Network guideline-recommended
treatment options, comparing a formulary with erlotinib versus
without erlotinib. The model considers second- and third-line
treatments and includes differential outcomes that may
influence direct medical costs (grade 3/4 adverse effects of treatment such as neutropenia and anemia). The incidence of
advanced NSCLC and adverse effects related to treatment (all
agents) are based on the Surveillance, Epidemiology, and End
Results Cancer Registry and published results of clinical trials.
Drug and treatment costs are obtained from publicly available
sources. One-way and multiway uncertainty analyses were used
to evaluate the impact of varying assumptions and data values
on outcomes.
RESULTS: The base case considers a health plan of 500,000
enrollees. Assuming that erlotinib represents 22% of secondline treatments and 50% of third-line, total costs of treating
Stage IIIb/IV NSCLC patients over 1 year are $418,710 without
erlotinib and $388,226 with erlotinib (difference: $30,484;
90% CI, $8,722-$ 90,170). Erlotinib lowers costs compared
with standard chemotherapy through reductions in standard
chemotherapy-related infusion costs and costs of managing
adverse events. Sensitivity analyses demonstrate that
the budget impact is most sensitive to the incidence of NSCLC,
the proportion of patients receiving treatment, the proportion
receiving erlotinib versus standard agents, the cost of erlotinib,
and duration of treatment. Outcomes were relatively insensitive
to the incidence of serious erlotinib-related adverse events.
CONCLUSION: Adding erlotinib as a second- and third-line treatment option may reduce total health plan costs compared with
standard therapy.
www.amcp.org
Vol. 11, No. 7
September 2005
JMCP
■■ CLINICAL AND FINANCIAL IMPACT OF ERYTHROPOIETIN:
A STUDY OF PRE–END-STAGE RENAL DISEASE PATIENTS
IN A HIGH-RISK SPECIALTY-CASE COORDINATION PROGRAM
Fein IA*, Mullany LD, Wells BA, Cumbaa V, McDermott C, Fein AB,
Elkins AL. University of Florida School of Public Health at AvMed
Health Park, 4300 NW 89th Blvd., Gainesville, FL 32602
OBJECTIVE: To assess the clinical and financial impact of
erythropoietin (EPO) in the management of anemic patients
with chronic kidney disease (CKD).
METHODS: A retrospective cohort study of 4,097 predialysis
CKD patients enrolled in a specialty coordination program of
the AvMed Health Plans, a not-for-profit health maintenance
organization, from January 2000 and December 2003.
Administrative databases were used to examine the clinical
characteristics and health care costs of the group. Two hundred
seventy-one members received EPO, but after those with aplastic anemia or renal transplants were eliminated as well as those
undergoing chemotherapy, hemodialysis, or peritoneal
dialysis, the final study group had 52 patients with 52 casematched controls. Clinical and demographic characteristics of
both the case and control groups were virtually identical to
nationally reported data for CKD patients.
RESULTS: CKD patients experienced a continuous deterioration
of renal function over time (creatinine clearance fell 16.3% over
9 months, P < .0001), accompanied by a progressive 5-fold
increase in global per-member-per-month (PMPM) health care
expenditures over the 3-year period prior to the initiation of EPO
therapy [r2 =.651, P<.0001]. Following EPO therapy, renal function was effectively stabilized, measured by creatinine clearance,
and sustained over a mean follow-up time of 9 months. Mean
hemoglobin levels fell from 11.52 g/dL (± 1.56 SD) to 9.90 g/dL
prior to EPO therapy and were corrected to starting levels
(11.52 g/dl, ± 1.44 SD) with the use of EPO. The rise in PMPM
expenditures was halted and remained stable over the 9-month
mean follow-up period. Health care expenditures were comprehensive, including all drug costs.
CONCLUSIONS: Treatment with EPO in anemic patients with predialysis CKD is clinically effective and is associated with a stabilization
of both the progression of renal disease and global PMPM costs.
■■ CMS COMPETITIVE DRUG ACQUISITION PROGRAM:
SHAPING FUTURE MCO PROCESSES?
Baker JJ*, McClard C. The Resource Group, PO Box 70, Pickton, TX
75471
INTRODUCTION: This case study details the Centers for Medicare
& Medicaid Services (CMS) Competitive (Drug) Acquisition
Program (CAP) processes that become effective January 1,
2006, and explores the potential impact on managed care
organization processes.
JMCP
September 2005
Vol. 11, No. 7
METHODS: The primary objective of the case study was 2-fold:
to develop a process model of the CMS CAP for Part B drugs as
it relates to physician providers, and to explore the potential for
those government processes to shape managed care organization
(MCO) processes in the near future. Government sources and
physician interviews and observations were utilized to create
the CAP process model. MCO interviews and relevant survey
results were utilized to create a likely future scenario for the
MCO process changes.
RESULTS: The case study identified sequential processes that are
required of physician providers who will participate in the CMS
CAP. Although the CAP processes were initially designated as
primarily administrative in nature, the physician practice personnel involved professional clinical labor as well as clinical
labor in more than 50% of practices interviewed. More than 70%
of practices interviewed indicated they planned to adopt the
government-mandated process as their primary model. A
significant portion of physician practices indicated they intended
to request specific changes from MCOs in the next
contracting cycle. When the framework for potential changes,
constructed utilizing physician responses, was ranked using a
5-point scoring system, 3 particular categories emerged as most
likely to impact MCO processes in the near future.
CONCLUSIONS: CAP stipulates a series of mandatory processes
for participating physician providers. This case study’s model
demonstrates that processes utilized by the government model
have a significant potential for migrating to MCOs in the near
future. Process changes will most likely be triggered by provider
requests and/or negotiations when contract expiration dates are
pending. The model also demonstrates that the processes
utilized are capable of being successfully replicated.
■■ COMPARISON OF HEALTH CARE UTILIZATION
AND INCIDENCE OF ADVERSE EVENTS OF LEVETIRACETAM
WITH GABAPENTIN IN EPILEPTIC PATIENTS USING A
RETROSPECTIVE CLAIMS ANALYSIS
Grossman P.* UCB Pharma, 1950 Lake Park Dr., Smyrna, GA 30064
OBJECTIVE: To compare health care utilization and incidence of
adverse events (AEs) in epileptic patients initiating levetiracetam (LEV) or gabapentin (GBP).
METHODS: A retrospective cohort analysis of epileptic patients
was conducted using data from a U.S. claims database. Patients
without any LEV or GBP prescription for 6 months prior to
therapy initiation were followed for 3 months to 1 year (July
2001-December 2003). GBP patients were matched to LEV
patients by clinical characteristics, seizure, and therapy types.
Comparison of health care utilizations used Wilcoxon rank-sum
tests. Risk of AEs was assessed using Cox proportional hazards
models.
RESULTS: Treatment groups (n = 816 in each) were comparable:
www.amcp.org
mean age ~39 years, ~63% women, 64% generalized seizures,
65% adjunctive therapy. LEV patients refilled their prescriptions
more often than GBP (mean/patient/year: 8.3 vs. 6.0,
P<0.001). Utilization was significantly lower in LEV than in GBP
for physician office visits (18.1 vs. 20.8, P < 0.01), emergency
room visits (8.5 vs. 11.0, P < 0.01), other outpatient visits (35.7
vs. 40.6, P < 0.05), and medications other than antiepileptics
(24.4% vs. 36.5, P < 0.001). Diagnostic tests and inpatient
services were comparable (not significant) in both groups.
Absence of AEs during follow-up was 38% for LEV versus 29%
for GBP (P < 0.001). Risk of AEs was significantly lower in LEV
than GBP: hazard ratio: 0.76, 95% confidence interval,
0.68-0.86, P < 0.001; median time to first AE: 42 days LEV, 28
days GBP.
CONCLUSION: LEV patients showed lower utilization of most
common health care services than GBP patients. Rate and risk
of AEs were significantly lower in LEV despite a higher refill rate
relative to GBP.
■■ COST-EFFECTIVENESS AND BUDGETARY IMPACT
OF BIOLOGIC THERAPIES FOR MODERATE-TO-SEVERE
PLAQUE PSORIASIS
Broder M*, Laouri M, Ding L, Kwon P. Partnership for Health
Analytic Research, LLC, 1950 Sawtelle Blvd., Suite 280, Los Angeles,
CA 90025
OBJECTIVE: To calculate, from the payer perspective, the relative
cost-effectiveness and budget impact of efalizumab and etanercept for psoriasis in a million-member health plan.
METHODS: We used information on dosing and treatment-related
utilization from product labeling information and subsequent
clinical trials for 2 biologic drugs (efalizumab and etanercept)
indicated for moderate-to-severe plaque psoriasis to construct a
cost-effectiveness model that calculates cost per patient with
a successful outcome. Success was defined as achieving a ≥75%
improvement in the Psoriasis Area and Severity Index score
(PASI 75) after 24 weeks of treatment. Cost of treatment was
determined by adding costs in 4 categories: (1) drugs (average
wholesale price), (2) administration, (3) monitoring (platelet
counts or tuberculosis test), and (4) adverse events. For budget
impact, we assumed that a million-member plan would have
2,000 adults with moderate-to-severe plaque psoriasis and 200
would use biologic therapy. To calculate cost, we assumed 100%
of etanercept users stepped down their dose from 100 mg/week
(weeks 1-12) to 50/mg/week (weeks 13-52) as recommended in
the product labeling.
RESULTS: Cost per PASI 75 responder was $20,438 for
efalizumab and $24,351 for etanercept. Projected annual treatment costs were $3.8 million for efalizumab and $4.7 million
for etanercept, assuming all patients step down and continue at
50 mg/week for 40 weeks. Annual treatment costs increased if
www.amcp.org
patients on etanercept had lower step-down rates: $5.4 million
(75%); $6.1 million (50%); and $6.8 million (25%).
CONCLUSIONS: Efalizumab is more cost effective than etanercept. The budget impact of etanercept is unpredictable since it
is unknown how many patients will step down from 100
mg/week to 50 mg/week. Efalizumab dosing is stable and predictable. Cost-effectiveness and dosing stability may be particular advantages when efficacy and safety are comparable and the
ability to accurately predict drug costs is a key consideration.
■■ COST-EFFECTIVENESS OF LINEZOLID VERSUS
VANCOMYCIN IN THE TREATMENT OF COMPLICATED
SKIN AND SOFT-TISSUE INFECTION DUE TO PROVEN OR
SUSPECTED MRSA IN PATIENTS AGED 65 YEARS OR OLDER
McCollum M*, Sorensen S, Liu L. University of Colorado at Denver
and Health Sciences Center, 4200 East Ninth Ave., Box C-238, Denver,
CO 80262
INTRODUCTION: We estimated costs of treating complicated skin
and soft tissue infections (cSSTI) with linezolid (LZD) versus
vancomycin in a subset of patients aged > 65 years enrolled in
the largest clinical trial of cSSTI due to proven or suspected
methicillin-resistant staphylo- coccus aureus (MRSA).
METHODS: Clinical trial patients aged > 65 years and admitted to
U.S. hospitals with cSSTI due to proven or suspected MRSA
were included in these analyses. Infection-related costs were
estimated by applying nationally representative 2003 per-diem
hospital costs for days in medical/surgical, intensive care, or stepdown units. Costs of administering intravenous (IV) therapy
were applied to IV treatment duration. Medications were valued at
wholesale acquisition cost. Hospitalization costs and total costs
(hospital plus outpatient) were estimated. Cure rates were
determined from the subset of patients who were clinically
evaluable.
RESULTS: Of the 717 patients in the clinical trial, 163 were aged
> 65 years (87 LZD, 76 vancomycin). No significant clinical or
demographic differences at baseline were observed between
groups. Average hospitalization costs for patients treated with
LZD were $4,511 versus $6,478 for vancomycin (P < 0.001).
Average total costs for patients treated with LZD was $6,009
versus $7,329 for vancomycin (P = 0.03). Average length of stay
for LZD patients was 6.8 days compared with 10.3 days for
patients treated with vancomycin (P<0.001). No significant differences were observed in clinical cure rates between linezolid
and vancomycin (LZD = 89%, vancomycin = 81%, P = 0.2).
CONCLUSION: Treatment with linezolid (IV/orally) for cSSTI due
to proven or suspected MRSA in patients aged >65 years results in
lower hospital and total costs compared with vancomycin.
Clinical cure rates were similar in both groups. Lower costs for
patients treated with linezolid are enabled by transition to oral
therapy and earlier hospital discharge.
Vol. 11, No. 7
September 2005
JMCP
■■ CUMULATIVE EXPOSURE MODELS:
CYCLOOXYGENASE-2 INHIBITORS AND CARDIOVASCULAR RISK
Shaya FT*, Gu A, Weir MR, Skolasky R. University of Maryland
School of Pharmacy, 515 West Lombard St., Room 256, Baltimore, MD
21201
INTRODUCTION: Little is known about the association of length of
exposure to cyclooxygenase-2 inhibitors (COX-2s) and the risk
of cardiac events. This study determines the impact of extended exposure to COX-2s among high-risk Medicaid patients and is
based on a previous propensity-adjusted model that showed no
added risk of cardiac events in COX-2 versus nonsteroidal antiinflammatory drug (NSAID) users in this Medicaid population.
METHODS: Selecting COX-2 users alone, we analyzed all medical
and prescription claims of all continuously enrolled Medicaid
patients, with at least 1 prescription for a COX-2 between
January 1, 2000, and January 1, 2003, and no such prescriptions
in the first 6 months. We used both direct adjustment and
propensity score methods and assessed length of exposure to
COX-2s as a risk factor for postuse cardiac events, defining risk
as a categorical variable (<30, 30-59, 60-89, 90-119, and >120
days), then as a continuous variable (divided by 30). The models
are adjusted for age, gender, race, location (urban/suburban/ rural),
and clinical risk factors.
RESULTS: A total of 1,784 patients used COX-2s, 25% for fewer
than 30 days. From the categorical analysis, there are significant
increases in the likelihood of a postuse cardiac event given
increased use compared with fewer than 30 days of cumulative
exposure. From the analysis of exposure to COX-2s as a continuous variable, in the propensity-adjusted model, each 30-day
increase in exposure corresponds to a concomitant but nonsignificant 2% increase in risk of cardiac events. For the direct
adjusted model, there is a concomitant 5.5% significant
increase in postuse cardiac events.
CONCLUSION: Among Medicaid COX-2 users, the risk of cardiac
events is associated with longer exposure to COX-2s only when
exposure is categorized in 30-days increments but not when
used as a continuous variable, suggesting a nonlinear relationship between exposure and events.
■■ DEPRESSION COMPLIANCE
RETROSPECTIVE CASE ANALYSIS
Straub PA.* Precision Healthcare Delivery, 2301 River Rd., Suite 302,
Louisville, KY 40206
patients receiving drug therapy for depression during 2003.
There were no time limits for treatment on drug therapy. A total
of 150 patients who were identified were sent a letter by their
physicians and asked to participate in a discussion about their
therapy with a registered pharmacist. The patients who
responded to the letter were asked to:
a) provide demographic information such as age, gender, etc.;
b) indicate chronic disease states, number of prescriptions
taken each day;
c) provide the name of medication used for depression therapy;
d) identify a stressful event in the last 12 months;
e) rank themselves according to a self-reported medicationtaking behavior scale;
f) discuss whether they skip doses due to cost;
g) discuss whether they experience embarrassment because of
medication; and
h) discuss whether they are receiving any counseling or group
therapy.
RESULTS: There were 76 patients interviewed for this study.
Despite treatment recommendations, 77.6% of the patients
were noncompliant due to at least one reason, whether it was
side effects, cost, or forgetfulness. A patient could have had
more than one reason for being noncompliant. The results also
showed that there was no correlation between noncompliance
and the medication being taken for depression.
CONCLUSIONS: It can be concluded that, in this population,
patient education regarding the value of their medications was to
be encouraged. The results were shared with the physicians
to increase their awareness of patients not properly taking their
medications; 27.6% of patients were noncompliant due to side
effects. A discussion with their physician regarding these unwanted
side effects can determine if patients need to switch medications. The study showed that 51.3% of the patients admitted
that they either skip doses, split their pills, or asked their physician
for samples, so cost was a major issue with compliance. Patientassistance program phone numbers for assistance and other
discount card information was given to the staff to distribute to
these patients. The other reason given for noncompliance was
forgetfulness; 31.6% of patients admitted to being forgetful
about taking their medication. Morisky tear-off sheets were placed
in patient waiting areas for patients to privately score themselves for
medication non-adherence and encourage them to discuss their
results with a staff member.
OBJECTIVE: To review and document nonspecific, nonbranded
drug therapy of patients being treated for depression with issues
related to noncompliance. The study was to determine the causes
of the noncompliance and provide recommendations for better
compliance.
METHODS: A managed care database was used to identify
JMCP
September 2005
Vol. 11, No. 7
www.amcp.org
■■ DOSING DISTRIBUTION PATTERNS
AND ASSOCIATED COSTS OF ERYTHROPOIETIC AGENTS
IN PATIENTS WITH PREDIALYSIS CHRONIC KIDNEY DISEASE
FROM 3 LARGE MANAGED CARE ORGANIZATIONS
Barron JJ, Mody SH*, Boudreaux MY, Yang W, Groesbeck M,
McKenzie RS. Ortho Biotech Clinical Affairs, LLC, 430 Route 22 East,
Bridgewater, NJ 08807
INTRODUCTION: Since few reports exist describing current dosing
patterns of epoetin alfa (EPO) and darbepoetin alfa (DARB) in
patients with predialysis chronic kidney disease (pCKD), this
retrospective, observational study was performed to analyze
dosing patterns and associated costs of EPO and DARB from a
managed care practice perspective.
METHODS: Anemic pCKD patients who were aged ≥18 years,
had >2 EPO or DARB claims, and were newly initiated on
erythropoietic therapy between the fourth quarter of 2001 and
the third quarter of 2004 were identified from medical claims of
3 regionally diverse health plans covering approximately
10.2 million lives. EPO and DARB use was identified via
Healthcare Common Procedure Coding System (HCPCS) codes
in medical claims with doses calculated using billed units.
Dosing frequency, mean weekly dosing, and drug costs (using
2004 wholesale acquisition prices) were calculated for each
group.
RESULTS: A total of 325 EPO and 163 DARB patients met the
inclusion criteria. Mean age (years, EPO 69.5±13.8, DARB 69.6
± 12.7, P = not significant NS), gender distribution (EPO 45.0%
male, DARB 45.0% male, P = [NS]), and prevalence of comorbid
conditions were similar between therapeutic groups. Weekly
and extended (every [Q] 2 weeks [Ws]) dosing frequency were
utilized in patients receiving EPO (QW: 23.4%, Q2W: 34.8%,
> Q3W: 29.6%) and DARB (QW: 5.5%, Q2W: 46.0%, > Q3W:
44.5%), with an average interval between treatments of 18.6 ±
15.9 days for EPO and 20.7 ± 8.6 days for DARB patients. The
average weighted weekly dose was 8,516 units for EPO and 57
mcg for DARB, which corresponded with estimated mean weekly
costs of $100 for EPO and $241 for DARB. Similar dosing
patterns and cost differences were observed for patients
completing 4, 8, and 12 weeks of therapy.
CONCLUSIONS: Extended EPO and DARB dosing (> Q2W) was
common among anemic pCKD patients. However, costs associated
with these treatments differed between therapies, providing a
cost advantage to those patients being administered EPO.
www.amcp.org
■■ ECONOMIC BURDEN OF GOUT TO THE EMPLOYER
Kleinman NL, Patel P, Brook RA*, Melkonian AK, Smeeding JE, JosephRidge N. The JeSTARx Group, 18 Hirth Dr., Newfoundland, NJ
07435-1710
OBJECTIVE: To determine the economic burden of gout associated
with medical costs and work loss from an employer perspective.
METHODS: Medical, pharmacy, workers’ compensation (WC),
short- and long-term disability (STD, LTD), and sick leave (SL)
costs in employees with gout, as identified by an International
Classification of Diseases, Ninth Revision (ICD-9) code of 274.xx,
were examined in a database consisting of 2001 through 2004
claims, payroll, and demographic data from more than 250,000
employees from multiple large U.S.-based employers.
Regression modeling was used to measure the cost differences
between employees with gout and employees without gout
while controlling for age, job tenure, gender, salary, region, and
other factors.
RESULTS: Data were available for 1,171 employees with gout
and a control group of (247,867) employees without gout. The
gout group’s costs (per patient per year) were almost twice as
high ($6,871 higher) summed across all direct medical and
work loss measures (P < 0.0001). The individual differences in
medical and pharmacy costs were $1,401 and $427, respectively
(both P < 0.0001). Work absence costs had differences of $697
(WC, P < 0.0001), $358 (STD, P < 0.0001), -$25 (LTD,
P < 0.0001), and $307 (SL, P < 0.0001).
CONCLUSIONS: The economic impact of gout can be costly to
employers not only in terms of direct health care costs but also
from potential work loss due to absenteeism. Interventions
focused on identifying and managing the underlying cause of
gout have the potential to produce significant savings in
medical and pharmaceutical costs.
■■ ECONOMIC EVALUATION OF CONTROLLED-RELEASE
OXYCODONE (CRO [OXYCONTIN TABLETS]) VERSUS
OXYCODONE/ACETAMINOPHEN (OXY/APAP [PERCOCET])
FOR OSTEOARTHRITIS PAIN OF THE HIP OR KNEE
Marshall DA*, Strauss ME, Pericak D, Math M, Buitendyk M,
Codding C, Kim SS, Torrance GW. Innovus Research Inc., 1016-A
Sutton Ave., Burlington, Ontario L7L 6B8, Canada
INTRODUCTION: Controlled-release oxycodone (CRO) is efficacious
for persistent moderate-to-severe osteoarthritis pain, based on
well-controlled trials. Additionally, decision makers require
evidence of effectiveness in routine practice and cost-effectiveness
compared with standard therapy.
