MediPass Primary Care Physician Network Analysis

MediPass Primary Care Physician Network Analysis: Primary Findings
Allyson Hall, PhD
Heather Steingraber, MAC (ABT)
Steve Shaffer
Christy Harris Lemak, PhD
Jianyi Zhang, PhD
Florida Center for Medicaid & the Uninsured
College of Public Health and Health Professions
University of Florida
352/273-5059
Sponsored by
The Agency for Health Care Administration
June, 2004
Introduction
Assuring that there is adequate primary care capacity is critical to the success of Medicaid primary care case
management programs, such as MediPass. Several earlier studies documented that physician participation in
Medicaid varies considerably throughout the country, and there has been a decline in participation in recent years.
In 2001, about 85 percent of all physicians throughout the country participated in Medicaid. However, for most
physicians, their Medicaid practice tends to be small – contributing less than 10 percent to overall total practice
revenues (Cunningham, 2002). Nationally, the number of physicians who will not accept new Medicaid patients
increased from 26 percent to 31 percent between 1999 and 2001 (Schoenman, 2003). And, in Alabama and
Georgia, the introduction of a primary care case management program similar to MediPass, actually resulted in
lower physician participation in Medicaid (Adams, et al, 2003). Low reimbursement levels, burdensome paperwork,
and persistently negative opinions about Medicaid are said to be some of the reasons for low physician
participation in some communities (Berman et al, 2003; Perloff et al, 1995; Bindman at al, 2003).
In Florida, a 2002 evaluation concluded that Florida MediPass had excess primary care capacity. At the time,
there were ov er 4,685 contracted providers or groups providing a capacity that was 10 times greater than the
enrolled MediPass population.
The report noted that in order to keep costs down, Medicaid may want to consider
reducing the size of the MediPass network so as to exert control over providers and to assure greater efficiency
within the program(Florida Center for Medicaid Issues, 2002).
This earlier study relied on data that did not consider whether physicians were accepting new Medicaid
patients. Physicians may be contracted with MediPass, but limit their practice to a few existing patients. Thus, while
it may appear that a community or region has adequate coverage, this may in fact not be the case if only a few
physicians are taking additional patients.
In addition, the MediPass program has instituted a requirement that for
all program enrollees, a PCP must be located within 30 minutes drive time of their home. Therefore, simply looking at
the number of physicians in an area without appropriate consideration of the drive time rule may lead to an
overestimation of PCP capacity.
To address this issue, the Florida Center for Medicaid and the Uninsured contacted MediPass primary care
physicians to learn if they are accepting new Medicaid patients. Mapping analysis then examined the extent to
which there is sufficient MediPass primary care capacity throughout the state.
Methods
Sample Participating MediPass primary care physicians were identified from lists supplied by Medicaid area offices
and from a list generated by QueryPath. The various lists were combined resulting in a master list of over 3,000
unique providers identified by their unique provider IDs.
Survey The main goal of the project was to collect information on whether physicians were accepting new MediPass
patients and to understand some of the barriers MediPass patients may encounter in trying to make appointments
with their PCPs. It was important to ensure that interviewers engaged in realistic dialogue with office staff, and that
2
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
interviewers could easily document findings from these conversations. Therefore interviewers used a basic script
that asked: whether the provider was accepting new patients in general, whether they were accepting MediPass
patients specifically, and if so, how long it would take to get an appointment. Physicians’ offices were generally not
aware that they were being surveyed.
In a training session, interviewers role-played and tested the script with several initial calls. Modifications were
made based on these pilot calls. During and after the call other information was also recorded, such as the length
of time the interviewer spent on hold, and the number of times they were transferred. This information was gathered
in order to capture the ‘hassle factor’ involved in trying to get an appointment.
This information will be reported
elsewhere. The data sheet used to collect this information is included in Appendix 1.
Data Analysis Florida Medicaid requires a MediPass Providers to be within a 30-minute drive of every enrollee. This
criterion was used to create MediPass provider service areas based on 30-minute drive times out in every direction
from the provider’s address. Creating drive time service areas for every accepting MediPass provider in the state will
allow for the identification of gaps in service.
Geographic Information Systems (GIS) is a standard tool for mapping and exploring variation in the demand
and supply of health care services (McLafferty, 2003). In this analysis GIS was used to place a point on a map for
every MediPass provider by geocoding based on the provider’s address or zip code. MediPass beneficiaries were
also geocoded by zip code. Service area polygons were then created by a network analysis of the statewide
3
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
street system radiating out from each provider’s point location. The drive time computation took into account
distances and speed limits along the network. Traffic congestion and other travel costs were also taken account by
the GIS system. Service areas for each provider were merged together to provide complete coverage area for the
state for all accepting providers and for specific provider types, such as pediatricians or ob-gyns. Maps and other
data were then produced to show service areas statewide and highlight gap areas where no provider was within a
30-minute drive time of beneficiary zip codes.
