Analytics In Healthcare and Capturing the Patient Journey
Transcription
Analytics In Healthcare and Capturing the Patient Journey
Analytics In Healthcare and Capturing the Patient Journey APRIL 8, 2016 SAS HEALTH USERS GROUP SHIRLEY LI Objective Agenda ▪ Objective of presentation ▪ Overview of analytics in healthcare (in my view) ▪ Capturing the Patient Journey with Data ▪ What is the patient journey? ▪ How do we capture it? ▪ What datasets are available? ▪ Factors to consider? ▪ What information is captured? ▪ How is the information captured? Disclaimer: All information presented are my owns thoughts and interpretation Overview of Healthcare Analytics What does SAS mean to you? http://blogs.sas.com/content/sascom/2014/04/03/predictive-analytics-described-in-one-word/ Except…. HOT BUTTER STUDIO @ 2012 Goal HOT BUTTER STUDIO @ 2012 How to get there? Analytical Life Cycle Is Data Preparation Really That Hard? Forbest.com “Cleaning Big Data: Most Time-Consuming, Least Enjoyable” http://blogs.sas.com/content/hiddeninsights/2013/10/11/how-well-are-you-managing-the-analytical-life-cycle/ How to capture the patient journey? Vision and mission of a healthcare organization ▪ Vision ▪ Working together to create the best health systems in the world. ▪ Mission Statement ▪ Together, we will improve the performance of our health systems by driving quality, accountability, innovation and value. What does the patient journey look like? https://www.cancercare.on.ca/ocs/qpi/dispathmgmt/pathways/colopath/ What really does the patient journey look like? “Pathways are flowcharts that show a high-level overview of the care a cancer patient in Ontario should receive. Pathways focus on one type of cancer, during a specific phase of the cancer journey, with the understanding that the patient journey differs from one cancer to another.” https://www.cancercare.on.ca/ocs/qpi/dispathmgmt/pathways/colopath/ What really does the patient journey look like? www.deepjapan.org How do we capture this in the available datasets? Two possible sources for surgery data Health Analysts’ Toolkit, Health Analytics Branch, Winter 2012 What information are captured in the datasets? ▪ Discharge Abstract Database (DAD): ▪ “patient-level data are collected at the time of service in participating institutions. After the discharge, a medical records coder at the hospital completes an abstract according to instructions in the CIHI abstracting manual.” (pg. 7) ▪ OHIP Claim’s history database: ▪ “provider billings through the Ontario Health Insurance Plan (OHIP)” (pg. 7) ▪ “CHDB is designed for the assessment and processing of claims, its use for other purposes—such as measuring utilization of services or estimating conditions based on diagnoses—is secondary. Care must be taken with interpretation and analysis.” (pg. 7) Health Analysts’ Toolkit, Health Analytics Branch, Winter 2012 What is the relationship between the two databases? Comparison of the two databases: unit of analysis DAD CHDB ▪ Each observation represents a discharge ▪ Each observation represents one claim (NOT visit or activity) ▪ Patients transferred to another facility (after discharge) possibly for the same condition are in multiple observations ▪ e.g. A visit to the doctor’s office can be captured by multiple observations depending on what procedures were performed Not absolutely comparable Comparison of the two databases: how information is captured DAD CHDB ▪ Surgery information is captured by CCI codes ▪ Surgery information is captured by OHIP fee codes ▪ “CCI specifies more precisely than ever before what interventions and services health professionals provide. ” ▪ Schedule benefit: ▪ Multiple fields to capture interventions http://www.health.gov.on.ca/english/providers/program /ohip/sob/physserv/sob_master20160406.pdf ▪ Modifier fee codes for additional payment (e.g. laproscopic surgery E747) Not absolutely comparable CCI Code Surgery Code Description Surgery Category Disease Site 1NK87 DN Excision partial, small intestine endoscopic (laparoscopic) approach; Enterocolostomy anastomosis technique Excision partial, small intestine endoscopic (laparoscopic) approach; Enteroenterostomy anastomosis technique Excision partial, small intestine open