METHODS: An open-label, active-controlled, randomized,
naturalistic 4-month study of analgesic effectiveness and costeffectiveness of CRO versus oxycodone/acetaminophen
(oxy/APAP) was conducted. Outcomes and resource use were
Vol. 11, No. 7
September 2005
JMCP
collected by telephone. Effectiveness was measured in 485
patients as the proportion having at least 20% improvement
from baseline in Western Ontario and McMaster Universities
Osteoarthritis Index pain score. Quality-adjusted life-years
(QALYs) were calculated from Health Utilities Index-3 scores.
Cost-effectiveness was measured as cost/patient improved and
QALYs gained from societal and health care perspectives using
generic oxy/APAP (base case). Uncertainty was evaluated using
multiple 1-way sensitivity analyses and cost-effectiveness
acceptability curves.
RESULTS: In the study, 62.2% versus 45.9% (P = 0.0003) of
patients improved with CRO and oxy/APAP, respectively. Mean
QALYs gained over 4 months with CRO compared with
oxy/APAP was 0.0105 (P = 0.1673). Mean societal cost/patient
over 4 months was US$6,792 versus US$6,929 (P = 0.3345) for
CRO and oxy/APAP, respectively. CRO was both more effective
and less costly than oxy/APAP, using the societal perspective
(includes costs associated with time loss). Using a health care
perspective (excludes costs associated with time loss), costeffectiveness of CRO was US$4,500/patient improved and
US$69,856/QALY gained.
CONCLUSIONS: From the societal perspective, CRO was both
more effective and less costly than oxy/APAP. From the health care
perspective, CRO, compared with generic oxy/APAP, fell within
the acceptable range of cost-effectiveness if decision makers
were willing to pay between US$50,000/QALY and
US$100,000/QALY. These findings should be considered in
decisions about treating osteoarthritis pain.
■■ EFFECTIVENESS OF A DIRECT-TO-PATIENT
SPECIALTY PHARMACY COMPLIANCE PROGRAM
Allen JD*, Zabriski S, Brown J. McKesson Specialty, 4343 N.
Scottsdale Rd., Suite 150, Scottsdale, AZ 85251
INTRODUCTION: The therapy persistence for patients enrolled in
a hepatitis C direct-to-patient specialty pharmacy compliance
program was compared with a retail distribution control group
in a matched-pair cohort fashion.
METHODS: Patients new to hepatitis C therapy who enrolled in
a Specialty Pharmacy Compliance Program (n = 279) were
matched retrospectively in a 1 to 5 ratio with a control group of
patients receiving therapy through a retail distribution channel
and analyzed over a 1-year period. The primary end point was
the percentage of patient persistence, defined as the percentage
of patients in a given month of the study period who had drugs
on hand as measured by prescription fill date and days supply.
Secondary measures included compliance and length of therapy.
Patients were excluded if they had any medication fills from a
retail outlet during the study period.
RESULTS: Of the 279 patients enrolled, 227 were available for
analysis. Beginning at 2 months, and extending each month to
JMCP
September 2005
Vol. 11, No. 7
the end of the evaluation period, the specialty pharmacy cohort
had significantly more persistence than the retail comparator
group (P < 0.05). Specialty patients were also more compliant
at 6 and 9 months and utilized 37% more units than did the
control group over the same period.
CONCLUSION: A specialty pharmacy compliance program,
including relationship marketing, call center support, and
distribution of a hepatitis C product, improves patient persistence
with therapy.
■■ EFFECTIVENESS OF DEPRESSION TREATMENT
PREDICTS SUBSEQUENT HEALTH SERVICES COSTS
Simon GE, Khandker RK.* Wyeth Research, 500 Arcola Rd.,
Collegeville, PA 19426
INTRODUCTION: Numerous cross-sectional studies demonstrate a
strong association between depression and use of health services,
but few longitudinal studies have examined whether remission
of depression is associated with decreased health services costs.
METHODS: Pooled data from 7 longitudinal studies of patients
beginning depression treatment were used to examine the relationship between clinical outcomes of acute-phase treatment
and health services costs over the subsequent 6 months.
Clinical outcomes were assessed by structured telephone interviews. Health services costs were assessed using health plan
accounting records.
RESULTS: Of 1,816 patients entering treatment and meeting
criteria for major depressive episode, 29% had persistent major
depression 3 to 4 months later, 37% were improved but did not
meet criteria for remission, and 34% achieved remission of
depression. Those with persistent depression had higher baseline
depression scores and higher health services costs before beginning treatment. After adjustment for baseline differences, mean
health services costs over the 6 months following acute-phase
treatment were $2,106 (95% confidence interval [CI], $1,684$2,545) for those achieving remission, $2333 (95% CI, $1,940$2,754) for those improved but not remitted, and $2,955 (95%
CI, $2,452-$3,509) for those with persistent major depression.
Average costs for depression treatment (antidepressant prescriptions, outpatient visits, mental health inpatient care)
ranged from $431 in the remission group to $599 in the
persistent depression group.
CONCLUSIONS: Clinical outcome of acute phase depression treatment predicts subsequent health services costs, and persistence of
depression is associated with 40% higher costs compared with
full remission. The excess costs associated with persistence of
depression are nearly twice as great as spending on depression
treatment.
www.amcp.org
■■ EFFECTIVENESS OF ROSUVASTATIN COMPARED WITH
OTHER STATIN THERAPIES ON TARGET LDL-CHOLESTEROL
GOALS IN A USUAL-CARE SETTING
Kamat SA*, Wertz DA, Yang GW, Bullano MF, Willey VJ. HealthCore,
Inc., 800 Delaware Ave., Fifth Fl., Wilmington, DE 19801
OBJECTIVE: To evaluate current cholesterol management trends in
clinical practice, low-density lipoprotein cholesterol (LDL-C)
goal attainment in patients treated with rosuvastatin compared
with other statins was examined using aggressive therapeutic
options suggested in the 2004 National Cholesterol Education
Program (NCEP) Report (Grundy et al. Circulation. 2004;110:
227-29).
METHODS: Patients newly initiated on rosuvastatin, atorvastatin,
simvastatin, pravastatin, lovastatin, or fluvastatin between
August 1, 2003, and September 30, 2004, were identified from
a Southeastern health plan’s administrative claims with integrated
lab results data for this retrospective, longitudinal cohort study.
Patients were excluded if they had prior dyslipidemic therapy in
the 12-month period preceding their initial statin fill. Patients
with at least 1 preinitiation and postinitiation LDL-C level were
followed until they switched, supplemented, or discontinued
their initial statin. Administrative claims were utilized to assign
patients an NCEP risk status and corresponding LDL-C goal.
Adjusted LDL-C goal attainment odds ratios were calculated
using multivariate regression techniques after controlling for
baseline differences between groups.
RESULTS: From the identified cohort (N=3,139), patients receiving rosuvastatin were slightly younger and had lower mean doses
compared with other statins (11 mg vs. 18-69 mg). LDL-C goal
attainment was higher with rosuvastatin compared with other
statins (58% vs. 29%-48%). After adjusting for age, gender,
preindex LDL-C, NCEP risk status, and therapy duration,
significantly (P <0.05) fewer patients achieved their LDL-C
goals with atorvastatin (odds ratio [OR] = 0.66; confidence
interval [CI], 0.49-0.89), simvastatin (OR = 0.53; CI, 0.380.74), pravastatin (OR = 0.22; CI, 0.15-0.33), fluvastatin
(OR = 0.16; CI, 0.09-0.28), and lovastatin (OR = 0.32; CI, 0.220.46) compared with rosuvastatin. Furthermore, dose-stratified
analysis revealed LDL-C goal attainment was significantly lower
(P < 0.05) with 10, 20, and 40 mg atorvastatin compared with
10 mg rosuvastatin (46%-48% vs. 57%).
CONCLUSION: Rosuvastatin patients were significantly more likely to attain their LDL-C goals compared with patients on other
statins. These data are among the first to illustrate the effectiveness of rosuvastatin over other statins in a usual-care setting and
reinforce findings from randomized controlled trials.
www.amcp.org
■■ EFFECTS OF IMPLEMENTING A GENERIC MEDICATION
STEP EDIT ON THE UTILIZATION AND COSTS OF
ANTIDEPRESSANTS IN A MANAGED CARE ORGANIZATION
Dunn JD*, Cannon HE, Burgoyne DS. Intermountain Health Care
Health Plans, 4646 West Lake Park Blvd., Suite N3, Salt Lake City,
UT 84120
OBJECTIVE: To evaluate the impact on utilization and costs of
implementing a generic medication step edit in the antidepressant category in an integrated managed care organization
(MCO). This report contains short-term results and an explanation of the methodology undertaken by this MCO. Long-term
data will be presented in a follow-up abstract.
METHODS: Antidepressants do not significantly differ in their
ability to treat depression or in their incidence of adverse events
(with some interpatient variability). Initial clinical medication
choices should be made based on cost considerations, with
specific treatments being prescribed at the discretion of the treating
provider. With the availability of multiple generic antidepressant
medications, using them first will improve the cost-effectiveness
of treatment and lower the cost of treatment for patients and MCOs.
Of the selective serotonin reuptake inhibitors (SSRIs) currently on
the market, fluoxetine (Prozac), paroxetine (Paxil), and citalopram
(Celexa) are available generically. In addition to the SSRIs,
2 other agents are available generically. Bupropion SR
(Wellbutrin) is a weak inhibitor of norepinephrine and
dopamine uptake while mirtazapine (Remeron) is a serotonin,
alpha-adrenergic, and histamine antagonist. Some of the antidepressants are U.S. Food and Drug Administration-labeled for
additional indications other than treatment of depression. The
vast majority of patients will both tolerate and respond to 1 of
these 5 medications. On January 1, 2005, Intermountain Health
Care (IHC) Health Plans and the IHC Behavioral Health Clinical
Program introduced their GenericStart! Program. Under this
program, for new starts, IHC Health Plans covers brand-name
antidepressants only after a trial of a generic antidepressant
medication (excluding tricyclic antidepressants, or TCAs). New
starts are defined as members with no claims history of antidepressant treatment within the previous 6 months. Branded
antidepressants were reevaluated for formulary positioning,
with different copays being applicable after the edit was met.
IHC Health Plans generally has a 3-tier benefit. Tier-1 is for generic
medications and has the lowest copay. Tier-2 is for branded
preferred medications, while tier-3 is reserved for branded nonpreferred medications that have the highest copay. In addition,
bupropion SR, citalopram, and paroxetine were added to IHC
Health Plans GenericSample Program (fluoxetine had been
available through this program since 2003). GenericSample is
a program for IHC Health Plans members that eliminates a
copay/coinsurance for the first fill of select generic prescriptions
when filled at a participating retail pharmacy. If the member has
not filled a prescription for the requested GenericSample drug
Vol. 11, No. 7
September 2005
JMCP
in the previous 6 months, the prescribed GenericSample drug
will not require a copay/coinsurance. All subsequent refills of that
drug will require the usual generic (lowest) copay/coinsurance.
Also, a recommendation for prescribing medications for the
treatment of depression was sent to physicians that focused on
the following messages: generic antidepressants offer a dramatic
improvement in cost-effectiveness over their brand-name
equivalents; they have an excellent efficacy and safety profile
and low expense; they should be considered as the initial choice
in a patient presenting with depression; most people respond
within the first 4 to 6 weeks of treatment, but a substantial
minority will respond after 8 to 12 weeks on an anti-depressant;
and because the possibility of a suicide attempt is inherent in
major depressive disorder and may persist until significant
remission occurs, close supervision of high-risk patients should
accompany drug therapy.
RESULTS: All generic medications were made available at tier-1,
with 4 being available with a zero copay the first time they are
filled. Paxil CR (paroxetine), Wellbutrin XL (bupropion), and
Effexor XR (venlafaxine) remained at tier-2. Lexapro (escitalopram)
and Cymbalta (duloxetine) remained at tier-3, while Zoloft
(sertraline) moved from tier-2 to tier-3. Ingredient cost, the
number of prescriptions, the ingredient cost per prescription,
and the per-member-per-month (PMPM) cost for the antidepressants remained fairly consistent throughout 2004 (prior
to the implementation of the GenericStart! Program). The
ingredient cost, the number of prescriptions, the ingredient cost
per prescription, and the PMPM costs for first-quarter 2004
were $5,403,976, $67,435, $80.14, and $4.22, respectively. For
fourth-quarter 2004, the comparable numbers were $5,371,476,
$67,028, $80.14, and $4.14. During first-quarter 2005, the
first-quarter after the GenericStart! Program was implemented,
the ingredient cost, the number of prescriptions, the ingredient
cost per prescription, and the PMPM costs all decreased
substantially. The respective values were $4,853,841, $62,689,
$77.43, and $3.76. This was despite the fact that the average ingredient cost per prescription of the branded antidepressants continued to increase over the 5 reported quarters. As expected, the
market share shifted more toward the generics during firstquarter 2005 compared with 2004. During first-quarter 2005,
3,928 prescriptions for generic antidepressants were filled
through the GenericSample Program at an expense to the MCO
of $22,891. This expense is very small compared with the savings generated through the increased use of generic antidepressants. Generic citalopram became available during fourth-quarter 2004. It had no impact on overall spend or utilization
during that quarter. As the price of citalopram and the other
generic antidepressants continue to decrease (and an MCO is
able to place a maximum allowable cost) and the prices of the
branded antidepressants continue to increase, there will be
more of an impact on the category, and savings will continue to
grow. The overall estimated annualized savings for this MCO
JMCP
September 2005
Vol. 11, No. 7
due to this program exceeds $2 million.
CONCLUSIONS: Antidepressants are generally a top 5 mostutilized category for most MCOs. As more antidepressants
become available generically, MCOs can implement utilization
controls to improve formulary compliance and reduce costs to
both the MCO and its members. The implications for the medical
side should be negligible, while the long-term savings for the
MCO should be significant. As more medications become available
generically, this process can be expanded and built upon,
especially in large disease categories such as the proton pump
inhibitors or HMG-CoA reductase inhibitors (statins).
■■ EFFECTS OF IMPLEMENTING EDITS TO CORRECT DOSING
INEFFICIENCIES AMONG THE ATYPICAL ANTIPSYCHOTIC
MEDICATIONS IN A MANAGED CARE ORGANIZATION
UT 84120
INTRODUCTION: The ability of managed care organizations
(MCOs) to balance high-quality pharmaceutical care with
improved cost efficiency is becoming increasingly more challenging because of a variety of issues. Certain drug categories
are exhibiting inefficiencies regarding appropriate utilization
and dosing regimens. With the atypical antipsychotics
(aripiprazole [Abilify], olanzapine [Zyprexa], quetiapine
[Seroquel], risperidone [Risperdal], and ziprasidone [Geodon]),
there is substantial off-label use and suboptimal dosing. Dose
optimization is one method of addressing the rising costs
associated with the use of atypical antipsychotics. An example
of atypical antipsychotic dose optimization would be recommending the administration of a single 10 mg tablet in place
of two 5 mg tablets if a patient was prescribed 10 mg/day of
olanzapine. Another example would be a claims edit that would
ensure that one 4 mg tablet was dispensed if a patient was
receiving four 1 mg risperidone tablets per day. The keys to
implementing a dose-optimization program include (1) maintenance medication being available in multiple strengths,
(2) clinical evidence (pharmacokinetics, study data) supporting
once-daily administration being available, and (3) similar
average wholesale price (AWP) among the different dosage
strengths of each drug. This quality-based cost-containment
approach ensures that patients still receive the same medication
at the same daily dosage; however, the dosing regimen is
simplified, which may improve compliance. Medically necessary
exceptions (which are clinically supported) to this rule are
always allowed.
OBJECTIVE: To evaluate the impact of inefficient dosing of
atypical antipsychotics and the success of implementing a
pharmacy claims edit in an integrated MCO. This report
contains preliminary data and an explanation of the methodology
www.amcp.org
undertaken by this MCO. Long-term data and atypical antipsychotic class results will be presented in a follow-up abstract.
METHODS: In 2003, Intermountain Health Care (IHC) Health
Plans began an investigation of the potential impact of implementing a dose optimization program involving the atypical
antipsychotics. A prescribing efficiency analysis was conducted to
identify the extent to which inefficient prescribing of olanzapine
and risperidone was occurring at IHC. Based on the findings of
this analysis, a pharmacy claims edit was put in place for all the
atypical antipsychotics. This edit was designed to flag prescriptions
at the pharmacy level for the atypical anti-psychotics in which
inefficient dosing was present. Pharmacists were required to
automatically optimize the dosing regimens based on
predetermined algorithms.
RESULTS: The average age of patients prescribed olanzapine and
risperidone in this MCO was 38 years and 28 years, respectively.
There were 880 members receiving olanzapine. The most
commonly prescribed daily doses for olanzapine (representing
95% of the total sample, or 836 patients) that were incorporated
into this cost analysis were 2.5 mg, 5 mg, 7.5 mg, 10 mg,
15 mg, and 20 mg. Of the 46 members using 7.5 mg/day, only
35 (76%) were receiving one 7.5 mg tablet. Eight members (17%)
were using three 2.5 mg tablets and three members (7%) were
using one and a half 5 mg tablets. Of the 203 members using
olanzapine 10 mg/day, 147 (73%) were receiving one 10 mg
tablet, 47 (23%) were using two 5 mg tablets, 8 (4%) were
using four 2.5 mg tablets, and 1 (≤1%) was using one 7.5 mg
tablet in combination with one 2.5 mg tablet. The average number of days on medication per year was estimated at 230 for
olanzapine-treated patients. The average cost per patient per
day for olanzapine was $9.10, based on AWP. If the dosing was
optimized to the appropriate strength of 1 tablet once daily, the
average cost per patient per day would be $8.32, resulting in a
savings of $0.78 per patient per day (or 8.5%). Multiplying this
estimated savings per patient per day by the number of days on
medication per year by an estimated 70% dose conversion
would result in an estimated $112,000 annual savings, based
on AWP. In comparison, there were 473 members receiving
risperidone. The most commonly prescribed daily doses for
risperidone (representing 80.5% of the total sample, or 381
patients) that were incorporated into this cost analysis were
1 mg, 2 mg, 3 mg, 4 mg, and 6 mg. Of the 144 members using
risperidone 1 mg/day, 108 (75%) were receiving one 1 mg
tablet, 32 (22%) were using two 0.5 mg tablets, and 4 (3%)
were using four 0.25 mg tablets. Of the 18 members using
risperidone 4 mg/day, only 6 (33%) were using one 4 mg tablet,
while 9 (50%) were using two 2 mg tablets and 3 (17%) were
using four 1 mg tablets. The average days on medication per
year for risperidone was estimated at 187. Multiplying a savings
of $1.11 per patient per day if the dosing were optimized to the
appropriate strength of 1 tablet once daily ($5.71 minus $4.61)
by the number of days on medication per year by an estimated
www.amcp.org
70% dose conversion would result in an estimated $55,000
annual savings, based on AWP. By simply optimizing the dosing
of olanzapine and risperidone, this MCO estimated a potential
savings of more than $167,000. Therefore, effective July 1,
2004, IHC Health Plans implemented a dose optimization program involving the atypical antipsychotics. The data do not
include diagnosis information, prior or concomitant medication
use, net costs (rebates), or prescriber (physician) information.
CONCLUSIONS: Atypical antipsychotics are becoming an increasingly
difficult category of medication to manage because of substantial off-label use and suboptimal dosing. Programs to optimize
dosing can have a positive impact on the cost of these medications and may improve compliance. One-year data from this MCO’s
dose optimization program will be available in the fall of 2005.
■■ EFFECTS OF MANAGING HIGH-QUANTITY LONG-ACTING
OPIOID UTILIZATION IN A MANAGED CARE ORGANIZATION
UT 84120
OBJECTIVE: To evaluate the impact of implementing quantity
limits, requiring medication management agreements (MMA), and
case managing members receiving high quantities of long-acting
opioids in an integrated managed care organization (MCO).
METHODS: Effective May 1, 2003, Intermountain Health Care
(IHC) Health Plans limited the amount of OxyContin a member
may receive from 4 tablets/day up to a maximum of 160 mg/day.
Exceptions to this limit are evaluated on a case-by-case basis
and require each member using more than 160 mg/day to sign
an MMA with IHC Health Plans. It requires the member to commit to using 1 physician and 1 pharmacy for his or her pain medications. The purpose of this agreement is to protect the health
care providers’ interests while ensuring that members are receiving
the most appropriate treatment possible. Also, IHC Pharmacy
Services works closely with an IHC case manager, who is available to assist where necessary. This same protocol has been
extended to other long-acting opioids (i.e., Duragesic, Kadian,
Avinza) and people using multiple physicians and pharmacies
for short-acting opioids. Prior to the changes being implemented,
letters were sent to members, and letters and clinical information
including specific lists of patients on high quantities of long-acting
opioids or patients using other additional physicians to obtain
pain medications or using multiple pharmacies were sent to
health care professionals. In addition, pharmacists from the
MCO educated health care professionals via clinical in-office
presentations.
RESULTS: For the 2 quarters prior to the changes, the number of
OxyContin prescriptions totaled 2,417 and 2,495. By thirdquarter 2004, this number had been decreased to 2,072. The
daily average consumption for OxyContin 10 mg was reduced
Vol. 11, No. 7
September 2005
JMCP
from a high of 3.16 prior to the change to a low of 2.78 after the
change. The changes for OxyContin 20 mg, 40 mg, and 80 mg
were 2.92 to 2.65, 3.12 to 2.91, and 5.37 to 3.08, respectively.
Total ingredient cost for all quarters beginning with first quarter
2003 were $542,780, $540,862, $586,543, $599,669,
$558,895, $539,674, and $567,567. This relative flat trend in
ingredient cost was despite an average 15% price increase in
OxyContin over the same time period. The number of members
filling a prescription for OxyContin during first quarter 2003
was 964l in third quarter 2004, 863 members filled a prescription
for OxyContin. Some members switched to other newer longacting opioids such as Avinza or Kadian. Overall, the number of
members using a long-acting opioid did not change. By third
quarter 2004, more than 400 members had signed an MMA.
Pharmacy Services and Case Management were actively coordinating care for nearly half of the members with an MMA. This
coordinated effort ensures that costs are simply not shifted to
facilities such as emergency rooms or to other agents.
CONCLUSIONS: The requirement of an MMA, in conjunction with
quantity limits and case management, improves the appropriate
use of potentially abused and very expensive opioids. Quantity
limits aid in the identification of potential opioid misuse, and
utilization numbers become more consistent with recommended
guidelines. Members are not denied medication. They are only
required to more appropriately use the health care system.
Theoretically, overall care is improved by using 1 physician and
1 pharmacy. Overall costs are also decreased.
■■ EMPLOYED RHEUMATOID ARTHRITIS
PATIENTS EXPERIENCED DECREASED WORK LOSS
ON ETANERCEPT TREATMENT
Paulus HE, Weaver AL, Yu EB, Xia HA, Chiou CF*, Louie J. Amgen,
Inc., 1 Amgen Center Dr., MS 28-3-B, Thousand Oaks, CA 91320
OBJECTIVE: To assess the effects of etanercept on work loss in
rheumatoid arthritis (RA) patients.
METHODS: Patients aged ≥18 years with active RA who started
on or switched to etanercept could enroll in RADIUS 2, a prospective, multicenter, observational study. Data on employment
status and the number of times patients missed work for half a
day or more in the past month were collected at every visit.
Only patients who were on etanercept monotherapy (ETA) or
etanercept combination therapy with methotrexate (ETA+MTX)
were included in this analysis. Missing data were imputed using
last observation carried forward. For patients who reported that
they were “full-time,” “part-time,” “temporarily employed,” or
“other” at baseline, the number of missed workdays was analyzed comparing baseline with month 6 and month 12 using
the Wilcoxon Signed Rank Test. This analysis was based on data
available through November 5, 2004.
RESULTS: At baseline, there were 1,146 patients on ETA and
JMCP
September 2005
Vol. 11, No. 7
1,659 patients on ETA+MTX. At baseline, 41%, 11%, 1%, and
4% of ETA patients and 45%, 10%, 1%, and 4% of ETA+MTX
patients reported that they were “full-time,” “part-time,”
“temporarily employed,” or “other,” respectively. The number of
times that patients had missed work for half a day or more in
the past month because of their RA declined significantly from
baseline (mean for ETA = 1.74, mean for ETA+MTX = 1.43) to
month 6 (mean for ETA = 0.75, mean for ETA+MTX = 0.67)
and to month 12 (mean for ETA = 0.91, mean for ETA+MTX =
0.71) (P <0.0001 at both time points for both groups).
CONCLUSION: RA patients on etanercept, either with or without
methotrexate, reported significantly fewer missed days of work
after 6 or 12 months of treatment. These improvements may be
translated into savings in indirect costs to society.
■■ ESZOPICLONE COADMINISTERED WITH FLUOXETINE
FOR INSOMNIA ASSOCIATED WITH MAJOR DEPRESSIVE
DISORDER: EFFECTS ON SLEEP AND DEPRESSION
Skolly S*, Fava M, Buysse D, Rubens R, Wessel T, Caron J, Amato D,
Roth T. Sepracor, Inc., 84 Waterford Dr., Marlborough, MA 01752
INTRODUCTION: Insomnia and major depressive disorder (MDD)
coexist; currently, no treatment standards exist that address
hypnotic administration during antidepressant therapy. This
study evaluated the efficacy of eszopiclone on insomnia in
patients administered fluoxetine for MDD.
METHODS: Patients who met Diagnostic and Statistical Manual of
Mental Disorders—Fourth Edition (DSM-IV) criteria for both MDD
and insomnia received 10 weeks of fluoxetine QAM (every
morning) and were randomized to nightly eszopiclone 3 mg (n=
270) or placebo (n=275) for 8 weeks. Subjective sleep and daytime
function were assessed weekly by telephone. Depression was
assessed with the Hamilton Depression Rating Scale (HAMD17)
every 4 weeks and with the Clinical Global Impression
Improvement (CGI-I) and Severity scales (CGI-S). Depression
response is ≥50% decrease from baseline HAMD17, and
remission is HAMD17 ≤7.
RESULTS: Completion rates were similar. Compared with placebo, eszopiclone was associated with significantly decreased
WASO (wake time after sleep onset) and increased TST (total
sleep time) at each treatment week (P <0.05), significantly
decreased latency at all time points (P <0.05) except week 8,
higher ratings across the treatment period in sleep quality and
depth (P <0.005), and higher ratings of daytime alertness, ability
to concentrate, and physical well-being (P <0.05). Patients
coadministered eszopiclone experienced significant reductions
in HAMD17 total scores compared with placebo coadministration
at Week 4 (P = 0.01) and Week 8 (P = 0.002). After removing
HAMD17 insomnia items, differences remained significant at
Week 8 (P <0.05). At Week 8, significantly more eszopiclone
patients were responders (P <0.0011) and remitters (P <0.03).
www.amcp.org
CGI-I and CGI-S scores were significantly improved with
eszopiclone coadministration at all time points beyond week 1
(P <0.05). Treatment was well tolerated, with similar adverse
event and dropout rates. Unpleasant taste was more common
with eszopiclone.
CONCLUSIONS: In this study, patients coadministered eszopiclone and fluoxetine experienced statistically significant
improvements in several insomnia and depression measures.
Eszopiclone was well tolerated and associated with significantly
improved sleep and daytime function in patients with insomnia
and coexisting MDD.
■■ EVALUATION OF COMORBIDITIES AND COSTS
IN PATIENTS WITH BENIGN PROSTATIC HYPERPLASIA
AND CARDIOVASCULAR DISEASE
Curtice TG*, Shah MB, Bramley TJ. Boehringer Ingelheim Pharmaceuticals, 900 Ridgebury Rd., Ridgefield, CT 06877
INTRODUCTION: To evaluate common comorbidities and annual
costs in patients with benign prostatic hyperplasia (BPH) also
having cardiovascular disease (CVD).
METHODS: A retrospective case-control analysis was conducted
in a large, national, managed care database to assess comorbidities and costs for patients with BPH and CVD. The study population (cases) consisted of males, aged 35 years or older, having evidence of CVD within 6 months prior to the index BPH
date (first identified BPH diagnosis) and having 18 months of
continuous eligibility between January 1997 and June 2004.
A control cohort was matched by age and index CVD date (first
indication of CVD) at a maximum of 4:1 versus cases.
Comorbidities, total costs, and CVD-related costs (adjusted to
2004 dollars) were evaluated over the 12-month period following the index BPH date (using the matched case’s index BPH
date for the controls). Log-transformed costs were modeled using
regression analyses, and differences in comorbidities between
cohorts were assessed using chi-square tests.
RESULTS: 82,828 patients having BPH with CVD met inclusion
criteria and were matched with 271,440 controls.
Hyperlipidemia (60% vs. 47%), diabetes (23% vs. 20%), and
erectile dysfunction (10% vs. 4%) were more common with
cases (patients having BPH and CVD) than controls (all
P <0.0001). Average annual total costs were $7,810 for cases
and $5,335 for controls. Average annual CVD costs were $2,204
for cases and $1,613 for controls. Controlling for potential
confounders, average costs in the case cohort were 45% higher
(P <0.001) for total costs and 46% higher for CVD-related costs
compared with the control cohort (P <0.0001).
CONCLUSIONS: Patients having BPH with CVD are more
frequently diagnosed with comorbidities and have higher costs
when compared with a CVD-matched cohort. Although
patients with CVD commonly have increased health care costs
www.amcp.org
associated with their disease, this study highlights additional
incremental costs among patients with BPH and CVD.
■■ EVALUATION OF MEDICATION EFFICACY
IN A MANAGED CARE ORGANIZATION WHEN USING
3 ORAL ANTIDIABETIC MEDICATIONS VERSUS METFORMIN
PLUS INSULIN IN THE TREATMENT OF TYPE 2 DIABETES
Mitchell M*, Hanson D, Cannon EH, Burgoyne DS, Dunn JD.
Intermountain Health Care Health Plans, 4646 West Lake Park Blvd.,
Suite N3, Salt Lake City, UT 84120
INTRODUCTION/BACKGROUND: Controlling hyperglycemia
reduces the risk of microvascular disease and alleviates the
classic symptoms of diabetes mellitus including, but not limited
to, polydipsia, polyphagia, and polyuria. As part of an integrated
health system, health care providers work together to ensure
that cost-effective treatment is part of a care process model for
diabetic members. Direct member data need to be evaluated
often in order to guarantee that the best service possible is being
provided to health plan members.
OBJECTIVE: The primary objective of this study was to evaluate
glycemic control in type 2 diabetic members on 3 oral antidiabetic medications versus members using insulin plus oral
metformin therapy.
METHODOLOGY: The health system’s electronic medical record
system was used to identify a subset of type 2 diabetic members
starting January 1, 2004, through December 31, 2004. These
members were then assigned to 2 groups according to their
medication regimen. The first group consisted of all members
taking 3 different oral anti-diabetic medications. Members of this
group needed to have at least 6 prescriptions adjudicated
through the system throughout the calendar year. A combination
medication (e.g., glyburide/metformin 2.5/500) counted as
2 different medications for the data pull. The second group
consisted of all members using any class of insulin (long-acting,
short-acting, or combination) therapy plus oral metformin.
Members in this group required at least 6 prescriptions for
metformin and at least 6 prescriptions for insulin.
In order to properly compare these 2 groups, adherence was
also reviewed in terms of “Total Length of Therapy,”
“Medication Possession Ratio,” and “Persistence.” The primary
objective was a comparison of the efficacy of treatment determined by A1c values between the 2 groups. A1c values were
retrieved from the electronic medical record system and were
reviewed and evaluated by 3 pharmacists and 1 computer data
analyst. Results were tested for normality using the AndersonDarling Normality test and found not to be normalized.
The Mann-Whitney test was used to determine statistical difference between the groups.
RESULTS: Glucose control: The oral group (members taking 3
oral medications) contained 191 members. The insulin/
Vol. 11, No. 7
September 2005
JMCP
metformin group contained 138 members. The oral group
showed better glucose control (median A1c = 7.37) compared
with the metformin/insulin group (median A1c = 8.02). This
difference was statistically different (P = 0.0004).
Adherence and demographics of medication use: The total
length of therapy is the time from the first fill to the last fill,
including the days supply from the last fill. The percentage of
time the patient actually possesses the medication during the
course of therapy is the medication possession ratio. It can be
reported by the length of therapy of the patient or by a specific
time period, such as 90, 180, or 365 days. The possession ratio
reported here is the length of therapy (LOT), 180 days, and 365
days. (Remember that members were required to have at least
6 months of therapy.) Insulin is not reported in the insulin/
metformin group with the medication possession ratio or the
persistence measure because the correct insulin length of therapy
is often hard to determine and/or is misrepresented when
adjudicated at the point of sale. “Persistence” over time is one
of the best measures of adherence; it measures how many
patients continue to have their medications filled. Even if there
is a gap in therapy, a patient is considered persistent until his
last fill of medication.
CONCLUSION: In this retrospective analysis, hyperglycemia was
better controlled, according to A1c values, in members taking
3 different oral medications compared with patients taking metformin and insulin together. This is despite similar adherence
measures. The next step in this evaluation of these treatment
regimens is to consider the cost-effectiveness of each group.
■■ EXTENT OF BLOOD PRESSURE CONTROL AMONG
DIABETIC PATIENTS WITHIN THE MANAGED CARE SETTING
Doyle JJ*, Frech FH, Marshall D. Novartis Pharmaceuticals
Corporation, 59 Route 10, East Hanover, NJ 07936
OBJECTIVE: Diabetes is associated with a 2- to 10-fold increased
risk of coronary events. Approximately 75% of cardiovascular
disease in diabetic patients may be attributed to hypertension.
The objective of this study is to determine and compare blood
pressure (BP) control in patients with diabetes among 4 health
plan populations.
METHODS: A retrospective chart review of a random sample of
diabetic patients (identified using Health Plan Employer Data
and Information Set [HEDIS] specifications for the
Comprehensive Diabetes Care measure) was conducted within
4 commercial health plans during 2004. The level of BP control
was stratified by 4 criteria according to the Seventh Report of
the Joint National Committee on Prevention, Detection,
Evaluation, and Treatment of High Blood Pressure: (1) diastolic
(DBP) level >80 mm Hg, (2) systolic (SBP) level >130 mm Hg,
(3) SBP >130 or DBP >80 mm Hg, and (4) BP <130/80 mm Hg,
defined as controlled.
JMCP
September 2005
Vol. 11, No. 7
RESULTS: Of the 1,260 usable patient records, 16.9% of the
patients were aged 18-44 years, 73.6% were aged 45-64 years,
and 9.5% were aged >65 years; 50.3% were female. More than
50% of diabetic patients in all plans had an SBP that met or
exceeded 130 mm Hg (range 50.6%-57.7%), and more than
50% of diabetic patients in 3 of 4 plans had a DBP level that met
or exceeded 80 mm Hg (range 34.8%-56.6%). In 2 of the 4
plans, fewer than one third of diabetic patients met the recommended BP target of <130/80 mm Hg (range 28.0%-42.5%).
CONCLUSIONS: Diabetic patients are frequently not treated to
their BP goal, putting them at higher risk for cardiovascular
complications. More aggressive treatment by physicians and
education by health plans are needed to achieve BP goals for the
diabetic population.
■■ FINANCIAL IMPACT OF A PROVIDER
AND PHARMACY OUTREACH PROGRAM DESIGNED
TO CONVERT MEMBERS FROM PRESCRIPTION TO
OVER-THE-COUNTER PROTON PUMP INHIBITORS
Devlin KA*, Seneviratne V, Milevich GE, Belazi DT, Tegenu M.
AmeriHealth Mercy/PerformRx, 200 Stevens Dr., Philadelphia, PA
19113
INTRODUCTION: An outreach program was initiated to encourage
the use of therapeutically equivalent, cost-effective over-thecounter (OTC) drugs.
METHODS: Proton pump inhibitors (PPIs) are one of the most
heavily promoted and prescribed therapeutic categories of medications in the United States. The introduction of Prilosec OTC
in September of 2003 presented health plans with yet another
opportunity to utilize therapeutically equivalent OTC products
as a cost-effective alternative to available prescription medications.
A “therapeutic alternative” outreach program was initiated in a
Medicaid health plan where state regulations did not allow for
the use of copays or tiers as incentives for formulary compliance
among members. Medical providers were given the names of
their patients who were currently being treated with prescription
PPIs and asked to consider Prilosec OTC as a therapeutic
option. The health plan’s pharmacy network was also notified
of the program and requested to reevaluate inventory to help
ensure product availability and to remind providers of the
availability of Prilosec OTC as a formulary alternative when
discussing patient therapy options. The PPI prior authorization
requirements were lifted for Prilosec OTC during the first
6 months of therapy but remained in place for the prescription
PPIs. Providers were given updated PPI utilization reports on
their patients on a quarterly basis throughout 2004.
RESULTS: Prior to implementing this program, the average cost
for a one-month supply of a PPI was approximately $120.
As the program progressed, that average cost decreased to
approximately $52. Annual savings attributed to the program
www.amcp.org
amounted to more than $3.6 million.
CONCLUSIONS: Despite product availability issues during 2004,
a provider and pharmacy outreach program to encourage the
use of Prilosec OTC was extremely successful. The promotion of
OTC products as cost-effective alternatives to prescription products
should be considered by health plans when identifying costcontainment solutions.
■■ FINANCIAL IMPACT OF A PROVIDER EDUCATION
PROGRAM ON THE APPROPRIATE USE AND DOSING
OF ATYPICAL ANTIPSYCHOTICS
Seneviratne V*, Devlin KA, Milevich GE, Belazi DT, Tegenu M.
AmeriHealth Mercy/PerformRx, 200 Stevens Dr., Philadelphia, PA
19113
INTRODUCTION: The impact, from a payer’s perspective, of a
provider education program on the appropriate use and dosing
of atypical antipsychotics in a Medicaid population is evaluated.
METHODS: Atypical antipsychotics tend to be one of the most
expensive classes of medications used by predominately
Medicaid populations, if not the most expensive. Compounding
resource issues faced by such health plans is the reality that
atypicals are often not used in accordance with U.S. Food and
Drug Administration labeling and recognized treatment guidelines, such as those created by the Texas Medication Algorithm
Project (TMAP). Unjustified combination therapy with 2 or
more atypical antipsychotics has become a common practice
even though guidelines do not recommend this type of combination
therapy until all other therapeutic options have been exhausted.
A provider education program was developed to encourage
prescribing patterns that are in accordance with the TMAP
guidelines. Providers were faxed an illustrative algorithm of the
guidelines and a report of their patients receiving duplicate
therapy with atypical antipsychotics. They were asked to evaluate the patients’ current regimens and determine if there were
opportunities for changes that would improve the patients’ care.
A clinical pharmacist was available via telephone should the
provider have any questions or would like to discuss an action
plan for making changes to a patient’s regimen. Throughout
2004, these reports were sent every quarter, and results were
evaluated routinely.
RESULTS: Prior to initiating the program, approximately 925
members were receiving combination therapy with 2 or more
atypicals each month. By the end of the year, that number
decreased to 634, and the total cost savings attributed to the
program was approximately $1.1 million.
CONCLUSION: A provider education program on the appropriate
use and dosing of atypical antipsychotics was successful in
improving practice patterns and should be considered by health
plans that experience similar prescribing patterns in their
population.
www.amcp.org
■■ FINANCIAL SAVINGS OF A GENERIC DRUG SAMPLING
PROGRAM IN A MANAGED HEALTH CARE SETTING
Young SC*, Pervanas HC. Anthem Prescription Management, Anthem
BC/BS, PO Box 7101, IN25A-549, Indianapolis, IN 46204
INTRODUCTION: A program was designed and implemented by a
pharmacy benefits management (PBM) company to counter
increasing drug costs and encourage the use of generic medications by its members.
METHODS: The program was available to health plan members
with eligible pharmacy benefits in the participating states. PBM
clinical pharmacists provided physician offices with sample
request forms listing the 10 generic medications offered in the
program. Physicians provided these forms to patients who had
medical conditions that could be treated by one of the listed
generic medications. The patient presented the request form
along with a prescription to his or her pharmacy to receive
a complimentary supply (up to $10) of the generic medication.
RESULTS: The retrospective analysis of pharmacy claims data
from January 1, 2004, through December 31, 2004, was
conducted to approximate the total savings associated with the
program. A total savings of $1,101,675.61 was attributed to this
program for 2004, with 18,866 participants. The average ingredient cost savings per prescription was $70.81 for the health plan
and $26.86 in copayment dollars for the member.
CONCLUSION: Development and implementation of this generic
drug sampling program provided a significant cost savings to
the health plan and saved copay dollars for the members who
continued to take these medications.
■■ HEALTH CARE COSTS OF PATIENTS
WITH PERSISTENT ASTHMA
Colice G, Crivera C*, Varghese S, Faruqi R, Wu EQ, Birnbaum H, Daher
M, Marynchenko MB. sanofi-aventis, 200 Crossing Blvd., M/S BX2-6600D, Bridgewater, NJ 08807-0890
INTRODUCTION: Health care costs of persistent asthma patients were
compared with health care costs of individuals without asthma and
across asthma severity levels within asthma patients.
METHODS: A persistent asthma patient sample (<65 years) was
selected from an administrative database (1999-2003). Patients
were included in the sample if they met the following criteria:
(1) had 1 asthma diagnosis (International Classification of
Diseases, Ninth Revision [ICD-9]: 493.xx), (2) had no diagnosis of
chronic obstructive pulmonary disease (COPD), (3) satisfied the
2005 Health Plan Employer Data and Information Set (HEDIS)
criteria of persistent asthma between July 2002 and June 2003.
The asthma sample was subsequently divided by asthma severity
according to Leidy’s Reliever and Oral Steroid Method, and
recommended inhaled corticosteroids (ICS) dosage by asthma
severity in the 2004 Global Initiative for Asthma guidelines for
Vol. 11, No. 7
September 2005
JMCP
asthma management. A matched sample of individuals without
asthma (2 controls:1 patient) was randomly selected based on
patient demographic characteristics. Annual health care costs,
measured as payments to medical providers and pharmacies,
were compared between asthma and nonasthma control samples,
within the overall asthma sample by severity level, and limited
to patients receiving long-term ICS treatment. Descriptive statistics
were evaluated for statistical significance using paired t tests.
Costs were adjusted to 2004 dollars using the medical
Consumer Price Index.
RESULTS: An average asthma patient uses $6,452 in health care
costs annually, whereas the cost for an average control individual with matching characteristics is $2,040 (difference: $4,412,
P <0.01). Severe asthma patients, compared with moderate and
mild patients, had higher total direct health care costs ($7,933
vs. $6,314 and $4,840, respectively; P <0.01). Health care cost
of persistent asthma patients receiving long-term ICS treatment
(severe: $7,635, moderate: $5,594, mild: $3,679) were lower
compared with the average asthma patients of the same severity. The cost reduction ($1,161) was greatest in mildpersistent patients.
CONCLUSIONS: Persistent asthma is an expensive, chronic condition across all severity levels, with health care costs increasing
as severity worsens. Long-term ICS treatment can reduce health
care costs of asthma, especially in the mild-persistent patients.
■■ HEALTH OUTCOMES OF MS PATIENTS
MANAGED BY A SPECIALTY PHARMACY
Duong LP*, Cantwell AE, Pennybacker BJ, Cocolicchio CJ, Lewis CA.
CuraScript Pharmacy Inc., 6272 Lee Vista Blvd., Orlando, FL 32822
INTRODUCTION: Specialty pharmacies have well-established
programs for providing multiple sclerosis (MS) medications and
monitoring patients’ therapy adherence. Our specialty pharmacy
expanded its program to include the assessment and evaluation
of health outcomes.
METHODS: A prospective, open-label study was implemented
with an Institutional Review Board-approved protocol in
February 2003. The objective was to compare health outcomes
between patients beginning therapy for MS with interferons
(IFNs) and those beginning therapy for MS with the noninterferon,
glatiramer acetate (GA). Consecutive MS patients were invited to
participate in a telephone survey assessing general health status,
depression (Center for Epidemiologic Studies Depression
Scale), and fatigue (Fatigue Severity Scale and Daily-Fatigue
Impact Scale). Verbal consent was provided for a baseline
survey prior to beginning therapy; a mailed written informed
consent form had to be returned prior to initiation of follow-up
surveys at months 1, 3, 6, and 12. Included in this mailing was
the Guy’s Neurological Disability Scale, which assessed the
patients’ disabilities in 12 domains commonly affected by MS.
JMCP
September 2005
Vol. 11, No. 7
Other variables thought to potentially impact the outcome
measures were collected: concomitant medications, compliance
with MS therapies, relapses, adverse events, and demographic characteristics.
RESULTS: A total of 103 patients completed 12 months of
follow-up as of February 2005; 56 began therapy with IFN and
47 began with GA. Mean age was 45.9 years (range: 28 to 65
years); 88% were female and 85% were white. There were no
statistically significant differences between the treated groups in
disability, depression, or fatigue at baseline. However, there
were significant or near-significant declines in fatigue severity,
the impact of fatigue, and depression over the course of the
1-year study period.
CONCLUSIONS: Assessing health outcomes as part of a specialty
pharmacy program allowed us to advise patients and their
physicians about problematic symptoms. In addition, we
tracked symptom change, showing reduced fatigue and depression over the 12-month period.
■■ THE HIDDEN VALUE OF PRIOR AUTHORIZATION (PA):
REIMBURSEMENT OUTCOMES AND ESTIMATED SAVINGS
FOLLOWING A POINT-OF-SALE REJECTION FOR VARIOUS
PA DRUGS
Cowan C*, Ma J, Semelman S, Grootendorst P. ESI Canada, 5770
Hurontario St., 10th Fl., Mississauga, Ontario L5R 3G5, Canada
INTRODUCTION: Although prior authorization (PA) is a popular
drug- cost-containment tool, the approval rate of PA requests
has been reported to be as high as 90%. Little is known, however, about the rate of successful PA application following a
point-of-sale rejection and its attendant effect on drug use and
costs. This study analyzed claims data for 8 drugs commonly
requiring PA on Canadian employer-sponsored drug plans to
determine the impact of PA on patient reimbursement outcomes
and savings to the plan.
METHODS: Study drugs were grouped by therapy class:
osteoarthritis (celecoxib, rofecoxib), erectile dysfunction (sildenafil, tadalafil), rheumatoid arthritis (etanercept, infliximab),
and obesity (orlistat, sibutramine). All plans administered by
ESI Canada, a large pharmacy benefit manager, with at least one
of the study drugs managed on a PA program were included.
Retrospective claims data for all patients (n = 4,510) with a
point-of-sale PA rejection from October 2003 through
September 2004 were examined. Patients reimbursed for the PA
drug after the point-of-sale rejection were compared with those
with no subsequent paid claims. The estimated decrease in drug
costs was based on utilization patterns of patients reimbursed
over the study period.
RESULTS: Overall, 27.5% of patients (1,240) with a point-of-sale
PA rejection pursued reimbursement and were eventually
approved. Approval rates varied by therapy class: 16.5% (erecwww.amcp.org
tile dysfunction), 20.7% (osteoarthritis), 30.4% (obesity), and
72.5% (rheumatoid arthritis). The estimated decrease in drug
costs within each therapy class was calculated to be 48.6%
(erectile dysfunction), 44.3% (osteoarthritis), 44.5% (obesity),
and 9.5% (rheumatoid arthritis).
CONCLUSIONS: The value of PA programs in managing drug
utilization should be measured from the point-of-sale rejection.
The rate at which patients seek reimbursement following a PA
rejection varies by therapy class and has been found to be relatively low compared with the reported approval rate. This translates into savings for employer plans.
■■ IMPACT OF A CARDIOVASCULAR RISK
REDUCTION PROGRAM ON LIPID GOAL ATTAINMENT
Burns N*, Carter T, Hall K, Evans TS. Southeastern Indiana Health
Organization, 417 Washington St., Columbus, IN 47202
INTRODUCTION: A cardiovascular risk-reduction program was
evaluated by comparing baseline Health Plan Employer Data
and Information Set (HEDIS) cholesterol measures with comprehensive follow-up chart review.
METHODS: HEDIS 2004 low-density lipoprotein cholesterol
(LDL-C) control was determined. The plan’s medical management
nurses employed provider- and patient-directed interventions to
address cardiovascular risk reduction and to increase awareness
of the importance of cardiovascular medication adherence.
Interventions included cardiovascular health screenings at
employer health fairs, monthly patient education programs by
mail, employer educational lunches, pharmacist-patient communication programs, and communication program for the
plan’s staff and nurses. A follow-up chart review
determined member cardiovascular risk factors and National
Cholesterol Education Program (NCEP) goal attainment.
RESULTS: Baseline HEDIS measurements were 63.3% for LDL-C
<130 mg/dL and 36.7% for LDL-C <100 mg/dL. Chart review was
conducted on 356 members with a diagnosis of hyperlipidemia.
LDL-C <100 mg/dL was documented in 51.0% of members
with coronary heart disease or CHD risk equivalents, and LDLC <130 mg/dL was documented in 73.2% of members with
2 or more cardiovascular risk factors. Overall NCEP goal attainment was achieved in 66.6% of members evaluated.
CONCLUSIONS: Provider- and patient-directed cardiovascular
risk reduction interventions had a positive impact on member
lipid goal attainment and may positively impact future HEDIS
measures. Continued use of this multifaceted approach may
result in overall improved cardiovascular health for the plan’s
member population.
www.amcp.org
■■ IMPACT OF A TARGETED DISEASE INTERVENTION
ON HMG-COA REDUCTASE INHIBITOR UTILIZATION
IN PATIENTS WITH DIABETES
Tjioe D*, Chang E, Nguyen CD, Stroup JR. Prescription Solutions,
3515 Harbor Blvd., Costa Mesa, CA 92626
INTRODUCTION: In May 2004, Prescription Solutions implemented the HMG-CoA Reductase Inhibitor (Statin) Use in Diabetes
Targeted Disease Intervention (TDI) to promote the proper use
of these agents in patients with diabetes.
METHODS: This program was an interactive provider-based
pharmacy intervention designed to increase awareness of
diabetic dyslipidemia treatment recommendations from the
National Cholesterol Education Program Adult Treatment Panel
III and the American Diabetes Association. This program
involved analysis of medical claims data to identify members
who were diabetic or had a diabetes-related diagnosis. Providerspecific reports were generated listing patients who might
benefit from lipid-lowering therapy, and on May 4, 2004, these
reports were sent to physicians in the intervention cohort. The
statin use of patients under the care of intervention and control
physicians was then compared. It was hypothesized that members
who were included in the intervention mailing would be more
likely to initiate statin therapy after the intervention compared
with those in the control group.
RESULTS: The Statin Use in Diabetes TDI was associated with a
nearly 2-fold increase in the percentage of members who initiated
statin therapy during the postintervention period. The statin initiation rate was significantly higher (one-sided test with P value
= 0.0497) in the intervention cohort compared with the control
cohort (13.2% vs. 7.7%), with an adjusted odds ratio of 1.8.
CONCLUSIONS: Patients whose physicians received TDI components were more likely to initiate statin therapy than patients in
the control cohort. These data suggest that the Statin Use in
Diabetes TDI had a positive impact on appropriate statin utilization within this high-risk patient population.
■■ IMPACT OF A TARGETED DISEASE INTERVENTION
TO OPTIMIZE ANGIOTENSIN-CONVERTING ENZYME
INHIBITOR AND ANGIOTENSIN RECEPTOR BLOCKER
UTILIZATION WITHIN A MANAGED CARE ORGANIZATION
Jahansouz F, Lew KH,* Vanderplas AM, Jarrar M, Stroup JR.
Prescription Solutions, 3515 Harbor Blvd., Costa Mesa, CA 92626
INTRODUCTION: An educational initiative was implemented to
optimize angiotensin-converting enzyme inhibitor (ACEI) and
angiotensin receptor blocker (ARB) therapy in patients with
diabetes.
METHODS: This pharmacy intervention highlighted the benefits
of blood pressure control in patients with diabetes to physician
providers. Educational materials presenting evidence from
Vol. 11, No. 7
September 2005
JMCP
published clinical trials and recommendations from the
American Diabetes Association were sent to primary care physicians
providing care to patients with diabetes. The mailing also
included a report listing the provider’s patients with diabetes
who had no history of ACEI or ARB utilization within the
previous 6 months. Twenty percent of identified physicians
were randomly selected as a control group and did not receive
the intervention. Six months following the intervention, ACEI
and ARB utilization of the previously identified patients under
the care of intervention and control physicians was compared to
evaluate the program’s impact. Patients not continuously enrolled
through the postintervention review period were excluded from
the evaluation analysis.
RESULTS: A total of 2,621 physicians were identified who
provided care to 29,224 patients with diabetes not previously
receiving an ACEI or ARB. A total of 24,997 patients were
available for follow-up analysis. The ACEI and ARB educational
initiative was associated with a significant increase in the
percentage of members initiating ACEI or ARB therapy during
the postintervention period. Within the intervened cohort,
3,635 of 20,045 patients (18.1%) initiated either ACEI or ARB
therapy compared with 821 of 4,952 patients (16.6%) from the
control group (P = 0.01).
CONCLUSIONS: Physicians who received educational materials
and a list of members with diabetes were more likely to initiate
ACEI or ARB therapy in their identified patients than control
physicians. These data suggest that the intervention had a positive impact on optimizing ACEI and ARB therapy within this
high-risk patient population where clinical evidence supports it
use.
■■ IMPACT OF CLINICIAN EDUCATION, COLD CARE KITS,
AND CLINICIAN PEER REVIEW ON CLINICIAN PRESCRIBING
OF ANTIBIOTICS FOR THE TREATMENT OF UPPER
RESPIRATORY TRACT INFECTIONS AT A STAFF-MODEL
HEALTH MAINTENANCE ORGANIZATION
Brzozowski DF*, Durvasula R, Murphy G. Yale University Health
Services, 17 Hillhouse Ave., New Haven, CT 06520
INTRODUCTION: The impact of clinician education, cold care kits,
and clinician peer review on clinician prescribing of antibiotics
for upper respiratory tract infections (URIs) was determined at
our staff-model health maintenance organization (HMO).
METHODS: Starting in October 2003, the Department of
Medicine and Department of Pharmacy used clinician education,
cold care kits, and clinician peer review as methods to promote
appropriate utilization of antibiotics for upper respiratory
infections (URIs) in 3 clinical departments: Internal Medicine,
Student Medicine, and Pediatrics. First, clinicians at our staffmodel HMO were educated on appropriate use of antibiotics for
URIs at mandatory staff education programs in which treatment
JMCP
September 2005
Vol. 11, No. 7
guidelines for URIs and barriers to appropriate use of antibiotics were discussed. Second, the Department of Pharmacy
stocked the Internal Medicine, Student Medicine, and Pediatrics
departments with cold care kits to be given to patients presenting
to the clinic with URIs. Thirdly, the Department of Medicine
and Department of Pharmacy developed a clinician peer review
program. On a quarterly basis, the department head provided
each clinician with a letter containing their specific antibiotic
utilization for the treatment of URIs versus the department average.
The results of the audit are being used as part of the clinician
performance review, and clinicians are encouraged to seek further
education on appropriate use of antibiotics as necessary.
RESULTS: Overall, antibiotic prescriptions for treatment of URIs
from October to December 2002 compared with October to
December 2004 were reduced from 33.3% to 26.5%, a 20.4%
reduction in antibiotic utilization. Antibiotic utilization in
Student Medicine dropped from 45% to 26.6% (41% reduction), Internal Medicine dropped from 30.2% to 24% (20.5%
reduction), and Pediatrics increased from 26.8% to 28.2%
(5.2% increase). From October to December 2004, our
clinicians dispensed 196 cold care kits to health plan members.
A total of 110 cold care kits were dispensed in Student
Medicine, 55 in Internal Medicine, and 31 in Pediatrics.
CONCLUSIONS: A multipronged approach (clinician education,
cold care kits, clinician peer review) is effective at lowering utilization of antibiotics for treatment of URIs. At our staff-model
HMO, this method had the most significant impact in our
Student Medicine department.
■■ IMPACT OF PHYSICIAN INTERVENTION ON THE
MANAGEMENT OF CONGESTIVE HEART FAILURE PATIENTS
Geierman JL*, Uchida KM. Beaver Medical Group, 242 Cajon St.,
Redlands, CA 92374
INTRODUCTION: The impact of a physician intervention program
targeting congestive heart failure (CHF) patients was determined through retrospective analysis of patient charts.
METHODS: CHF is a chronic and progressively debilitating disease
and a major public heart concern in the United States. Despite
overwhelming evidence to support their efficacy, angiotensinconverting enzyme inhibitors (ACEI) and β-adrenergic blocker
use in CHF patients is often suboptimal. Chart reviews of 775
patients were analyzed to establish a baseline of CHF management at Beaver Medical Group. Following implementation of a
physician-based intervention program, a subsequent chart
review of 125 of the patients was conducted to determine the
impact of the intervention program on CHF management and
outcomes.
RESULTS: 53 of 125 (42.4%) patients were taking a β-blocker,
and 87 (69.6%) patients were taking an ACEI or angiotensin
receptor blocker (ARB) at baseline. Following physician education
www.amcp.org
programs, 66 of 110 (60.0%) patients and 75 of 110 (68.2%)
patients were taking a β-blocker and ACEI or ARB, respectively.
Medication data were not available for 15 patients. In addition,
of the 125 patients, only 30 (24.0%) of the patients remained
on the same β-blocker regimen and 16 (12.8%) remained on
the same ACEI/ARB regimen. Almost 25% and 35% of the
patients had their β-blocker and ACEI/ARB, respectively, changed
to another drug within the same class during the study period.
CONCLUSIONS: Physician education may result in an improvement in management of patients with CHF. Physicians change
drug regimens frequently. More study is required to determine
whether these prescribing changes are due to formulary issues,
financial concerns of patients, clinical reasons, or other
confounding variables.
■■ IMPROVING PATIENT HEALTH OUTCOMES
USING PHARMACY CLAIMS DATA:
A PHYSICIAN-TARGETED POLYPHARMACY PROGRAM
Lee G*, Stroup J, Tjioe D, Chang EY, Lew KH. Prescription Solutions,
3515 Harbor Blvd., Costa Mesa, CA 92626
INTRODUCTION: A physician-targeted polypharmacy program
was implemented within an employer population to reduce
inappropriate use of pharmaceuticals; we then evaluated the
program’s impact.
METHODS: Members with a medication profile meeting defined
criteria for polypharmacy (duplicate therapies, drug-drug or
drug-disease interactions, and drugs not recommended for use
in the elderly) were identified from pharmacy claims data over
a 3-month review period for a 35,000+ member employer. The
last prescriber of the polypharmacy agent was notified by mail
of the polypharmacy incident and relevant clinical information.
A second mailing was issued to those not responding to the
initial report. Pharmacy claims were analyzed retrospectively
3 months following the intervention to evaluate the polypharmacy program’s impact. Members who became ineligible prior
to the end the postintervention review period were excluded
from analysis. Baseline polypharmacy cases not present during
the postintervention period were considered resolved.
RESULTS: Among 2,924 members with 3,988 polypharmacy
incidents identified at baseline, a total of 2,217 members with
2,932 polypharmacy incidents were available for evaluation. The
most frequently identified polypharmacy incident was for use of
drugs not recommended for the elderly (61.7%), followed by
drug-disease interactions (30.1%) and drug-drug interactions
(14.8%). Following the intervention, 54.8% of polypharmacy
cases were resolved. Resolution was highest for the polypharmacy categories “drug-drug interactions” (61.6%) and “drugdisease interactions” (58.7%) and lowest for “duplicate therapies” (50.4%). Resolution of at least one incident, and all baseline polypharmacy incidents, occurred in 59% and 52.4%,
www.amcp.org
respectively, of identified members.
CONCLUSIONS: The physician-targeted polypharmacy program
was effective in eliminating a majority of polypharmacy incidents identified over a 3-month period, reducing the potential for
adverse complications in 1,310 of 2,217 members examined
during the postintervention period. An ongoing program may
help follow up on unresolved polypharmacy incidents and also
ensure that all new polypharmacy cases are quickly identified
and evaluated for resolution.
■■ ORAL DIABETIC MEDICATION ADHERENCE
IS CORRELATED WITH A1C GOAL ATTAINMENT
IN A MANAGED CARE DIABETES
DISEASE MANAGEMENT PROGRAM
Lawrence DB*, Long LB, Israel G, Mohn LA, Parris ES. Pfizer, Inc.,
175 Beresford Creek St., Charleston, SC 29492
INTRODUCTION: Poor medication adherence is a significant barrier
to positive clinical outcomes. The purpose of this evaluation is
to determine the relationship between adherence with oral
diabetic therapy and glycosylated hemoglobin (A1c) goal attainment in a diabetes disease management program.
METHODS: This was a retrospective, descriptive evaluation of
patients enrolled in a managed care diabetes disease management program. A dataset analysis containing demographic,
enrollment, pharmacy claims, and clinical lab data was
performed. Continuously enrolled patients with a documented
A1c obtained at least 90 days after the initial sulfonylurea and
metformin prescription index dates were included. The medication possession ratio (MPR) was calculated from the prescription
claims records and correlated with the A1c value.
RESULTS: Forty-two percent of patients on sulfonylurea therapy
and 46% of metformin were reaching an A1c goal of <7.0, and
the average MPRs were 0.76 (+0.31) and 0.69 (+0.3), respectively.
The average MPR for sulfonylurea-utilizing patients reaching
and not reaching the A1c goal was 0.82 (+0.29) and 0.72
(+0.31), respectively (P <0.0001, students t test). The average
MPR for metformin patients reaching and not reaching the A1c
goal was 0.77 (+0.3) and 0.62 (+0.3), respectively (P <0.0001,
students t test). A Pearson correlation showed significant positive associations between A1c and the MPR (sulfonylureas:
r= -0.295, P <0.001; and metformin: r = -0.285, P <0.001).
CONCLUSIONS: This evaluation shows that oral diabetic medication
adherence is a significant factor in A1c goal attainment for
diabetes patients. Nonadherence to medications should be considered and evaluated when a patient is not reaching a clinical goal.
Barriers to adherence should be assessed and interventions to
improve adherence implemented.
Vol. 11, No. 7
September 2005
JMCP
■■ OUTCOMES OF A PROTON PUMP
INHIBITOR FORMULARY COMPLIANCE PROGRAM
Nelson CL*, Williams CE, Handke LJ. Blue Cross and Blue Shield of
Nebraska, 7261 Mercy Rd., Omaha, NE 68180
OBJECTIVE: The goal of the study was to determine if utilization
of formulary proton pump inhibitor (PPI) agents could be
increased by providing formulary and benefit design information
along with formulary brand PPI mail-in rebates to targeted
members with claims for non-formulary PPIs.
METHODS: A prospective cohort study was conducted.
The intervention cohort Blue Cross and Blue Shield of Nebraska
(BCBSNE) was compared with a nonintervention control Blue
Cross and Blue Shield plan (BCBS). Formulary options for PPIs
were identical for both health plans. Members met study criteria
if they had continuous enrollment, 3-tier drug copay benefit,
and at least one nonformulary PPI claim during the 3-month
period prior to mail date November 22, 2004. The analysis
period was November 29, 2004, to March 1, 2005. Inclusion
criteria were met by 3,379 BCBSNE members (~400,000 lives)
and 2,028 BCBS comparison group members (~320,000 lives).
The BCBSNE intervention detailed the 3-tier benefit by formulary
status of PPIs and copay differentials. Formulary-brand PPI
mail-in rebates were included, which reimbursed the member
up to $30 in first prescription rebate dollars. The primary end
point measured was the incidence of generic or formularybrand PPI claims during the analysis period. Statistical 2-sided
chi-square analyses performed with SPSS version 13.0 (SPSS,
Inc., Chicago, IL).
RESULTS: There were 946 (28%) BCBSNE members and 473
(23%) BCBS control members excluded because they had no
PPI claims during the follow-up period. In the 3 months after
the letter, 265 of 2,433 (11%) intervention group members and
33 of 1,518 (0.02%) control group members had a claim for a
generic or formulary PPI agent claim, P <0.001. In the intervention
group, 130 of 265 (49%) of claims were for a generic PPI. Of
the intervention group members, 232 (86%) maintained generic or
preferred-brand formulary agents throughout the follow-up
period, P = 0.028.
CONCLUSION: Targeted formulary education regarding formulary
PPI agents and brand-formulary mail-in rebate incentives directed
toward members utilizing nonpreferred PPI agents were associated
with a statistically significant switch to formulary agents.
■■ PATTERNS OF ANTIDEPRESSANT USE
AND COST IMPLICATIONS OF PRODUCT SWITCHING
Kruzikas D, Khandker RK*, McLaughlin T, Tedeschi M. Wyeth
Research, 500 Arcola Rd., Collegeville, PA 19426
INTRODUCTION: The study examines patterns of antidepressant
use, including drug switching and related resource utilization.
JMCP
September 2005
Vol. 11, No. 7
METHODS: Using retrospective claims of managed care enrollees
using a national database (PharMetrics), the study follows
newly diagnosed depression patients (aged 18+ years) with
newly prescribed anti-depressants. We identified the proportion
of switchers from commonly prescribed selective serotonin reuptake inhibitors (SSRIs: fluoxetine, citalopram, sertraline, and
paroxetine) to serotonin and norepinephrine reuptake inhibitors
(SNRIs: venlafaxine), and vice versa. We then aggregated health
care costs for a 1-year period following diagnosis for various
switcher groups. Multivariate regression analyses determined
predictors of switching and factors influencing overall and
depression-related costs while controlling for confounding factors.
RESULTS: Of the 48,950 patients included in the study population,
89% were treated with SSRIs and 11% with SNRIs. Between
12% to 15% of patients switched antidepressants. Of the SSRI
switchers, 29% switched to an SNRI. Increased likelihood of
switching was associated with female gender, Medicaid coverage,
prior anxiolytic use, treatment by a psychiatrist or psychologist,
and paroxetine as the index medication. Compared with SSRI
nonswitchers, costs for SSRI switchers were 36% higher for all
causes and 58% higher for depression-related causes. In contrast,
compared with SNRI nonswitchers, costs for SNRI switchers
were 27% higher for all causes and 5% higher for depressionrelated causes. Thus, relatively more costly patients are switching from SSRIs to SNRI than vice versa. In addition, among SSRI
patients switching to SNRI, costs increased with the number of
switches. Mutlivariate analyses confirmed that switching was
associated with higher overall and depression-related costs.
CONCLUSIONS: Switching among antidepressants is quite frequent among depression patients. Switchers incur significantly
higher overall and depression-related costs, and, in general,
more costly SSRI patients end up switching antidepressants.
■■ PERSISTENCE WITH ANTIHYPERTENSIVE
MONOTHERAPY IN A MANAGED CARE SETTING
Patel BV*, Remigio-Baker RA, Thaker DJ, Preblick R. MedImpact
Healthcare Systems, Inc., 10680 Treena St., 5th Fl., San Diego, CA
92131
INTRODUCTION: Persistence in taking frequently prescribed classes
of anti-hypertensive (AHY) therapy was assessed through a
longitudinal, retrospective analysis of pharmacy records.
METHODS: Pharmacy claims data (2001-2003) from the
MedImpact database were used to identify patients with (1) at
least 1 prescription claim from January 1, 2001, to December
31, 2003 (index date), (2) 6 months of negative medication
history for AHY therapy, and (3) continuous benefit-eligibility
6 months preindex and 1-year post-index date. Patients were
followed for 1 year to assess persistence and medication ownership ratio (MOR: the proportion of members with therapy at
end of follow-up). Multiple variable linear regression was used to
www.amcp.org
assess differences in MOR, adjusted for age, gender, business
segment, comorbidity, and concurrent cardiovascular-related
medication utilization. Pair-wise comparisons were performed.
RESULTS: The study cohort consisted of 304,818 patients who
initiated AHY monotherapy: 103,064 on β-blockers (BBs),
99,154 on angiotensin-converting enzyme inhibitors (ACEIs),
47,229 on calcium channel blockers (CCB), 41,651 on diuretics
(D), and 13,720 on angiotensin receptor blockers (ARBs).
A higher proportion of initial ARB patients (53.3%) remained
persistent at 12 months past the index date compared with
ACEI (49.3%), BB (42.0%), CCB (41.7%), and D (29.9%).
Adjusted MORs were greatest for ARB patients (44.8%),
followed by ACEI (43.8%), BB (37.3%), CCB (36.5%), and
D (27.0%). All pair-wise comparisons were significant
(P <0.0001), except BB versus CCB, for persistence.
CONCLUSIONS: Initiating AHY with ARBs versus other AHY classes tends to have a greater persistence and proportion of members on
therapy after 1 year. These findings have important implications for
managed care since optimal medication-taking behavior
influences the extent of blood pressure control and the
reduction of long-term cardiovascular risks. Further research is
needed to assess whether clinical differences exist among AHY
therapies as well as the relationships between utilization and
health outcomes.
■■ PHARMACOECONOMIC ANALYSES OF
ANGIOTENSIN RECEPTOR BLOCKER THERAPY
IN PATIENTS WITH MILD-TO-MODERATE HYPERTENSION
Joish VN, Brixner DI.* Pharmacotherapy Outcomes Research Center,
University of Utah, 421 Wakara Way, Salt Lake City, UT 84108
INTRODUCTION: Seven angiotensin II receptor blockers, also
known as sartans (candesartan, eprosartan, irbesartan, losartan,
olmesartan, telmisartan, and valsartan), have been approved for
the treatment of hypertension. The purpose of this study was to
conduct a cost-efficacy analysis (CEA) based on clinical outcomes and costs to determine the least costly agent per treatment
success from a managed care perspective.
METHODS: Efficacy data were derived from published head-tohead clinical studies or meta-analyses that evaluated
angiotensin receptor blockers (ARBs) in mild-to-moderate hypertension patients. The primary efficacy outcome was diastolic
blood pressure reduction in mm Hg at week 6 to 8 from
baseline. Average efficacy rates were determined by calculating
averages of the efficacy rates in the individual studies. Costs to
treat were averaged for multiple dosing regimens. Costs included
were 2005 average wholesale prices from FirstData Bank.
RESULTS: Eighteen published studies were identified, and efficacy
data from these studies were used in the CEA model. Exclusions
were applied for using doses or durations that differed from the
primary efficacy outcome. The highest average diastolic blood
www.amcp.org
pressure reduction from baseline was for losartan
hydrochlorothiazide 100 mg/25mg (17.5 mm Hg), and the lowest
was for valsartan 160 mg (5.3 mm Hg). The lowest average cost
per diastolic blood pressure reduction was for eprosartan, at an
average of $0.14, and highest for candesartan and valsartan, at
an average of $0.24.
CONCLUSION: Using this simple straightforward pharmacoeconomic method, eprosartan mesylate had the lowest cost-efficacy
ratio compared with other sartans in mild-to-moderate hypertensives.
■■ PREVALENCE AND FACTORS ASSOCIATED WITH
VACCINATION RATES AMONG U.S. ADULTS AT HIGH RISK
OF VACCINE-PREVENTABLE HEPATITIS
Cantrell CR*, Chen H, DeBartlo C. GlaxoSmithKline, Five Moore Dr.,
Research Triangle Park, NC 27709
INTRODUCTION: This study sought to estimate the prevalence of
adults at high risk of vaccine-preventable hepatitis (hepatitis A
virus [HAV] and hepatitis B virus [HBV]) in the United States
and their vaccination rates. Secondly, the study investigated the
association between vaccination rates and demographic
(age, gender, location of birth, and race) and social economic
(annual household income, education level, and marital status)
characteristics.
METHODS: Four years (1999-2002) of publicly available
National Health and Nutrition Examination Survey data were
utilized. Survey participants aged 20 to 59 years were selected.
Survey participants were considered at high risk of vaccinepreventable hepatitis if they belonged to a “risk population
group” where their situation and/or behavior placed them at a
higher risk of contracting hepatitis, as identified by the Centers
for Disease Control and Prevention. All prevalence
estimates were weighted to represent the total U.S. population,
using 4-year interview and examination weights. Logistic
regression was utilized to identify demographic and social
economic factors associated with vaccination rates.
RESULTS: The study included 6,237 survey participants who
represent 153,919,438 adults aged 20 to 59 years in the United
States. Of those adults, 12,347,634 (8.0%) were at high risk of
HAV, 18,849,536 (12.3%) were at high risk of HBV, and
2,141,907 (1.4%) were at high risk of both HAV and HBV. The
vaccination rates among these high-risk groups were 13%,
23.6%, and 13.4%, respectively. The most prevalent risk groups
were persons with sexually transmitted diseases and persons
using illegal drugs. Within the higher-risk population,
single males between the ages of 20 and 29 years were significantly (P <0.05) less likely to be vaccinated than their counterparts.
CONCLUSION: Among persons identified at high risk of vaccinepreventable hepatitis (HAV, HBV, or both), only a small proportion
Vol. 11, No. 7
September 2005
JMCP
of this population had evidence of hepatitis vaccination (13%,
23.6%, and 13.4%, respectively).
■■ REAL-WORLD DOSING BEHAVIORS FOR
SELF-ADMINISTERED ANTITUMOR NECROSIS FACTORS
Vishalpura TV*, Grogg A, Abeyta K. Applied Health Outcomes, 4114
Woodlands Pkwy., Suite 500, Palm Harbor, FL 34685
OBJECTIVE: To assess real-world dosing patterns for self-administered anti-tumor necrosis factors (TNF).
METHODS: Retrospective analyses of National Data Corporation’s
database were conducted. Patients with a pharmacy claim for
etanercept (25 mg, 50 mg) or adalimumab (40 mg) for various
therapeutic areas, including inflammatory, gastroenterology, and
dermatology, from November 2004 through January 2005 were
analyzed. Prescription claims were stratified into 3 cohorts:
new, continuing, or switched. Patients in the new prescription
cohort were those without a claim 3 months preceding the
study period. Patients in the continuing prescription cohort
were those with a claim 3 months preceding the study period.
Patients in the switched prescription cohort were defined as
those switched from 25 mg to 50 mg etanercept. Average dose
per patient and percentage deviation from reference dosing were
calculated for etanercept (25 mg, 50 mg) and adalimumab for
each cohort and therapeutic area.
RESULTS: For both drugs, patients in the new cohort had greater
dose deviations and higher average doses than those in the continuing cohort across all therapeutic areas. Among patients in
the new cohort, both agents had doses well above the reference
dose. Dermatology patients utilized the highest average dose for
both drugs. The average dose was 94% higher than the onceweekly reference dose among patients receiving the etanercept
50 mg formulation. Etanercept patients in the 25 mg continuing
cohort received doses that were slightly below the twice-weekly
reference dose, and those in the 50 mg continuing cohort
received doses that were 25% above the once-weekly reference
dose. In the adalimumab continuing cohort, doses were 10%
above the 40 mg every-other-week reference dose. Specifically,
for gastroenterology, adalimumab had the highest average dose,
at 66% above the reference dose.
CONCLUSION: Overall, in the cohorts examined through a retrospective review of claims data, this study illustrates that doses well
above reference doses are commonly prescribed for the self-administered anti-TNFs.
■■ RESOURCE USE AND ANEMIA TREATMENT COSTS
AMONG CANCER PATIENTS TREATED WITH EPOETIN ALFA
OR DARBEPOETIN ALFA
Harley C, Muser E*, Nelson M, McKenzie RS, Piech CT. Ortho
Biotech Clinical Affairs, LLC, 1061 Brook Mont Dr., O’Fallon, MO
63366
INTRODUCTION: This study examines resource use and treatment
costs in cancer patients receiving epoetin alfa (EPO) or darbepoetin alfa (DARB) in a managed care setting.
METHODS: A retrospective analysis was conducted using medical
claims from a large U.S. health plan from January 1, 2002,
through December 31, 2003. Cancer patients aged 18 years and
older with at least 2 doses of EPO or DARB were eligible for
inclusion. Patients with renal failure or who were switched
between erythropoietic agents were excluded. Treatment
episode was defined as the period from first EPO/DARB claim
to last claim (with no gaps in treatment greater than 30 days)
with a 3-month maximum duration of treatment. Resource use
and treatment costs were examined for each group.
RESULTS: 4,753 EPO and 1,601 DARB patients met inclusion
criteria. Patient age was similar between groups with a higher
proportion of men in the EPO group compared with the DARB
group (EPO 34%, DARB 29%, P <.001). Treatment duration
(approximately 8 weeks) and proportion of patients transfused
were similar. While the number of hemoglobin determinations
and cancer-related outpatient visits overall were similar, the
number of anemia-related outpatient visits were lower in the
DARB group (EPO 6.92 vs. DARB 4.89, P<.001). Mean anemiarelated costs (per patient per episode) were significantly lower
for the EPO group than for the DARB group, respectively ($3,943
vs. $5,723, P <.001).
CONCLUSIONS: In this managed care population, mean anemiarelated costs were 45% higher for the DARB group than for the
EPO group. A similar number of cancer-related office visits and
hemoglobin determinations were observed between groups
despite less-frequent DARB dosing, suggesting that factors other
than frequency of erythropoietic agent administration determine resource utilization in this population.
■■ SYNCHRONICITY: A STUDY TO EVALUATE THE
EFFECTIVENESS OF DARBEPOETIN ALFA AT 300 MCG EVERY
3 WEEKS ON CLINICAL OUTCOMES IN CANCER PATIENTS
WITH ANEMIA DUE TO CHEMOTHERAPY
Boccia R, Silberstein P, Tchekmedyian S, Tomita D, Rossi G*, Otterson
GA. Amgen, Inc., One Amgen Center Dr., Thousand Oaks, CA 91320
OBJECTIVE: To assess the effectiveness of darbepoetin alfa
administered at 300 mcg every 3 weeks (Q3W) in achieving and
maintaining hemoglobin (Hb) levels recommended by current
evidence-based guidelines.
JMCP
September 2005
Vol. 11, No. 7
www.amcp.org
METHODS: This interim analysis of a multicenter, open-label,
16-week study of patients with chemotherapy-induced anemia
(CIA) includes 1,225 patients who received at least 1 dose of
darbepoetin alfa. Study end points included the proportion of
patients who achieved and maintained Hb levels within the
range recommended by current evidence-based guidelines
(11 to 13 g/dL), the proportion of patients who required red
blood cell (RBC) transfusions, and patient-reported changes in
symptoms of fatigue measured using the Functional Assessment
of Cancer Therapy-Fatigue (FACT-F) scale.
RESULTS: Seventy-nine percent of patients achieved the target
Hb range, of which 72% maintained Hb levels within the range
of 11 to 13 g/dL, consistent with current evidence-based guidelines. The mean (95% confidence interval) change in Hb from
baseline to the end of the treatment period was 1.5 g/dL (1.3,
1.6) (last-value-carried-forward approach). Darbepoetin alfa
reduced the proportion of patients who required RBC transfusions from 11% in the first month to 4% in the last month of
the study. Increases in Hb levels from the start to the end of the
study period were associated with clinically significant improvements in FACT-F.
CONCLUSIONS: Darbepoetin alfa administered at 300 mcg Q3W
is well tolerated by cancer patients with CIA. Most patients
achieved and maintained Hb levels within the range recommended by the current evidence-based guidelines, required
fewer RBC transfusions, and demonstrated improvements in
symptoms of fatigue. Administering darbepoetin alfa Q3W may
simplify the treatment of CIA in the oncology practice by synchronizing darbepoetin alfa therapy with chemotherapy, thereby
reducing disruption to the life of patients and their caregivers.
resource use were compared between the preperiod and postperiod. Statistical tests were conducted using McNemar’s tests
for categorical variables and Wilcoxon tests for continuous
variables. The most common TPM-associated adverse events
observed in pivotal trials included paresthesia, cognitive
impairment, fatigue, anorexia, and nausea.
RESULTS: The study included 1,749 TPM prophylaxis patients.
The mean ± SD age was 43 ± 10 years, and 90% were female.
The mean ± SD number of TPM claims per patient in the
6-month postperiod was 5.5 ± 1.8. Changes in preacute versus
postacute medication and resource use were as follows: 15%
decrease in triptan prescriptions (P <.0001); 7% decrease in
other abortive medications (P <.0001); 8% decrease in outpatient services (P = .045), which included a 45% decrease in
emergency room services (P<.0001) and a 50% decrease in procedures such as computed tomography scan and magnetic
resonance imaging (P =.0003); and 52% decrease in hospital
admissions (P =.001).
CONCLUSION: This is the first study to examine the impact of
TPM prophylaxis on the resource use of migraineurs in a realworld setting. In this setting, use of TPM was associated with
significantly lower acute medications and health care resource
use. Future studies should examine the longer-term impact of
TPM prophylaxis.
■■ TOPIRAMATE IN THE PROPHYLAXIS OF MIGRAINE:
IMPACT ON RESOURCE UTILIZATION
Vermilyea JA, Feliu AL, Rupnow MFT*, Carter CT, Blount A,
Ollendorf DA. PharMetrics, Inc., 1125 Trenton-Harbourton Rd.,
Titusville, NJ 08560
OBJECTIVE: To describe the demographic and pharmacy/medical
resource utilization profile of patients diagnosed with migraine
who received topiramate (TPM) prophylaxis.
METHODS: The Pharmetrics Patient-Centric Database was used
to obtain pharmacy and medical claims data from January 1,
1999, to September 30, 2004. Migraine patients were required to
have continuous enrollment 6 months prior to and after an
index TPM prescription. In the 6 months immediately preceding
the TPM index prescription, patients had to have ≥1 triptan
prescription with a minimum of 6 and a maximum of 180 triptan
equivalents (TEs). Patients with claims for other U.S. Food and
Drug Administration (FDA)-approved prophylaxis therapy in
the postperiod were excluded. Demographics, comorbidities,
prescriber specialty, and migraine-related inpatient and outpatient resource use were assessed using descriptive statistics.
RESULTS: The study included 1,749 TPM prophylaxis patients.
Mean ± SD age was 43 ± 10 years, with 83% of patients aged
between 25 and 54 years. Ninety percent were female and 81%
were commercially insured. Neurology providers initiated 54%
of TPM prescriptions. The most common comorbidities were
OBJECTIVE: To evaluate the effect of topiramate (TPM) prophylaxis on acute medication use and medical resource utilization
among patients diagnosed with migraine.
METHODS: The Pharmetrics Patient-Centric Database was used
to obtain pharmacy and medical claims data. The first TPM
claim between January 1, 1999, and September 30, 2004, was
considered the index date. Triptan use was standardized to triptan equivalents (TE) based on the maximum daily dose allowed.
Analysis included patients continuously enrolled for at least
6 months preindex and postindex date, with ≥6 TEs (but ≤180
TEs) during 6 months preindex date and ≥3 claims for TPM
prescription. Patients with use of other U.S. Food and Drug
Administration-approved prophylaxis therapy in the postperiod
were excluded. Migraine-related inpatient and outpatient
www.amcp.org
■■ TOPIRAMATE IN THE PROPHYLAXIS OF MIGRAINE:
PATIENT PROFILES FROM A PAYER DATABASE
Vermilyea JA, Feliu AL, Rupnow MFT*, Carter CT, Blount A, Dodd S,
Ollendorf DA. PharMetrics, Inc., 1125 Trenton-Harbourton Rd.,
Vol. 11, No. 7
September 2005
JMCP
depression and psychosis (16%), allergy (11%), gastrointestinal
disorders (8%), and anxiety (6%). Mean ± SD Charlson
Comorbidity score was 0.20 ± 0.6. Prior to receiving TPM, 78%
had a prescription for FDA- and non–FDA-approved prophylaxis therapy and 72% used abortive prescription medications
in addition to triptans. The mean ± SD number of TEs per person
was 23.2 ± 19.4. Six months prior to receiving TPM, these
migraineurs had a mean ± SD of 0.64 ± 3.32 emergency room
visits, 2.12 ± 2.99 physician office visits, 0.08 ± 1.72 procedures,
and 0.41 ± 1.72 lab tests, cumulating to 4.20 ± 8.03 migrainerelated outpatient services. Additionally, they had 4 hospitalizations
and 18 days hospitalized per 100 migraineurs.
CONCLUSION: In the 6 months prior to TPM initiation,
migraineurs showed patterns of extensive acute and prophylactic
pharmacy resources as well as health services utilization. This is
the first study to examine the TPM patient profile from a claims
perspective. Future studies should examine the impact of TPM
prophylaxis in reducing migraine-related resource consumption.
■■ TREATMENT RATES AND GLYCEMIC CONTROL
IN U.S. ELDERLY TYPE 2 DIABETIC PATIENTS, 2001-2002.
Suh DC, Shin H, Wogen J, Plauschinat C.* Novartis Pharmaceuticals
Corporation, One Health Plaza, East Hanover, NJ 07936
OBJECTIVE: To characterize the elderly type 2 diabetes (T2DM) population in the United States, we evaluated patient characteristics,
treatment rates, and glycemic control in a U.S.-representative,
weighted sample of T2DM patients aged ≥65 years.
METHODS: The National Health and Nutrition Examination
Survey 2001-2002 (n = 11,039) was used to identify persons
with T2DM aged ≥65 years, defined by physician diagnosis at
>30 years or current or past use of insulin and/or oral agents.
Body mass index was used to classify patients as normal weight
(<25 BMI), overweight (25-29 BMI), or obese (>30 BMI). Data
were analyzed using SAS and SUDAAN statistical software.
RESULTS: Age-adjusted prevalence was 14.22% (SE = 0.86),
estimating that 4.3 million elderly persons in the United States
have been diagnosed with T2DM. Of these, 56.9% were aged
65 to 74 years, 34.3% 75 to 84 years, and 8.8% >85 years. Sixty
percent were female, and 80% were non-Hispanic white, 9%
non-Hispanic black, and 9% Hispanic. Approximately 40%
were overweight and 46% obese. Mean glycosylated hemoglobin
(A1c) for elderly T2DM patients was 6.99%; 42.3% had A1c
>7%. 83.7% of patients were treated with an oral agent and/or
insulin; half (49.9%) of treated patients had A1c >7%. 56.1% of
patients treated with only oral agents, 33.7% of patients treated
with only insulin, and 38.5% of patients treated with oral
agents and insulin had A1c <7%.
CONCLUSIONS: Though 84% of elderly T2DM patients were
treated with oral agents and/or insulin, glycemic control rates in
this population were strikingly poor, as 49.9% of treated patients
JMCP
September 2005
Vol. 11, No. 7
had A1c >7%. Almost half (46%) of elderly T2DM patients were
obese. This study highlights a significant unmet need for more
aggressive treatment of elderly T2DM patients to achieve
American Diabetes Association-recommended A1c goal.
■■ TREATMENT SATISFACTION AND PATIENT
FUNCTIONING WITH OXYBUTYNIN OR TOLTERODINE
IN PATIENTS WITH OVERACTIVE BLADDER
Wan GJ, Chen A*, Bolge SC. Ortho Urology, Ortho-McNeil
Pharmaceuticals, Inc., 1000 Route 202, Raritan, NJ 08869
OBJECTIVE: To compare satisfaction and patient functioning with
immediate-release (IR) and extended-release (ER) formulations of
oxybutynin or tolterodine in patients with overactive bladder
(OAB).
METHODS: Cross-sectional data were obtained from the
Consumer Health Sciences 2004 National Health and Wellness
Survey, a nationally representative sample of a noninstitutionalized U.S. civilian population. Patients were currently taking
either IR or ER oxybutynin (n = 162) or IR or ER tolterodine
(n = 277) to treat OAB. Patient satisfaction with treatment (PST)
was measured using a 5-point scale (1 = not at all satisfied to
5 = extremely satisfied). Satisfaction rates were computed as the
percentage of patients reporting a 4 or 5 on the PST scale.
Patient functioning was assessed using the mental and physical
component summary scores of the 8-item Short-Form Health
Survey and the activity impairment score of the Work
Productivity and Activity Impairment Questionnaire. A P value
<0.05 in a 2-tailed test was considered significant.
RESULTS: The mean age of patients was 58 years, and 84% were
female. More patients reported that they were satisfied with
oxybutynin (66%; 107 of 162) versus tolterodine (48%; 132 of
277; P <0.001), with an adjusted odds ratio of 1.98 (95% CI,
1.31, 3.00; P <0.001). Those satisfied with their studied OAB
medication reported statistically significantly better patient
functioning compared with those who were dissatisfied.
CONCLUSIONS: In this study, significantly more patients were
satisfied with oxybutynin versus tolterodine. Satisfaction with
OAB therapy and patient functioning are important factors to
consider in the selection of OAB medication by health care
professionals and payers.
■■ TREATMENT WITH ETANERCEPT IMPROVES
PATIENT-REPORTED OUTCOMES IN PATIENTS WITH
MODERATE-TO-SEVERE PSORIASIS
Gottlieb AB, Chiou CF*, Woolley JM, Lalla D, Jahreis A. Amgen, Inc.,
One Amgen Center Dr., MS 28-3-B, Thousand Oaks, CA 91320
OBJECTIVE: To evaluate the effect of etanercept therapy on patientreported outcomes in patients with moderate-to-severe psoriasis.
www.amcp.org
METHODS: A 12-week, double-blind, multicenter clinical trial
was conducted in the United States and Canada to confirm the
efficacy and safety of etanercept 50 mg twice weekly in patients
with psoriasis. Patients with stable moderate-to-severe psoriasis
were randomized to receive subcutaneous etanercept 50 mg
twice weekly or placebo. Patient-reported outcomes included
the Dermatology Life Quality Index (DLQI), a disease-targeted
health-related quality-of-life questionnaire. The DLQI has
10 questions that cover 6 life areas (symptoms and feelings,
daily activities, leisure, work and school, personal relationships,
and treatment satisfaction). Lower scores indicate better health
status; a 0 score means that a patient is “not at all” bothered by
psoriasis in the 6 life areas. Patients who received at least one
dose of study drug and provided baseline DLQI assessments
(N = 308 for etanercept; N = 304 for placebo) were included.
RESULTS: Relative to patients on placebo, patients on etanercept
achieved statistically significantly greater percentage improvements in DLQI total score by week 1 (P < 0.05). At week 12, the
mean percentage improvement in DLQI was 69.1% for patients
receiving etanercept compared with 22% for patients on placebo
(P < 0.0001). Furthermore, at week 12, the improvements in
each of the 6 DLQI life areas for patients receiving etanercept
were statistically significantly superior to those for patients
receiving placebo (P < 0.001). At week 12, significantly more
patients on etanercept achieved a 0 score on the DLQI (28% on
50 mg twice weekly and 3% on placebo; P < 0.0001).
CONCLUSIONS: Treatment with etanercept improves patient-reported outcomes in patients with moderate-to-severe psoriasis.
Combined with clinical efficacy, these results support the benefit
of etanercept in patients with moderate-to-severe psoriasis.
■■ UTILIZATION OF CHOLINESTERASE INHIBITORS
IN THE TREATMENT OF ALZHEIMER’S DISEASE
Mucha L*, Cuffel B, McCrae T, Mark T, Wang S. Thomson Medstat,
Inc., 125 Cambridge Park Dr., Cambridge MA 02140
days on therapy, and medication possession ratio.
METHODS: Data were from MarketScan’s Medicare Claims
Database from 2001 through 2003 (n = 3,177). Inclusion criteria were (1) age 65 years or older, (2) at least 1 claim with an
International Classification of Diseases, Ninth Revision (ICD-9)
code for Alzheimer’s disease (331.0), (3) at least
1 cholinesterase inhibitor prescription preceded by a 6-month
period without any such prescription, and (4) at least
18 months of continuous enrollment. A 30-day gap between
prescription end date and fill date indicated discontinuation.
Cohorts were defined by their starting medication (index therapy)
and were similar in most patient characteristics. Logistic regression
models tested for cohort differences in discontinuation and
switching, controlling for demographics, region of the country,
type of insurer, and Charlson index.
RESULTS: Fewer patients started on galantamine (73%) and
rivastigmine (79%) reached an effective dose than those started
on donepezil (99%, P <.0001). Donepezil patients had higher
rates of medication persistence starting at month 5 compared
with rivastigmine, and at month 7 compared with galantamine
(P <.05). Switching rates were higher for rivastigmine (odds
ratio [OR]=1.57, P<.0001) than donepezil patients but were not
different for galantamine patients. Similarly, discontinuation rates
were higher for rivastigmine (OR = 1.22, P <.05) than donepezil
patients but were not different for galantamine patients.
Donepezil patients spent more days on index therapy than
galantamine (215 days) or rivastigmine (206 days) patients.
Cohorts did not differ in medication possession ratio.
CONCLUSIONS: Choice of initial therapy affects medication
adherence in a cognitively impaired Medicare population.
Patients initiated on donepezil were more likely to reach an effective dose, persist on therapy, and have more days on the initial
therapy than patients started on other cholinesterase inhibitors.
INTRODUCTION: This study compared use of cholinesterase
inhibitors by patients diagnosed with Alzheimer’s disease on
time-to-effective-dose, persistence, discontinuation, switching,
www.amcp.org
Vol. 11, No. 7
September 2005
JMCP
RESIDENCIES, FELLOWSHIPS, AND OTHER PROGRAMS
Managed Care Pharmacy Residencies,
Fellowships, and Other Programs
T
he following is a partial list of available managed care
pharmacy residency and fellowship programs and other
programs compiled as of August 2005. The residencies
listed were submitted by AMCP members in response to AMCP’s
call for residency program listings. This is not a comprehensive
list of all available programs. AMCP provides it solely as a service
to its readers. This list does not imply AMCP’s endorsement of
any particular program nor does AMCP guarantee the availability
of any of the programs listed. AMCP does not assume responsibility for any errors that may appear in these listings. If you are
aware of additional residency and fellowship programs not listed
here, please contact AMCP at (800) TAP-AMCP.
■■ AMERICA SERVICE GROUP
Managed Care
Accredited:
Yes
Length of Program:
1-year, 2-year option
Number of Positions: 1
Affiliation:
Various universities
Application Deadline: February 15
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Yes
Educational/Special Requirements: Graduate of accredited
school of pharmacy
Fringe Benefits: 3 weeks paid personal leave, paid sick leave,
full medical and dental, tuition assistance toward specified
degree programs
Special Features: America Service Group is a closed-model
HMO that provides comprehensive at-risk health care; Secure
Pharmacy Plus is the pharmacy benefits manager.
Contact Information:
Peter Mikhail
Vice President, Clinical Services
American Service Group/Secure Pharmacy Plus
416 Mary Lindsay Polk Dr., Suite 515
Franklin, TN 37067
(615) 771-1457
(615) 771-4557 (fax)
[email protected]
pharmacist license or eligibility for licensure
Fringe Benefits: Medical, dental, vision, holidays, vacation,
and attendance at national conference(s)
Special Features: Activities the resident will be involved in
include, but are not limited to, P&T committee participation
and presentations; formulary management and review; one-onone physician correspondence on evidence-based medicine
with current clinical studies reviewed and presented, focusing
on actual patients who may benefit from this information; new
drug review and its placement with available therapies; clinical
participation and set-up; client summary report write-up and
delivery; participation in various education conference and
clinical studies.
Nazly Westernoff
American Health Care
3001 Douglas Blvd., #320
Roseville, CA 95661
(916) 773-7227
(916) 773-7210 (fax)
[email protected]
■■ AMERICAN HEALTH CARE
Clinical Therapeutics / Managed Care Pharmacy
Accredited:
Being pursued
Length of Program:
12 months
Affiliation:
None
Application Deadline: Open
Starting Date:
July 1
Estimated Stipend:
$43,000 and up
Onsite Interview:
Yes
Educational/Special Requirements: PharmD with a California
JMCP
September 2005
Vol. 11, No. 7
■■ AON CONSULTING
Pharmacy Benefits Consulting
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Application Deadline: January 3
Starting Date:
July 5
Estimated Stipend:
$36,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD, managed care
course work preferred
Fringe Benefits: 2 weeks paid vacation, paid holidays, medical/dental insurance, travel budget, professional meetings
Special Features: Aon Consulting is the second largest
employee benefits consulting firm nationwide. This unique
program provides residents with the opportunity to help
www.amcp.org
Managed Care Pharmacy Residencies, Fellowships, and Other Programs
employers and health plans better manage and control the
prescription drug benefit for their employees or members.
Residents will be exposed to all aspects of pharmacy benefit
management consulting: plan design modeling, formulary
analysis, clinical programming, disease management, audits,
regulatory, trends and forecasting through proprietary actuarially
based models. Residents will interact with benefit administrators,
major PBMs, PPOs, HMOs, and disease management firms
nationally. Currently, there are 2 resident positions available,
based in Chicago, IL, and Philadelphia, PA.
Connie Perry
Vice President
Aon Consulting
330 East Kilbourn Ave., Suite 450
Milwaukee, WI 53202
(414) 225-5345
(414) 276-3929 (fax)
[email protected]
■■ APPLIED HEALTH OUTCOMES
Health Outcomes Research Fellowship
Accredited:
No
Length of Program:
2 years
Affiliation:
University of South Florida,
College of Public Health
Starting Date:
July 1
Estimated Stipend:
Contact program
Onsite Interview:
Yes
Educational/Special Requirements: PharmD or equivalent
(residency preferred)
Fringe Benefits: Competitive salary, health insurance, vacation,
401(k), tuition for required classes, and travel expenses to
1 national meeting per year
Special Features: This 2-year, degree-granting fellowship
provides a unique research and education experience in an outcomes consulting environment. Research activities include, but
are not limited to, quality improvement programs, database
analysis, economic modeling, and development of researchbased manuscripts. In addition, the fellow will obtain an MSPH
or MPH degree from the University of South Florida, College of
Public Health.
James H. Jackson IV, PharmD, MPH
Senior Consultant
Applied Health Outcomes
4114 Woodlands Pkwy., Suite 500
Palm Harbor, FL 34685
(727) 771-4100
www.amcp.org
(727) 771-4145 (fax)
[email protected]
Web: http://www.applied-outcomes.com
■■ BIOSCRIP
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD, licensure in the
state of Ohio
Fringe Benefits: 10 vacation days, 5 holidays, 4 floating holidays, 2 personal days, health benefits, and professional travel
allowance
Special Features: This 12-month residency program is
designed to provide a firm knowledge base of managed care
from the pharmacy benefit perspective and allows residents to
excel in clinical services, formulary management and research,
participation in formulary development and medication management, and provision of clinical services, including drug information, clinical program development, pharmacoeconomic
assessment, and disease state management as well as the
strengthening of written and verbal communication skills.
Suzanne Tschida, PharmD, BCPS
Director, Clinical Services
BioScrip
10900 Red Circle Dr.
Minnetonka, MN 55343
(800) 444-5951, ext. 3621
(952) 352-6785 (fax)
[email protected]
■■ BLUE CROSS AND BLUE SHIELD OF ALABAMA
Managed Care
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1 (flexible)
Estimated Stipend:
$32,000
Onsite Interview:
Yes
experience
Vol. 11, No. 7
September 2005
JMCP
Fringe Benefits: Paid vacation, personal holiday leave,
health/dental insurance, no on-call responsibilities
Special Features: The program provides the resident with the
opportunity to experience a true integrated medical and pharmacy system. The areas of focus will include pharmaceutical
care, drug information, formulary management, clinical program management, disease state management, and outcome
studies. The resident also will complete a research project suitable for publication. This program will incorporate communication and time management skills.
Jerry Wong, PharmD, MBA
Residency Director
Blue Cross and Blue Shield of Alabama
450 Riverchase Pkwy., East
Birmingham, AL 35244
(205) 220-6526
(205) 220-2939 (fax)
[email protected]
■■ BLUE CROSS AND BLUE SHIELD OF NEBRASKA
Accredited:
Being pursued
Length of Program:
12 months
Affiliation:
University of Nebraska Medical Center
Starting Date:
July 1
Estimated Stipend:
$36,000
Onsite Interview:
Required
Educational/Special Requirements: PharmD degree from an
ACPE-accredited college of pharmacy. Licensed pharmacist in
the United States or eligibility for licensure (successful candidate must be fully licensed at the start of residency). Application
requirements include: application, writing sample, 3 professional letters of recommendation, and onsite interview with
short formal presentation.
Fringe Benefits: 10 days paid vacation, holidays, health insurance, paid travel and registration to AMCP Educational
Conference and Annual Meeting
Special Features: The Blue Cross and Blue Shield of Nebraska
residency will offer exposure to various aspects of managed care
pharmacy through health plan and PBM exposure as well as
direct patient care through clinical rotations. The residency will
offer participation and exposure to the following areas: pharmacy benefit design, utilization management, formulary
management, pharmacy trend management, drug information,
health economic/outcomes research, pharmacy benefit management, disease state management quality assurance initiatives.
This unique program also allows the resident the option to
participate in teaching or didactic coursework through the
JMCP
September 2005
Vol. 11, No. 7
University of Nebraska Medical Center.
Lee Handke
Vice President of Pharmacy and Wellness
Blue Cross and Blue Shield of Nebraska
7261 Mercy Rd.
Omaha, NE 68180
(402) 398-3884
(402) 548-4683 (fax)
[email protected]
■■ BLUE SHIELD OF CALIFORNIA
Managed Care Pharmacy Systems
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
UCSF
Starting Date:
July 1
Estimated Stipend:
$46,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD degree from
an accredited school of pharmacy, completion of a pharmacy
practice residency or equivalent experience, 3 letters of recommendation, letter of intent, and onsite interview
Fringe Benefits: Health/dental/vision benefit; 20 days of paid
time off, including professional leave (with travel allowances);
and 9 holidays; no on-call responsibilities
Special Features: This residency instills the philosophy that
health care outcomes need to be considered from all relevant
perspectives (patient, provider, and payer). Residents participate in the development of drug policy, clinical guidelines,
pharmacy benefits, and population-based disease management,
pharmaceutical contracting support/analysis, and quality
improvement. This program teaches residents to conceptualize,
integrate, and transform accumulated experiences and knowledge into improved drug therapy for managed care patients.
Tara Abrams
Senior Clinical Pharmacist, Quality Improvement
Blue Shield of California Pharmacy Services
50 Beale St., 22nd Fl.
San Francisco, CA 94105
(415) 229-6424
(415) 229-6011 (fax)
[email protected]
www.amcp.org
■■ CAREMARK, INC.
Managed Care
Accredited:
Length of Program:
Number of Positions:
Affiliation:
AMCP/ASHP
12 months
3 (1 each in Maryland, Texas, and Arizona)
University of Maryland, University of
Arizona
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Yes
experience
Fringe Benefits: 2 weeks vacation, health insurance, free parking,
professional meetings, and other management and pharmaceutical industry experience
Special Features: Off-site rotations, university affiliation at MD
and AZ sites, ambulatory care clinic, disease management, industry experience
Melissa Jay
Caremark, Inc.
750 West John Carpenter Fwy.
Irving, TX 75039
(469) 524-5832
(469) 524-5858 (fax)
[email protected]
■■ CAREMARK, INC.
Managed Care Specialty-Analytics and Outcomes
Accredited:
No
Length of Program:
12 months
Affiliation:
University of Illinois at Chicago;
Midwestern University-Chicago College
of Pharmacy
Starting Date:
July 1
Estimated Stipend:
$38,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD with experiential
or internship-based experience in managed care/PBM industry
Fringe Benefits: Comprehensive medical, dental, and life
insurance plan; 2-week paid vacation; holidays; employee stock
purchase program; flexible spending program; travel budget
Special Features: Caremark is a leading pharmaceutical services
company, providing comprehensive drug benefit services to
approximately 24 million participants throughout the United
States. Caremark's clients include corporate health plans,
managed care organizations, insurance companies, unions,
government agencies, and other funded benefit plans. The
www.amcp.org
Analytics and Outcomes Residency will provide the resident a
unique opportunity to work on initiatives that foster pro-active
management of pharmaceutical and overall health care costs.
It offers the ability to work with large data sets and perform
various pharmaceutical cost analyses such as plan design modeling,
formulary analysis, and clinical outcomes. As part of a core sales
and account management team, the resident will have the
opportunity to interact directly with clients, consultants, and
various other benefit providers. While the focus is on analytics,
the resident will be exposed to various areas in pharmacy
benefit management such as clinical program development and
implementation, operations, sales, account management, clinical
sales support, marketing and communications, trade relations,
pharmaceutical services, and therapeutic services.
Anita Allemand
Director, Client Analytic Services, Inc.
Caremark, Inc.
2211 Sanders Rd.
Northbrook, IL 60062
(847) 559-3923
(847) 559-5475 (fax)
[email protected]
■■ CLINICAL PHARMACOLOGY SERVICES, INC.
Ambulatory Care/Clinical Research
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
June 1
Estimated Stipend:
$34,000
Onsite Interview:
Yes
Educational/Special Requirements: None
Fringe Benefits: Sponsorship to professional meeting and
Southwestern Residency Conference
Special Features: None
Daniel Buffington
Director
Clinical Pharmacology Services, Inc.
6285 E. Fowler Ave.
Tampa, FL 33617
(813) 983-1500
(813) 983-1501 (fax)
[email protected]
Web: http://www.cpshealth.com
Vol. 11, No. 7
September 2005
JMCP
■■ COVENTRY HEALTH CARE OF KANSAS, INC.
Pharmacy Benefits Management
Accredited:
No
Length of Program:
12 months
Affiliation:
University of Missouri–Kansas City
(UMKC)
Starting Date:
July 1
Estimated Stipend:
$32,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD degree from an
ACPE-accredited college of pharmacy, licensed pharmacist in
the United States or eligibility for licensure (successful candidate must be fully licensed at the start of residency), application, letter of intent, curriculum vitae, writing sample, official
transcripts, 3 professional letters of recommendation, onsite
interview with short formal presentation in Kansas City
Fringe Benefits: 2 weeks paid vacation; holidays; health insurance; paid travel and registration to AMCP Educational
Conference, AMCP Annual Meeting, and ASHP Midyear
Meeting
Special Features: Residency opportunities in all managed care
core competencies: pharmacy benefit management, utilization
management, formulary management, clinical consultation
service, drug information, new technologies assessment,
provider network relations, Medicare plan interventions,
disease state management, health and wellness initiatives, quality
assurance/improvement activities, marketing and sales,
contracting, health economics/outcomes, rotations with
pharmaceutical industry
Shawn Burke
Director of Pharmaceutical Services
Coventry Health Care of Kansas, Inc.
8320 Ward Pkwy.
Kansas City, MO 64114
(866) 795-3995
(866) 795-3992 (fax)
[email protected]
■■ DEPARTMENT OF VETERANS AFFAIRS MEDICAL CENTER
Pharmacy Practice-primary care emphasis
Accredited:
ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$35,000 plus benefits
Onsite Interview:
Yes
JMCP
September 2005
Vol. 11, No. 7
Educational/Special Requirements: PharmD or equivalent experience
Fringe Benefits: Vacation, paid holidays, sick days, and administrative time off for selected meetings
Special Features: ASHP-accredited pharmacy practice residency
with large emphasis in direct patient care in a primary care
setting. The pharmacy resident will work under a collaborative
practice agreement with a medical team to facilitate achievement
of therapeutic goals through evidence-based disease state
management. In addition, the resident will have learning
experiences in critical care, internal medicine, practice management, drug policy development, and education and teaching.
Upon completion of this residency program, the pharmacy
resident will have achieved a variety of advanced practice skills
that will enable the graduate to feel confident to function effectively
in multiple health care environments and roles.
Jo-Ann Caudill
Residency Program Director
Dept. of Veterans Affairs Medical Center
3200 Vine St.
Cincinnati, OH 45220
(513) 475-6322
(513) 475-6981 (fax)
[email protected]
■■ GROUP HEALTH COOPERATIVE
Managed Care Pharmacy Practice
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$40,000
Onsite Interview:
Yes
experience.
Fringe Benefits: Full medical coverage ($35/month paid by the
employee) and dental coverage (cost varies depending on the
plan chosen by employee) for the resident, 7 days of vacation;
paid registration and some fees to attend professional meetings
is provided
Special Features: The residents are trained in the role of the
pharmacist in the development and implementation of clinical
practice guidelines, formulary development and management,
and drug use policy development. In addition, residents are
trained to function as leaders in implementing pharmaceutical
care plans for specific patients in a managed care setting.
Jim Carlson
www.amcp.org
Director, Pharmacy Administration
Group Health Cooperative
12400 E. Marginal Way S.
Seattle, WA 98168
(206) 901-4420
(206) 901-4410 (fax)
[email protected]
Web: http://www.ghc.org/about_gh/employ/rxresidency.jhtml
■■ HARVARD VANGUARD MEDICAL ASSOCIATES
Pharmacy Practice with Emphasis in Managed Care
Accredited:
ASHP
Length of Program:
12 months
Affiliation:
Massachusetts College of Pharmacy
and Health Sciences
Starting Date:
July 1
Estimated Stipend:
$32,000
Onsite Interview:
Yes
Educational/Special Requirements: BS in pharmacy or
PharmD
Fringe Benefits: Comprehensive medical plan, 2 weeks paid
vacation
Special Features: Academic appointment: instructor of pharmacy practice
William McCloskey
Massachusetts College of Pharmacy & Health Sciences
179 Longwood Ave.
Boston, MA 02115
(617) 732-2167
(617) 732-2244 (fax)
[email protected]
Web: http://www.mcphs.edu
■■ HEALTHPARTNERS
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
Competitive
Onsite Interview:
Yes
Educational/Special Requirements: PharmD
Fringe Benefits: Health insurance, vacation, and holidays
Special Features: Travel/registration for 1 national meeting
www.amcp.org
Vyvy Vo
Clinical Pharmacy Program Manager
HealthPartners
8100 34th Ave. South
PO Box 1309
Minneapolis, MN 55440
(952) 967-5133
(952) 883-5875 (fax)
[email protected]
■■ HENRY FORD HEALTH SYSTEM
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
Competitive
Onsite Interview:
Yes
Educational/Special Requirements: PharmD, pharmacy practice residency desirable
Fringe Benefits: Health care, 2 weeks paid vacation, travel to
1 meeting
Special Features: The resident will design system enhancements and participate in ongoing utilization management, disease management, and compliance intervention programs.
He/she will participate in formulary management and design
clinical indications of effectiveness for guidelines. The resident
will formulate and answer a research question, using scientific
principles, with a strong emphasis on outcomes, pharmacoeconomics, and quality-of-life research. Experiences will
include exposure to our HMO (Health Alliance Plan) and
involvement with the Center for Clinical Effectiveness, health
system studies, quality improvement center, and clinical
pharmacy services. Direct patient care responsibilities in one of
our ambulatory clinics will be ongoing throughout the year.
Vanita Pindolia
VP, Pharmacy Care Management
Henry Ford Health System
30100 Telegraph Rd., Suite 200
Bingham Farms, MI
48025
(248) 723-0206
(248) 642-6094 (fax)
[email protected]
Vol. 11, No. 7
September 2005
JMCP
■■ HORIZON NJ HEALTH
Accredited:
No
Length of Program:
12 months
Affiliation:
Horizon BC
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Yes
experience; eligibility for New Jersey state licensure
Fringe Benefits: Paid vacation, full medical/dental/retirement
benefits
Special Features: Medicaid managed care HMO, unique focus
on government programs, pharmacy case management, formulary and disease state management, development of clinical
policies, outcomes research, assist with PharmD student oversight, ambulatory care experience, professional development
courses, attendance to at least 1 national conference
Samuel Currie
Director, Clinical Pharmacy Programs
Horizon NJ Health
210 Silvia St.
West Trenton, NJ 08628
(609) 538-0700
(609) 538-1698 (fax)
[email protected]
■■ HUMANA INC.
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$37,000
Onsite Interview:
Yes
Educational/Special Requirements: Graduate of an accredited
school of pharmacy, PharmD preferred, minimum GPA of 3.2
(on a 4.0 scale), and eligibility for pharmacy licensure
Fringe Benefits: Health, dental, 401(k), 3 weeks vacation,
2 floating holidays, relocation allowance, and travel expenses
paid for 2 professional meetings
Special Features: Humana Inc. has a broad. range of programming within the pharmacy management department. The residency will offer exposure to various aspects of managed care
pharmacy as well as direct patient care through clinical rotations at the Veterans Affairs hospital. The residency will offer
JMCP
September 2005
Vol. 11, No. 7
participation in corporate pharmacy and therapeutics committees, manufacturer relations, pharmaceutical contracting and
rebating, pharmacy benefit design, consumer relations, legislative
and drug policy issues relating to managed care and health benefit design, ePharmacy initiatives, and outcomes analysis.
Jane Stacy
Clinical Advisor
Humana Inc.
500 W. Main St., 16th Fl.
Louisville, KY 40202
(502) 580-1591
(502) 508-1591 (fax)
[email protected]
■■ IBA HEALTH PLANS/BLUE CARE NETWORK
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
Ferris State University/Kalamazoo
Center for Medical Studies/Pfizer
Starting Date:
July 1
Estimated Stipend:
$32,000
Onsite Interview:
Yes
Fringe Benefits: Health/dental, insurance, 2 weeks vacation,
paid holidays
Special Features: Travel/registration reimbursement for 1 national
conference and the Great Lakes Residency Conference
(pharmacy residents conference); no weekend or evening shifts
Teresa Klepser
Residency Director
MSU-KCMS/Ferris
1000 Oakland Dr.
Kalamazoo, MI 49008
(269) 337-6392
(269) 337-4474 (fax)
[email protected]
■■ INTERMOUNTAIN HEALTH CARE (IHC) HEALTH PLANS
Accredited:
Anticipated, July 2005
Length of Program:
1 year
Affiliation:
None
Starting Date:
June 30
www.amcp.org
Estimated Stipend:
$40,000
Onsite Interview:
Required
Educational/Special Requirements: Candidates must graduate
from an ACPE-accredited pharmacy program with a doctor of
pharmacy degree with a minimum GPA of 3.0 on a 4.0 scale.
Candidate must obtain Utah pharmacist licensure within the
first 60 days of the program.
Fringe Benefits: Health, dental, and life insurance; staff
discounts; 10 days of vacation leave; and 10 holidays
Special Features: Travel benefits to either the ASHP Midyear
Clinical Meeting or the AMCP Educational Meeting and the
Western States Conference
Jeffrey Dunn
Formulary and Contract Manager
IHC Health Plans
4646 W. Lake Park Blvd., Suite N3
Salt Lake City, UT 84120
(801) 442-7984
(801) 442-3006 (fax)
[email protected]
■■ KAISER PERMANENTE
Managed Care
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$41,220 (2005-2006)
Onsite Interview:
Yes
Educational/Special Requirements: This is a specialty residency in managed care with emphasis on pharmacy practice
management; BS/PharmD
Fringe Benefits: Medical benefits, 10 days sick leave/vacation,
5 holidays, $500 for national meeting plus expenses paid for
Western States Residency Conference
Special Features: Emphasis in administration/practice
management
Susan Downard
Area Pharmacy Manager
Kaiser Permanente
16601 E. Centretech Pkwy.
Aurora, CO 80011
(303) 326-6764
(303) 739-3574 (fax)
[email protected]
www.amcp.org
Mental Health (Specialty)
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 5
Estimated Stipend:
$41,220 (2005-2006)
Onsite Interview:
Yes
Educational/Special Requirements: PharmD required;
completion of pharmacy practice residency or possession of
equivalent clinical experience preferred
Fringe Benefits: Health insurance, holidays, travel assistance
Special Features: Group-model HMO setting
Daniel Dugan
Clinical Pharmacy Specialist in Mental Health
Kaiser Permanente of Colorado
16601 E Centretech Pkwy.
Aurora, CO 80011
(303) 467-5776
(303) 467-5805 (fax)
[email protected]
Primary Care (Specialty)
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 5
Estimated Stipend:
$41,220 (2005-2006)
Onsite Interview:
Yes
Educational/Special Requirements: PharmD required;
completion of pharmacy practice residency or possession of
equivalent clinical experience preferred
Fringe Benefits: Health insurance, holidays, travel assistance
Special Features: Group-model HMO setting
Rachana Patel
Clinical Pharmacy Specialist
Kaiser Permanente
1375 East 20th Ave.
Denver, CO 80205
(303) 764-4479
(303) 861-3668 (fax)
[email protected]
Vol. 11, No. 7
September 2005
JMCP
[This listing has not been not updated; contact the program for
current information.]
■■ KAISER PERMANENTE MEDICAL CARE PROGRAM—
CALIFORNIA
Pharmacy Practice and Drug Information Practice
Accredited:
ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Yes
Educational/Special Requirements: Applicants must be a
graduate of an accredited college of pharmacy and be licensed,
or be eligible for licensure, in California. Good communication
skills required. Resident will develop a project with targeted
care outcomes and present at the annual Western States
Conference.
Fringe Benefits: Medical, dental, and optical insurance; holidays;
vacation/sick leave
Special Features: Hospital and ambulatory care experiences in
the nation’s largest health maintenance organization, preventative
and disease state management in an integrated managed care
setting
Elaine Watanabe
Pharmacy Services Manager, Recruitment
Kaiser Permanente Pharmacy Operations Services,
California Division
9521 Dalen St.
Downey, CA 90242
(714) 796-4809
(714) 796-4826 (fax)
[email protected]
INLAND EMPIRE SERVICE AREA
General Pharmacy Practice
Accredited:
ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Yes
college of pharmacy and be licensed, or be eligibile for licensure,
in California; pharmacy school transcript; 3 letters of recommendation; letter of intent; and curriculum vitae
JMCP
September 2005
Vol. 11, No. 7
Fringe Benefits: 2 weeks paid vacation; health benefits, including
dental/optical; paid holidays; office space; reimbursement for
off-site experiences
Special Features: Kaiser Permanente is the nation's largest
nonprofit health plan, serving more than 8.4 million members
in 9 states. The Inland Empire Service Area consists of 2 medical
centers and 14 satellite medical offices for 598,000 members in
the San Bernardino and Riverside counties of southern
California.
Patricia Gray
Clinical Operations Manager
Kaiser Permanente Inland Empire Service Area
9310 Sierra Ave.
Fontana, CA 92335
(909) 427-3838
(909) 427-3830 (fax)
[email protected]
LOS ANGELES MEDICAL CENTER
Pharmacy Practice
Accredited:
ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Yes
Conference.
Fringe Benefits: Medical, dental, optical insurance; holidays;
vacation/sick leave
Special Features: The Kaiser Permanente Los Angeles Medical
Center is the tertiary care center for Kaiser Permanente in
southern California and provides comprehensive inpatient, outpatient, and ambulatory care services to Kaiser Permanente
members. This residency program provides development and
training for recently graduated pharmacists, with an emphasis
on pharmaceutical care and leadership to a diverse community.
This program will allow the residents to become familiar with
pharmacy practice in an integrated health care program.
Steve Litsey
Pharmacy Leader-Metro Service Area
Kaiser Permanente Pharmacy Operations Services
www.amcp.org
TRI-CENTRAL SERVICE AREA
1515 N. Vermont Ave., Suite 237
Los Angeles, CA 90027
(323) 783-8306
(323) 783-7609 (fax)
[email protected]
METRO LOS ANGELES
Managed Care
Accredited:
Yes
Length of Program:
12
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
35,000
Onsite Interview:
Yes
Conference.
Fringe Benefits: Paid insurance (medical, dental, vision), holidays, vacation/sick days
Special Features: The Kaiser Permanente Los Angeles Medical
Center is the tertiary care center for Kaiser Permanente in
southern California and provides comprehensive inpatient, outpatient, and ambulatory care services to Kaiser Permanente
members. This residency program provides development and
training for recently graduated pharmacists, with an emphasis
on pharmaceutical care and leadership to a diverse community.
This program will allow the residents to become familiar with
managed care pharmacy practice in an integrated health care
program.
Elizabeth Oyekan, PharmD
Residency Progam Coordinator
Kaiser Permanente Medical Care Program—Metro Los Angeles
Pharmacy Operations
1515 N. Vermont Ave., Suite 237
(323) 783-8306
(323) 783-7609 (fax)
[email protected]
www.amcp.org
Managed Care Organization
Accredited:
ASHP
Length of Program:
1 year
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Yes
Educational/Special Requirements: Recent graduate of an
accredited school of pharmacy and eligibility for California
licensure, college transcript(s), letter of intent, 3 letters of recommendation, and curriculum vitae
Fringe Benefits: 2 weeks paid vacation; 6 paid holidays; sick
leave; health benefits, including dental/optical (also dependents); uniforms; office space; and reimbursement for off-site
experiences
Special Features: The program provides the resident with an
opportunity to participate in the provision of health care in a
large managed care environment. Resident will gain experience
in various practice areas: acute care, ambulatory care, physician
drug education, drug information, outpatient pharmacy, and
practice management; ambulatory care practice includes anticoagulation, heart failure, hypertension, integrated CVD, oncology,
etc. Residents completing our program will have a strong
foundation for future pharmacy practice.
John Sie
Pharmacy Practice Residency Coordinator, Ambulatory Care
Clinical Pharmacy Services Supervisor
Kaiser Permanente Medical Care Program Tri-Central Service
Area, Pharmacy Operations
1011 Baldwin Park Blvd.
Baldwin Park, CA 90706
(626) 851-5602
(626) 851-5813 (fax)
[email protected]
WEST LOS ANGELES
Pharmacy Practice
Accredited:
Length of Program:
Number of Positions:
Affiliation:
Application Deadline:
Starting Date:
Estimated Stipend:
Onsite Interview:
Vol. 11, No. 7
September 2005
ASHP
12 months
1
None
January 15
July 1
$35,000
Yes
JMCP
Conference
Fringe Benefits: Medical, dental, and optical insurance; holidays;
vacation/sick leave
Special Features: Hospital and ambulatory care experiences in
the nation’s largest integrated care organization, preventative
and disease state management in an integrated managed care
setting. Flexible program molded to the resident’s interests.
Michael Cinnamond, PharmD
Inpatient Pharmacy Director, Residency Program Director
Kaiser Permanente Medical Care Program at West Los Angeles
6041 Cadillac Ave., B 310
(323) 857-2044
(323) 857-2870 (fax)
[email protected]
■■ KAISER PERMANENTE MID-ATLANTIC STATES REGION
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 5
Estimated Stipend:
$39,500
Onsite Interview:
Yes
Educational/Special Requirements: PharmD preferred, pharmacy
licensure eligibility in DC, Maryland, or Virginia
Fringe Benefits: Medical benefits, selected holidays, sick, vacation and education leave for Midyear Clinical Meeting and
Eastern States Conference
Special Features: In addition to helping develop new and
innovative programs, the resident will participate on the
Pharmacy & Therapeutics Committee, teach patient education
classes, provide pharmacy staff continuing education, assist in
educating pharmacy students, and complete residency projects and
presentations.
Katrin Fulginiti
Director, Managed Care Pharmacy Practice Residency Program
Kaiser Permanente-Mid-Atlantic
12201 Plum Orchard Dr.
Silver Spring, MD 20904
(301) 572-3330
(301) 572-3399 (fax)
JMCP
September 2005
Vol. 11, No. 7
[email protected]
Web: http://www.kaiserpermanente.org
■■ KAISER PERMANENTE OF CALIFORNIA—
SPECIALTY RESIDENCY IN DRUG INFORMATION
Medical Care/Drug Information
Accredited:
ASHP
Length of Program:
12 months
Affiliation:
None
Application Deadline: March 31
Starting Date:
July 1
Estimated Stipend:
$70,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD, pharmacy
practice residency program or equivalent experience, excellent
communication skills, eligibility for California licensure, 3 letters
of recommendation
Fringe Benefits: Medical, dental, optical insurance; 10 days
time off; attendance at 1 pharmacy conference
Special Features: None
Mirta Millares
Kaiser Permanente
12254 Bellflower Blvd., Suite 106
Downey, CA 90242
(562) 658-3630
(562) 658-3758 (fax)
[email protected]
■■ KAISER PERMANENTE OF COLORADO
Cardiology/Managed Care
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$41,220
Onsite Interview:
Yes
Educational/Special Requirements: PharmD required; prior
pharmacy practice residency or equivalent preferred
Fringe Benefits: Health benefits, travel support to ASHP
Midyear meeting and Western States Residency Conference
Special Features: This cardiology specialty residency provides
unique opportunities to acquire advanced knowledge and skills
in ambulatory management of anticoagulation, cardiac risk
reduction, and heart failure in a managed care setting.
Numerous educational opportunities exist. Experience in evaluating clinical and financial outcomes will be emphasized.
www.amcp.org
Completion of a residency research project required.
Anne Denham, PharmD
Clinical Pharmacy Cardiac Risk Service
Aurora, CO 80011
(303) 326-7663
(303) 326-7670 (fax)
[email protected]
Mental Health/Managed Care
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$41,220
Onsite Interview:
Yes
Educational/Special Requirements: PharmD required, prior
pharmacy practice residency or equivalent practical experience
preferred
Midyear Clinical Meeting and Western States Residency
Conference
Special Features: Experiences include inpatient and outpatient
psychiatric pharmacy practice sites among a full range of behavioral health conditions as well as neurology and chemical
dependence. Evaluation of clinical outcomes is emphasized and
completion of residency research is required.
Daniel Dugan
Clinical Pharmacy Specialist in Mental Health
Aurora, CO 80011
(303) 467-5776
(303) 467-5805 (fax)
[email protected]
Primary Care/Managed Care
Accredited:
No
Length of Program:
12 months
Affiliation:
None
www.amcp.org
Starting Date:
July 1
Estimated Stipend:
$41,220
Onsite Interview:
Yes
Educational/Special Requirements: PharmD required; prior
pharmacy practice residency or equivalent experience preferred
Midyear Clinical Meeting and Western States Residency
Conference
Special Features: This primary care/managed care specialty
residency provides the unique opportunity to acquire advanced
knowledge and skills in ambulatory care pharmacotherapy in a
managed care setting. Practice sites include primary care clinics,
centralized services (anti-coagulation, cardiac risk), and specialty
services (diabetes, cardiology, infectious disease, mental health,
nephrology). Residents participate in formulary management,
pharmacoeconomic evaluations, and target drug programs.
Numerous educational opportunities exist. Experience in
evaluating clinical and financial outcomes will be emphasized.
Completion of residency research project required.
Rachana Patel
Kaiser Permanente
1375 East 20th Ave.
Denver, CO 80205
(303) 764-4479
(303) 861-3668 (fax)
[email protected]
■■ KAISER PERMANENTE OF GEORGIA
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$34,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD preferred; eligibility
for Georgia licensure
Fringe Benefits: Medical/dental/vision insurance, holidays,
vacation/ sick leave
Special Features: Ambulatory care, acute care, drug information,
and administration
Suzanne Booth
Residency Program Coordinator
Kaiser Permanente
200 Crescent Centre Pkwy.
Tucker, GA 30083
Vol. 11, No. 7
September 2005
JMCP
(770) 496-3653
(770) 496-3487 (fax)
[email protected]
■■ KELSEY-SEYBOLD CLINIC/UNIVERSITY OF HOUSTON
Pharmacy Practice with an Emphasis on Managed Care
Accredited:
ASHP/AMCP
Length of Program:
1 year
Affiliation:
University of Houston
Starting Date:
July 1
Estimated Stipend:
30,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD or equivalent experience
Fringe Benefits: State of Texas benefits, 10 days vacation, support
to attend AMCP, ASHP, TSHP, and either Alcalde or Mid-West
Residency Conference
Special Features: Ambulatory-care-focused administrative and
clinical responsibilities, collaborations with other UH
residents
Sarah Lake-Wallace, PharmD
Kelsey-Seybold Clinic
Pharmacy Administration
8900 Lakes at 610 Dr.
Houston, TX 77054
(713) 442-6248
(713) 442-5253 (fax)
[email protected]
■■ MEDCO HEALTH SOLUTIONS, INC.
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
None
Starting Date:
June 26
Estimated Stipend:
$34,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD degree, eligibility
for New Jersey licensure, managed care rotation or experience
preferred
Fringe Benefits: Complete medical coverage, paid holidays
and vacation
Special Features: Professional development trainings, travel
and attendance at educational meetings
JMCP
September 2005
Vol. 11, No. 7
Doris Fishman
Senior Director, Clinical Therapeutics
Medco Health Solutions, Inc.
100 Parsons Pond Dr.
Mail Stop F2-3
Franklin Lakes, NJ 07417
(201) 269-6159
(201) 260-1035 (fax)
[email protected]
■■ NOVARTIS PHARMACEUTICALS CORPORATION
Health Economics & Outcomes Research Fellowship
Accredited:
No
Length of Program:
2 years (Rutgers site, flexible)
Affiliation:
Duke University; Scott & White
Health Plan/University of Texas at
Austin; Rutgers University
Application Deadline: December 31
Starting Date:
July 1
Estimated Stipend:
$35,000-$44,000
Onsite Interview:
Yes
Educational/Special Requirements: Advanced degree in
health services research, public health, health policy, pharmacy,
economics, medicine, or other related areas, with some experience in outcomes research
Fringe Benefits: Medical insurance, vacation
Special Features: The fellows will gain familiarity with outcomes
research principles/application and experience in designing
research studies that examine economic, clinical, and humanistic
outcomes. The first year is spent at an academic/managed care
institution and the second year with Novartis’s Health Economics
& Outcomes Research Department.
Feride Frech
Director, Health Economics & Outcomes Research
Novartis Pharmaceuticals Corporation
One Health Plaza
East Hanover, NJ 07936-1080
(862) 778-5094
(973) 781-3018 (fax)
[email protected]
■■ OPTIMA HEALTH PLAN/ SENTARA HEALTHCARE
Managed Care
Accredited:
Not at this time
Length of Program:
12 months
Affiliation:
None
www.amcp.org
Starting Date:
July 1
Estimated Stipend:
$31,720
Onsite Interview:
Yes
Educational/Special Requirements: PharmD from an accredited
school of pharmacy or equivalent experience, onsite interview,
and eligibility for Virginia licensure
Fringe Benefits: 2 weeks of paid vacation and uninterrupted
stipend during minor illness, a health insurance plan, travel
assistance for continuing education events and other professional activities
Special Features: Resident will have exposure to an integrated
health care system and system-wide pharmacy services, including
a drug information center. The resident will have the opportunity
to precept students from Virginia Commonwealth University and
Hampton University schools of pharmacy and work with medical
residents from the Eastern Virginia Medical School.