Findings
The series of maps and charts presented here summarize preliminary findings from the survey and mapping
exercise. A total of 3,800 physicians were identified as either being on an area office list or were listed in QueryPath.
Eighty-eight percent of all physicians surveyed said that they were accepting new patients. However, when
probed, a smaller percentage of offices (74 percent) responded that they were actually taking new MediPass
patients.
With a few exceptions, statewide coverage by MediPass physicians appears to be good. Using the 30-minute
drive-time rule discussed earlier, most of the state’s MediPass population appears to have appropriate access to a
primary care provider.
Areas of concern include parts of south Florida including Monroe county, and communities
in south central areas and in the Panhandle. Of particular concern is the lack of pediatric and obstetrics and
gynecology physicians in certain areas.
4
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 1 – Percent of Florida MediPass Physicians by County who are Accepting New Patients
The majority of MediPass providers are
100
80
89
88
88
90
83
92
93
90
86
continuing to accept new MediPass
89
83
76
74
71
65
70
patients. When asked, 88 percent of
60
offices said they will see new patients
40
However, when asked specifically about
20
Medicaid patients, 74 percent of
0
providers will see a new MediPass
Flo
Bro
D
rid
war ade
a
d
Du
val
Hil
Ora
P
P
lsbo
nge alm B inella
rou
s
eac
gh
h
Area Office
New Patients
New MediPass Patients
* Includes physicians who are either on an area office or QueryPath list
patient. Acceptance of new MediPass
patients varies across communities. In
Broward only 65 percent of identified
MediPass physicians will take a new
patient, compared to 86 percent in Duval
5
county.
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 2 – Percent Of MediPass Physicians On An Area Office List Who Are Accepting New Patients
Area offices supplied lists of
100
98 97
90
80
92
92
88
85
97 95
85
77
70
96
91
88
85
83
76
74
71
74
MediPass patients. These are the
66
60
59
providers who are accepting new
58
same lists that are given to MediPass
40
beneficiaries. Statewide, 77 percent
20
of physicians on an area office list
0
will see new MediPass patients. This
Flo 1
rid
a
2a
3
4
5
6
7
8
Area Office
New Patients
New MediPass Patients
9
10
11
ranges from a low of 58 and 59
percent in Areas 8 and 1 respectively,
to a high of 97 percent in Area 4.
6
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 3 – Percent of MediPass Physicians on An Area Office List who are Accepting New Patients:
Statewide, by Specialty
Ninety percent of the OB/GYN physicians
100
80
94
91
90
77
87
surveyed will see new MediPass patients,
90
92
90 91
compared to 65 percent of Internal
81
78
65
60
Medicine physicians. Although there are
40
high acceptance rates among OB/GYNs,
20
coverage is inadequate because of the
relatively small numbers of participating
0
All
FP/
GP
Inte
OB
/GY
rna
lM
N
ed
Ped
s
Area Office
New Patients
Oth
er
OB/GYN physicians (see Figures 16 and
17).
New MediPass Patients
7
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 4 – MediPass Providers Statewide
As expected, most of the MediPass
physicians are located in the major
metropolitan areas of Miami-Dade, Ft.
Lauderdale, West Palm Beach, Tampa,
Orlando, and Jacksonville…….
8
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 5 – MediPass Enrollees by Zip Code Area
…..which is where most of the MediPass
beneficiaries live.
9
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 6 – MediPass Provider Service Areas
The gray area represents areas within 30
minutes of a MediPass provider (also called
service area). Most of Florida’s populated
areas are within 30 minutes of a MediPass
provider. A few small, wealthy communities
are not within 30 minutes of a MediPass
provider.
10
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 7 – Ratio of MediPass Enrollees to Accepting Providers Located within a 30-minute Drive Time
Gainesville, Lake City,
Daytona Beach and Chiefland
Panama City and Perry
Most MediPass beneficiaries live within
30 minutes drive time of a primary care
provider. However, some communities
(identified by the blue squares) rely
Ft. Walton
Beach,
Century, and
Pensacola
heavily on a few providers to deliver
care to MediPass patients. MediPass
patients could experience access barriers
if these providers should withdraw from
Medicaid.
Port St. Lucie
However, Medicaid areas with large
numbers of providers may wish to
consider downsizing their networks in
Immokalee, Arcadia, and
Okeechobee
Key West and Marathon
order to exert control over providers and
to assure greater efficiency.
11
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 8 - Ratio of MediPass Enrollees to Accepting Providers within a 30-minute Drive Time: Pensacola and
Surrounding Areas
Zip codes 32536 (Century),
32506 , 32505, 32501
(Pensacola) and 32547 (Ft.
Walton Beach) have high
enrollee to provider ratios .