approach; Enterocolostomy anastomosis technique Excision partial, large intestine endoscopic (laparoscopic) approach; Colorectal anastomosis technique Excision partial, large intestine endoscopic (laparoscopic) approach; Colocolostomy anastomosis technique Excision partial, large intestine endoscopic (laparoscopic) approach; Enterocolostomy anastomosis technique Excision partial, large intestine open approach; Colorectal anastomosis technique Excision partial, large intestine open approach; Enterocolostomy anastomosis technique Excision partial, large intestine open approach; Colocolostomy anastomosis technique Excision total, large intestine endoscopic (laparoscopic) approach; Ileorectal (endorectal, ileoproctostomy) anastomosis technique A - Resection of colon without stoma, with anastomosis Colon Surgical Resection Indicator Nonresective A - Resection of colon without stoma, with anastomosis Colon Nonresective A - Resection of colon without stoma, with anastomosis A - Resection of colon without stoma, with anastomosis Colon Nonresective Colon Resective A - Resection of colon without stoma, with anastomosis Colon Resective A - Resection of colon without stoma, with anastomosis Colon Resective A - Resection of colon without stoma, with anastomosis A - Resection of colon without stoma, with anastomosis A - Resection of colon without stoma, with anastomosis A - Resection of colon without stoma, with anastomosis Colon Resective Colon Resective Colon Resective Colon Resective 1NK87 DP 1NK87 RE 1NM87 DE 1NM87 DF 1NM87 DN 1NM87 RD 1NM87 RE 1NM87 RN 1NM89 DF CCI codes for surgical resection List not exhaustive OHIP Fee Code S166 Fee Code Description Disease Site Colon S167* Resection with anastomosis – small and large intestine terminal ileum, cecum and ascending colon (right hemicolectomy) Resection with anastomosis – large intestine Surgical Resection Indicator Resective Colon Resective S168 Ileostomy - subtotal colectomy Colon Resective S169 Resection with anastomosis – total colectomy with ileo-rectal anastomosis Colon Resective S171* Colon Resective Colon Resective S188 Resection with anastomosis – left hemicolectomy with anterior resection or proctosigmoidectomy (anastomosis below peritoneal reflection and mobilization of splenic flexure) Resection with anastomosis – total colectomy with mucosal protectomy with ilea pouch, ileoanal anastomosis and loop ileostomy Bowel resection without anastomosis (colostomy and mucous fistula) Colon Resective E793 Laparoscopic or laparoscopic assisted Colon Modifier Code S172 OHIP fee codes for surgical resection Might not be exhaustive Comparison of the two databases: activity dates DAD CHDB ▪ Admission date ▪ Service date ▪ Discharge date ▪ Admission date ▪ Intervention date Not absolutely comparable Comparison of the two databases: summary ▪ Not 100% match ▪ Dates can be different (actual mismatch vs. coding variability) ▪ Type of surgery can be different (actual mismatch vs. coding variability) ▪ Institution which the surgery occurred can be different (actual mismatch vs. coding variability) Might requires both sources to fully capture surgical resection information. Reconciliation of mismatches is very time consuming! …”its use for other purposes—such as measuring utilization of services or estimating conditions based on diagnoses—is secondary” Summary ▪ Prior to thinking about advanced analytics, make sure you know what data sources you need and what information you need ▪ Sometimes multiple sources are needed (Warning! Matching can be challenging) ▪ Understand how the data is captured prior to you extract data ▪ Understand what is and is not captured in data sources ▪ Be aware of dates, unit of analysis, diagnosis codes (ICD-9, ICD10, ICD-O3, OHIP diagnosis codes), etc. ▪ Read database documentation and ask colleagues around you End result: Save time! Acknowledgments: ▪ All my colleagues in the Cancer Analytics Team within Analytics and Informatics, Cancer Care Ontario ▪ My team lead and manager: Kelly Woltman and Asmaa Maloul for their support of this presentation ▪ Disease Pathway Management, Clinical Program Quality Initiative, Cancer Care Ontario Questions? Thank you! Contact: [email protected] or [email protected]