Elizabeth Brusig, PharmD
Clinical Pharmacist
Optima Health Plan
4417 Corporation Lane
Virginia Beach, VA 23464
(757) 552-7519
(757) 552-7516 (fax)
[email protected]
■■ ORTHO-MCNEIL JANSSEN SCIENTIFIC AFFAIRS, LLC
Pharmaceutical Industry, Department of Outcomes Research
Accredited:
No
Length of Program:
24 months
Affiliation:
Thomas Jefferson University
Starting Date:
June 2006 (flexible)
Estimated Stipend:
Competitive
Onsite Interview:
Yes
Educational/Special Requirements: Contact program
Fringe Benefits: Contact program
Special Features: Fellows spend the first year of the program
working on outcomes research projects in an academic setting
and the second year in the pharmaceutical industry. Fellows
have the opportunity to take coursework in biostatistics,
epidemiology, economics, and other outcomes-related subjects.
Julie Locklear, PharmD, MBA
Associate Director, Outcomes Research
Ortho-McNeil Janssen Scientific Affairs, LLC
1125 Trenton-Harbourton Rd.
(609) 730-3664
(609) 730-2556 (fax)
[email protected]
■■ ORTHO-MCNEIL JANSSEN SCIENTIFIC AFFAIRS, LLC
Drug Information
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
Competitive
Onsite Interview:
Yes
Educational/Special Requirements: Contact program
Fringe Benefits: Contact program
Special Features: Contact program
Will Zachok
Associate Director, Medical Communications
Ortho-McNeil Janssen Scientific Affairs, LLC
1000 Route 202 S
Raritan, NJ 08869
(908) 218-6283
(908) 722-6402 (fax)
[email protected]
■■ OUTCOMES PHARMACEUTICAL HEALTH CARE
Medication Therapy Management Firm
Accredited:
ASHP/AMCP
Length of Program:
1 year
Affiliation:
University of Iowa
Starting Date:
July 5
Estimated Stipend:
$33,000
Onsite Interview:
Yes
Educational/Special Requirements: Candidates for residency
must possess a doctor of pharmacy degree from a school of pharmacy accredited by the American Council on Pharmaceutical
Education (ACPE) and be eligible for licensure in Iowa.
Fringe Benefits: For additional information, contact the
Outcomes office at (515) 237-0001 or [email protected].
Special Features: Outcomes’ clinical services residency is a
12-month managed care systems residency program. This residency offers a unique learning experience on how pharmacists
can impact the quality and rising cost of health care by providing
patient-oriented, care-based services. Residency activities
include account management, marketing and communications,
pharmacoeconomic data collection and analysis, claims assessment, development and implementation of disease management
www.amcp.org
Vol. 11, No. 7
September 2005
JMCP
programs, development and administration of research projects,
and training of pharmacists and support staff. The resident also
serves as a preceptor to PharmD clerkship students.
Patty Kumbera
Chief Operations Officer
Outcomes Pharmaceutical Health Care
601 E. Locust, Suite 200
Des Moines, IA 50309
(515) 237-0001
(515) 237-0002 (fax)
[email protected]
Web: http://www.getoutcomes.com
■■ PHARMACARE
Pharmacy Benefits Management
Accredited:
No
Length of Program:
12 months
Affiliation:
University of Pittsburgh School of
Pharmacy
Starting Date:
July 1
Estimated Stipend:
$31,000
Onsite Interview:
Yes
Fringe Benefits: Comprehensive benefits package, no weekends/holidays, attend professional managed care meetings
Special Features: The University of Pittsburgh School of
Pharmacy and EHS, one of the nation’s largest pharmacy chainbased prescription management firms, offers an opportunity to
practice in a dynamic PBM environment and gain a clinical and
administrative perspective in managed pharmacy benefit plans
for a wide variety of clients. Multifaceted experience will include
DUR criteria development, clinical intervention activities, P&T
activities, clinical systems development, and new business development/client services/ marketing support.
Teddi Gianangeli
Senior Clinical Management Specialist
Pharmacare
620 Epsilon Dr.
Pittsburgh, PA 15238
(412) 967-2300 ext. 5045
(412) 968-2676 (fax)
[email protected]
JMCP
September 2005
Vol. 11, No. 7
■■ PHARMACEUTICAL CARE NETWORK
Accredited:
No
Length of Program:
12 months
Affiliation:
None
Starting Date:
July 1
Estimated Stipend:
$37,500
Onsite Interview:
Yes
college of pharmacy and licensed, or be eligible for licensure, in
California
Fringe Benefits: Paid vacation and sick leave, health/dental/
vision benefits, and educational support to attend professional
meetings
Special Features: This residency will provide training in formulary
management, provider drug therapy education, drug benefit
design, outcomes analysis, and prior authorization. The resident
will participate in drug utilization review, P&T committee presentations, and interface with pharmacists from a variety of professional fields. The resident will learn to use PCN’s MedIntelligence
software to identify drug therapy problems and make appropriate
interventions. This program includes rotations in direct patient care
and drug information and an opportunity to rotate through the
California Pharmacists Association.
Philip Parsatoon
Vice President, Professional Services
Pharmaceutical Care Network
9343 Tech Center Dr., Suite 200
Sacramento, CA 95826-2563
(916) 361-4450
(916) 414-4650 (fax)
[email protected]
■■ PHARMACY MANAGEMENT CONSULTANTS/
OKLAHOMA UNIVERSITY
Accredited:
ASHP-Managed Care Pharmacy
Systems
Length of Program:
12 months
Affiliation:
Oklahoma University College of
Pharmacy
Starting Date:
July 1
Estimated Stipend:
$35,000
Onsite Interview:
Preferred
www.amcp.org
Educational/Special Requirements: Pharmacy degree
Fringe Benefits: 10 days vacation, health and dental coverage,
faculty appointment at the level of clinical instructor, sponsorship to at least 1 professional meeting
Special Features: Evaluate health plan formularies, develop
clinical guidelines, evaluate retrospective and prospective DUR
criteria, develop educational intervention programs, serve as
preceptor for pharmacy students; program includes experience
working with dual Medicare/Medicaid eligibles and with the
state PDP plan.
Elgene Jacobs
Pharmacy Management Consultants
1122 N.E. 13th St.
ORI W-4403
Oklahoma City, OK 73117
(405) 271-9039, ext. 47354
(405) 271-2615 (fax)
[email protected]
■■ PRESCRIPTION SOLUTIONS
Managed Care
Accredited:
ASHP
Length of Program:
12 months
Number of Positions: 1 or 2
Affiliation:
None
Starting Date:
July 5
Estimated Stipend:
$40,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD, license to practice
in California
Fringe Benefits: 2 weeks paid vacation, paid holidays, sick days,
travel reimbursement
Special Features: Drug therapy management, developing clinical
guidelines for appropriate drug use, formulary management,
legal and regulatory affairs; professional meetings and seminars:
required to attend the ASHP Midyear Clinical Meeting, CSHP
Seminar, AMCP Educational Conference, and Western States
Conference
Alex Gilderman
Director, Clinical Pharmacy Services
c/o Prescription Solutions
3515 Harbor Blvd.
Costa Mesa, CA 92626
(714) 825-6768
(714) 825-3742 (fax)
[email protected]
www.amcp.org
■■ RUTGERS UNIVERSITY/HORIZON BLUE CROSS
BLUE SHIELD OF NEW JERSEY
Managed Care Organization
Accredited:
No
Length of Program:
12 months
Affiliation:
Ernest Mario School of Pharmacy,
Rutgers University, State University of
New Jersey
Starting Date:
July 1
Estimated Stipend:
$30,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD, eligibility for
New Jersey state license
Fringe Benefits: Full medical coverage, dental and retirement
benefits
Special Features: No weekend or staffing requirements, teaching and preceptoring, mandatory, graduate courses offered at
Rutgers University, industry and PBM perspectives of managed
care
Saira A. Jan
Associate Professor, Rutgers University
Associate Director, Pharmacy Management
Horizon Blue Cross Blue Shield of NJ
Three Penn Plaza East, PP-13Q
Newark, NJ 07105
(973) 466-4575
(973) 466-6266 (fax)
[email protected]
■■ SCOTT & WHITE
Pharmacy Practice Managed Care
Accredited:
No
Length of Program:
1 year
Affiliation:
University of Texas at Austin
Starting Date:
July 1
Estimated Stipend:
$30,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD, eligibility for
Texas state license
Fringe Benefits: Full medical coverage and dental
Special Features: Minimal weekend staffing requirements;
teaching and preceptoring, if desired; industry and PBM
perspectives of managed care; marketing and sales exposure
Vol. 11, No. 7
September 2005
JMCP
(520) 629-4700 (fax)
[email protected]
John Jackimiec
Director of Pharmacy Managed Care
Scott & White
2601 Thornton Lane, Suite A
Temple, TX 76502
(254) 742-3144
(254) 742-3109 (fax)
[email protected]
■■ SOUTHERN ARIZONA VA HEALTH CARE SYSTEM
Pharmacy Practice
Accredited:
ASHP since 1982
Length of Program:
12 months
Affiliation:
University of Arizona
Starting Date:
July 1
Estimated Stipend:
$33,000 plus benefits
Onsite Interview:
Interview required, onsite preferred
Educational/Special Requirements: Curriculum vitae, 3 letters
of recommendation, college transcripts, 1 writing sample, U.S.
citizenship
Fringe Benefits: Residents accumulate 13 days of paid vacation
time during the residency year. Sick leave is accrued at the rate
of 4 hours every 2 weeks. Educational leave is provided to
attend the Arizona Society of Health-Systems Pharmacists
Annual Meeting, the ASHP Midyear Clinical Meeting, and the
Western States Conference for Pharmacy Residents, Fellows,
and Preceptors. Travel funds are available to offset some of the
expenses for these educational meetings. Residents have access
to the same medical and dental insurance plans that are offered
to full-time employees of the VA. These include a wide range of
HMO and PPO health plans. Tucson also has a wide range of
outdoor activities and excellent weather year round.
Special Features: This comprehensive residency includes a
balance of inpatient and outpatient clinical pharmacy experiences. Residents work in the areas of internal medicine, cardiology, neurology clinics, primary care, geriatrics, hospice,
surgery/nutritional support, mental health, practice management,
DUE, Pharmacy and Therapeutics Committee activities, drug
literature evaluation, drug policy development, teaching,
research, and enhancing communication abilities. Two months
of elective experience are available.
William Jones, MS, RPh
Southern Arizona Veterans Administration Health Care System
3601 South Sixth Ave.
Pharmacy Service 5-119
Tucson, AZ 85723
(520) 792-1450, ext. 6721
JMCP
September 2005
Vol. 11, No. 7
■■ UNITED DRUGS
Managed Care
Accredited:
Not at this time
Length of Program:
1 year
Affiliation:
AMCP
Application Deadline: June 15
Starting Date:
September 1
Estimated Stipend:
$33,000
Onsite Interview:
Yes
Educational/Special Requirements: Completion from an
accredited pharmacy school, PharmD preferred; ability to work
independently
Fringe Benefits: Understanding of managed care from the
perspective of a PBM, product development opportunities, and
ability to live in Arizona for a year
Special Features: Work hours negotiable
Jean Brown
Director of Clinical Services
United Drugs
7227 North 16th St., Suite 160
Phoenix, AZ 85020-5256
(602) 678-1179, ext. 229
(602) 678-0772 (fax)
[email protected]
[This listing has not been not updated: contact the program for
■■ UNIVERSITY AT BUFFALO
Ambulatory Care
Accredited:
Pending
Length of Program:
1 year
Affiliation:
University
Application Deadline: October 1
Starting Date:
January 1
Estimated Stipend:
Contact program
Onsite Interview:
Required
experience required
Fringe Benefits: Health, dental, vision insurance; paid vacation;
conference travel funds
Special Features: Unique experience designed to further refine
skills in pharmaceutical care in addition to developing skills in
www.amcp.org
program development and personnel and resource management. Resident is involved in coordination of clinical activities
in a high-volume lobby-based pharmacy and a health clinic
based ambulatory care pharmacy. The resident will participate
in medication histories, adherence counseling, and education
programs. Development and implementation of disease management initiatives, patient education, medical informatics and
supervision of PharmD students are also significant aspects of
the program. This residency will allow ample latitude for the
resident to explore interests and further develop skills as a practitioner. The resident will also be appointed as a clinical instructor at
the University at Buffalo, School of Pharmacy and Pharmaceutical
Sciences.
Gene Morse, PharmD
University at Buffalo
School of Pharmacy and Pharmaceutical Sciences
311 Hochstetter Hall
Buffalo, NY 14260
(716) 645-2828
(716) 645-2886 (fax)
[email protected]
■■ UNIVERSITY OF ILLINOIS AT CHICAGO
& WALGREENS HEALTH INITIATIVES
Fellowship— Outcomes Research
Accredited:
No
Length of Program:
2 years
Affiliation:
University of Illinois & Walgreens
Health Initiatives
Starting Date:
July 1
Estimated Stipend:
$36,000
Onsite Interview:
Educational/Special Requirements: Applicants should have a
PharmD or MD (or equivalent) and have completed a pharmacy practice or managed care residency
Fringe Benefits: _________________________________
Special Features: This is a 2-year fellowship jointly offered by
Walgreens Health Initiatives and the Center for
Pharmacoeconomic Research at the University of Illinois at
Chicago. The aim of the program is to train clinical pharmacists
to conduct research in drug therapy outcomes and pharmacoeconomics in the managed care setting. Knowledge and
experience will be gained in the use of research tools to
evaluate economic, humanistic, and clinical outcomes of drug
therapy. Presentation and publication of research findings in
peer-reviewed venues is expected. The fellowship is designed to
facilitate career opportunities in managed care, health provider
organizations, consulting, academia, or the pharmaceutical
www.amcp.org
industry.
Glen Schumock, PharmD, MBA
University of Illinois at Chicago
Center for Pharmacoeconomic Research
833 S. Wood St. (MC 886)
Chicago, IL 60612
(312) 996-7961
[email protected]
■■ UNIVERSITY OF MARYLAND SCHOOL OF PHARMACY/
CAREFIRST BLUECROSS BLUESHIELD
Accredited:
No
Length of Program:
12 months
Affiliation:
University of Maryland
Starting Date:
July 1
Estimated Stipend:
31,500
Onsite Interview:
Yes
Educational/Special Requirements: Graduate degree in pharmacy
Fringe Benefits: Health insurance, parking, support for national
meeting attendance and poster presentation
Special Features: Appointment as a clinical instructor at the
University of Maryland School of Pharmacy, ambulatory care
clinics at HMO, office with computer/references at managed
care organization
Catherine Cooke
Clinical Assistant Professor
University of Maryland School of Pharmacy
5106 Bonnie Branch Rd.
Ellicott City, MD 21043
(410) 480-5012
(410) 480-5296 (fax)
[email protected]
■■ UNIVERSITY OF TEXAS MEDICAL BRANCH
CORRECTIONAL MANAGED CARE
Pharmacy Practice with an Emphasis in Managed Care
Accredited:
ASHP
Length of Program:
12 months
Affiliation:
UTMB
Starting Date:
July 1
Estimated Stipend:
Contact program
Onsite Interview:
Yes
Educational/Special Requirements: Pharmacy degree from
Vol. 11, No. 7
September 2005
JMCP
accredited college of pharmacy, Texas license or eligibility for
Texas licensure
Fringe Benefits: UTMB is an AA/EO employer; closed major
holidays and weekends; generous vacation, holiday, and sick
leave; competitive benefits
Special Features: Program strengths include automated technology, telemedicine technology, ambulatory care and managed
care
Stephanie Zepeda
Assistant Director of Pharmacy
UTMB Correctional Managed Care
2400 Ave. I
Huntsville, TX 77340
(936) 437-5363
(936) 437-5311 (fax)
[email protected]
■■ UPMC HEALTH PLAN
Managed Care
Accredited:
No
Length of Program:
12 months
Affiliation:
UPMC Health System
Application Deadline: May 20
Starting Date:
July 1
Estimated Stipend:
$32,000
Onsite Interview:
Preferred
Educational/Special Requirements: Applicants must have
completed a doctor of pharmacy degree and be eligible for
Pennsylvania licensure. A pharmacy practice residency is recommended but not required for consideration. Please include
letter of intent, curriculum vitae, list of references (3) with contact information, and 3 letters of recommendation.
Fringe Benefits: Health, dental, eye care, life, and disability
insurance is available; vacation time is allotted and professional
travel and stipend is available
Special Features: UPMC Health Plan is the second largest
health insurer in western Pennsylvania, covering commercial,
medical assistance, and Medicare populations; it has integrated
resources from Community Care Behavioral Health, University
of Pittsburgh School of Pharmacy, and the UPMC Health
System.
Rae Ann Maxwell
Director, Pharmacy Services
UPMC Health Plan
One Chatham Center, 5th Fl.
112 Washington Pl.
Pittsburgh, PA 15219
(412) 454-5276
JMCP
September 2005
Vol. 11, No. 7
(412) 454-5295 (fax)
[email protected]
■■ VA SAN DIEGO HEALTHCARE SYSTEM
Managed Care Pharmacy Systems
Accredited:
Application submitted
Length of Program:
12 months
Affiliation:
University of the Pacific School of
Pharmacy
Starting Date:
July 1
Estimated Stipend:
$38,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD plus first-year
residency or equivalent experience, U.S. citizenship, personal
statement, curriculum vitae, transcripts, and 3 letters of recommendation
Fringe Benefits: 12-13 vacation and sick leave days, 11 federal
holidays, health and life insurance, free parking, paid leave,
tuition and travel to required events, office with up-to-date
computer systems
Special Features: This second-year residency will provide the
skills necessary for the practical application of pharmacoeconomic principles to formulary management and outcomes
research in integrated health care systems. Education will include
formal pharmacoeconomics training classes and hands-on
application of principles. Work activities will encompass the
VASDHS, Veterans Integrated Service Network 22 (VISN 22)
pharmacy benefits management (i.e. Southern California
Regional VA), and VA national formulary tasks. Out-of-state and
in-state travel is required. A university-affiliated, teaching,
integrated healthcare system with 100% computerized medical
records, cutting-edge patient safety, pharmacy-managed clinics,
pharmacist specialty practices, and a dedicated Pharmacy
Health Outcomes Division with 4 full-time pharmacoeconomists.
The pharmacy service has outstanding leadership with a long
positive track record for innovation and excellence at the local,
state, and national levels and a well-trained, well-published
staff, most with residencies. This is an exciting opportunity to
learn, and to start your career.
Anthony P. Morreale, PharmD, MBA, BCPS
Chief, Pharmacy Service
VA San Diego Healthcare System
Pharmacy Service (119)
3350 La Jolla Village Dr.
San Diego, CA 92161
(858) 552-8585, ext.. 3026
(858) 552-4391(fax)
www.amcp.org
[email protected]
Web: http://www.san-diego.med.va.gov
■■ WALGREENS HEALTH INITIATIVES
Pharmacy Benefit Manager
Accredited:
AMCP/ASHP
Length of Program:
12 months
Affiliation:
University of Illinois, Midwestern
University–Chicago College of
Pharmacy
Starting Date:
July 5
Estimated Stipend:
$36,000
Onsite Interview:
Yes
Fringe Benefits: Medical plan, 2-week vacation, holidays,
travel expense budget
Special Features: This managed care pharmacy residency
program is designed to allow the residents to work within the
various departments of a pharmacy benefits management firm,
including, but not limited to, care management, drug use policy,
PBM operations, clinical sales, and specialty pharmacy. The residents will gain practical experience and develop skills related
to disease management, health outcomes, medication management strategies, formulary management, drug utilization review,
drug information, and other clinical services. Additionally,
residents will have the opportunity to gain exposure to the
pharmaceutical industry, be involved in professional organizations, and precept pharmacy students.
Susie Min
Manager, Clinical Education and Shared Faculty
Walgreens Health Initiatives
1417 Lake Cook Rd., MS# L457
Deerfield, IL 60015
(847) 964-6740
(847) 374-2669 (fax)
[email protected]
www.amcp.org
■■ WELLPOINT PHARMACY MANAGEMENT
Accredited:
ASHP/AMCP
Length of Program:
12 months
Affiliation:
Blue Cross of California, PrecisionRx,
University of Southern California
Starting Date:
July 1
Estimated Stipend:
$50,000
Onsite Interview:
Yes
Educational/Special Requirements: PharmD from an ACPEaccredited college of pharmacy or equivalent experience,
eligibility for California licensure, good academic standing,
excellent written and verbal communication skills
Fringe Benefits: Health insurance; 2 weeks paid vacation; paid
holiday and sick days; attendance at the Western States
Conference, at a national pharmacy organization meeting, and
at a WellPoint Pharmacy Management National P&T meeting
Special Features: The program is designed to provide the
resident with an overall managed care experience. The resident
rotates through several areas within the PBM, including drug
information, therapy management, clinical account management,
clinical product development, clinical intervention and prior
authorization centers, mail-order and specialty pharmacy,
health informatics and policy research, and pharmaceutical
contracting and industry relations. The program also includes
rotations at the Blue Cross of California health plan and the
University of Southern California direct-patient-care sites.
Krista Yokoyama
Residency Program Director
WellPoint Pharmacy Management
8407 Fallbrook Ave.
MS AF-7
West Hills, CA 91304
(818) 313-5082
(818) 313-5110 (fax)
[email protected]
Vol. 11, No. 7
September 2005
JMCP

Pharmacist response to alerts generated from medicaid pharmacy

Transcription

Similar documents

PDF - First Nations Health Authority

Qmatic Pharmacy - Customer Flow Management

Pharmacies

Fyfe`s Friendly Pharmacy Address: 5101-50 St

Link to presentation - Leslie Dan Faculty of Pharmacy, University of

New cancer drug inquiry

Pharmacy Daily for 250712

Brochure

Harding Students Lobbied for Arkansas Therapeutic Substitution Law

It`s Digest Time Again