12
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 9- Ratio of MediPass Enrollees to Accepting Providers within a 30-minute Drive Time: Panama City
and Surrounding Areas
Zip codes 32401, 32404,
32405 (Panama City) and
32348 (Perry) have high
enrollee to provider ratios.
13
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 10- Ratio of MediPass Enrollees to Accepting Providers within a 30-minute Drive Time: Southwest
Florida
Zip codes 34142 (Immokalee),
34266 (Arcadia), 34972 and 34974
(Okeechobee) have high enrollee to
provider ratios
14
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 11- Ratio of MediPass Enrollees to Accepting Providers within a 30-minute Drive Time: North Central
Florida
Zip codes 32055 (Lake City), 32607,
32609(Gainesville) and 32177 (Palatka)
have high enrollee to provider ratios.
15
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 12 - Ratio of MediPass Enrollees to Accepting Providers within a 30-minute Drive Time: Fort Pierce
and Surrounding Areas
Zip codes 34950, 34982 (Ft.
Pierce) and 34972 (Okeechobee)
have high enrollee to provider
ratios.
16
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 13 - Ratio of MediPass Enrollees to Accepting Providers within a 30-minute Drive Time: The Florida
Keys
Zip codes 33040 (Key
West) and 33050
(Marathon) have high
enrollee to provider ratios.
17
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 14 – MediPass Pediatric Provider Service Areas
The brown color represent areas within 30
minutes of a MediPass pediatric provider.
The green areas are not located within 30
minutes of a MediPass pediatric provider.
Several communities, indicated by red dots
are without a pediatric provider.
18
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 15 – Ratio of MediPass Child Enrollees to Accepting Pediatric Providers Located within a 30-minute
Drive Time
Areas with high enrollee to pediatric
provider ratios include Immokalee,
Labelle, Okeechobee, Arcadia,
Gainesville, Tallahassee, Pensacola,
and Century. These areas have
relatively few providers.
19
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 16 – MediPass OB/GYN Provider Service Areas
The pink area represents areas within 30
minutes of a MediPass OB/GYN
provider. The green areas are not
located within 30 minutes of a MediPass
OB/GYN provider. Many communities,
indicated by red dots are without
OB/GYN physicians within 30 minutes.
20
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
Figure 17 – Ratio of Adult Female MediPass Enrollees to Accepting OB/GYN Providers Located within a 30minute Drive Time
Almost one-third the state appears to have
a high patient to provider ratio. This may
be an indication of poor access OB/GYN
services among women enrolled in
MediPass.
21
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
References
Adams EK, Bronstein JM, Florence CS (2003) The Impact of Medicaid Primary Care Case Management on OfficeBased Physician Supply I Alabama and Georgia Inquiry 40 (Fall 2003): 269-282.
Berman S, Dolins J, Tang, SF and Yudkowsly B ( 2002) Factors that Influence the Willingness of Private Primary Care
Pediatricians to Accept More Medicaid Patients Pediatrics 110(2 part 2): 239-248.
Bindman AB, Yoon J, and Grumbach K (2003) Trends in Physician
Participation in Medicaid. The California Experience Journal of Ambulatory Care Management 26(4): 334-343.
Cunningham PJ (2002) Mounting Pressures: Physicians Serving Medicaid Patients and the Uninsured. 1997-2001.
Center for Studying Health System Change. Tracking Report 6: 1-4.
Florida Center for Medicaid Issue (2002) Evaluation of the Florida Medicaid MediPass Program, University of Florida,
College of Public Health and Health Professions
Perloff JD, Kletke P and Fossett JW (1995) Which Physicians Limit their Medicaid Participation and Why? Health
Services Research 30 (1): 7 -26
McLafferty SL (2003) GIS and Health Care Annual Review of Public Health 24: 25-42.
Schoenman JA and Feldman J (2003) Results of the Medicare Payment Advisory Commission’s 2002 Survey of
Physicians. Washington DC: Medicare Payment Advisory Commission.
22
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
23
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS
APPENDIX 1 – DATA COLLECTION SHEET
Provider Name: «PROVIDER_NAME»
Phone: «OFFICE_PHONE»
Address: «ADDRESS»
City: «CITY»
County: «County»
Practice Type: «Practice_Type»
Specialty: «Specialty»
Attempt 1
Attempt 2
Date & Time
Interviewer
# Rings
Phone Tree? Y/N
# Transfers
Hold Time
Still Accepting New Patients
(Stop if No)
Still Accepting MediPass
Patients
(Stop if No)
How long to Appt?
(If >1 week) Even if sick?
Taking Kids? Y/N
After hours & Weekends?
Courteous & Professional?
VC, SC, SD, VD
Comments
Attempt 3
Attempt 4
25
PRELIMINARY FINDINGS – DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS