Integrating PROs for Clinical Care & Research Rachel Hess, MD, - PowerPoint PPT Presentation

Integrating PROs for Clinical Care & Research Rachel Hess, MD, MS @ r h e s s m d

EVERY PATIENT SHOULD HAVE ACCESS TO THE BEST CARE FOR THEM AND THAT CARE SHOULD BE…. • Timely • Personalized • Safe @ r h e s s m d

A growing body of evidence demonstrates that patients who are more actively involved in their health care experience better health outcomes and incur lower costs. -Robert Wood Johnson Foundation @ r h e s s m d

STRATEGIC INITIATIVES / POPULATION MANAGEMENT Treatment determination based upon outcome for • demographic – Use of Predictive Analytics to Tailor Treatment Plan Cost Benefit Analysis • Allow for more bundled payments • @ r h e s s m d

@ r h e s s m d

• Safe • Effective • Patient-centered • Timely • Efficient • Equitable @ r h e s s m d

Implementation @ r h e s s m d

PRINCIPLES FOR SETTING UP A HEALTH SYSTEM PRO COLLECTION SYSTEM Leadership needs to establish expectations • Patient Care Focus – Not Research • Decrease Burden of PRO collection on patients • – Short questionnaires – Collect only what is needed — Nothing More – Collect Remotely PRO results available at time of visit to provider • @ r h e s s m d

ENTERPRISE STANDARDIZATION Core General Health Assessment (E-VG-G-F-P) • Visual Analog Scale • PROMIS Physical Function • PROMIS Depression • Specialty Specific 1 Specialty Instrument (Can be CAT or legacy) • Minimize patient burden : 5-7 minutes (goal); no more than 10 minutes total (maximum) Set the “ interval ” by clinic for how often the questions are asked @ r h e s s m d

BUT WHAT ABOUT MY LEGACY SCALE? PROMIS CROSS-WALKS PANAS Neuro-QoL PHQ-9 PHQ-2 Upper Neuro-QoL Extremity Mobility PROMIS SF-36 PROMIS Physical DEPRESSION V1.0 Kessler 6 Function v1.2 Mental Health Scale SF-36 Physical HAQ- Beck Function Disability Depression Neuro-QoL Index CES-D Inventory-II Depression @ r h e s s m d

CHANNEL YOUR INNER DUNSON I’m concerned that this data isn’t available in Epic because the “ ortho model” didn’t • work well We are collaborating with the NCCN to identify outcome • I f you don’t think thru the “ownership” of these questions you are dead • measures for lung cancer. We're particularly interested in How are you going to get patients to do this? • At 5 mins a patient this is going to kill clinic productivity and slow down the physician PROs. • This isn’t physician workflow, this needs to be operationalized without breaking the • We plan to use the generic PROs that are currently in place • clinic workflow at HCH and to develop custom PROs for the Thoracic group. What, we barely have 60% of patients signed up on MyChart, how are you going to • get an 80% response rate? We have a number of projects theorizing that different • What about all my questions I already built in Epic? • aspects of advanced NSCLC care impact outcomes, I don’t want to have to click to see any data • including PROs. How does mEVAL and Epic integrate? • The Breast & Melanoma Teams are either done or close to • What is my liability for asking these questions? • being set up with similar projects. I don ’ t have time to deal with the answers! • @ r h e s s m d

HOW PATIENTS ENTER DATA DIRECTLY INTO EPIC Two Ways Epic PRO Questionnaire Home Home Clinic Clinic MyChart Epic Welcome Tablet Workstation Tablet (iPad or Windows) (Windows Only) @ r h e s s m d

CLINIC WORKFLOW QR Code generated OSS finds patient Patients check in with MRN encounter within with OSS information EPIC embedded Patient is handed Application (either OSS scans QR tablet to complete Welcome or Code with tablet all questionnaires mEVAL) is launched After completion, patient is instructed to return tablet N o t e : T h e r e w i l l @ r h e s s m d b e a m i n i m u m a n d m a x i m u m i n t e r v a l b a s e d o n c l i n i c w h i c h p a t i e n t s w i l l c o m p l e t e t h e

CREATING VALUE FOR THE PATIENT @ r h e s s m d

EPIC PRESENTATION OF RESULTS @ r h e s s m d

EPIC PRESENTATION OF RESULTS • The report also has a tab where the physician can see the patient’s answers for every specific question that they were asked in the surveys • This can be filtered by individual assessment or by the visit date that corresponds to the assessment results @ r h e s s m d

HOW MUCH DATA HAS BEEN CAPTURED Overall since October 2015 we have gather assessment data from over 65,000 unique patients and 120,000 encounters The more data we collect the better our ability will be to help patients make educated health care decisions by using this data of similar interventions 17 @ r h e s s m d

Technology @ r h e s s m d

EPIC INTEGRATION Question/answer pairs stored in application tables • Upon completion, stored algorithm is run and creates • rows in score table When scores are recorded, HL7 process runs and creates • HL7 messages HL7 engine picks up messages and sends them to Epic • ETL to BOE for reporting embedded in Epic (Radar • dashboard link, custom programing point) @ r h e s s m d

Administration Application Web App tables (EDW) HL7 Epic-BOE ETL Integration PRO entry Reports @ r h e s s m d

EDW: Patient data, forms, scores, Q/As, status HL7 Message HL7 Queue Score Table Generation HL7 Informatica ETL BizTalk (2 mins) BI Architecture HL7 PRO entry @ r h e s s m d

mEVAL Analytics @ r h e s s m d

ADMINISTRATION WEB APP TOOL Login to web app Search Results Filter patient list by selecting a provider and date. Use the date and MRN Search to find walk-in or same day appointments. Click on a patient name to display available assessments. @ r h e s s m d

DISTRIBUTION OF SCORES SHOWN BY PROCEDURE Distribution is shown over time including standard deviations to show the shift over time The Mean Age Adjusted Assessment Percentile can be tracked in any time interval from any range of days pre- surgery to post surgery @ r h e s s m d

DISTRIBUTION OF SCORES SHOWN BY PROCEDURE Summary data Count of responses broken out both by that are included in the percentile the graphed data rankings as well as by standard deviations Adjustable “Days From Surgery” element, where the interval the user would like can be chosen 25 @ r h e s s m d

ABILITY TO TRACK INDIVIDUAL PATIENT’S CARE Cost of each visit The Physical Function Provider who saw the percentile over the patient and visit date course of care @ r h e s s m d

ABILITY TO TRACK INDIVIDUAL PATIENT’S CARE Each visit number hyperlinks to a report with all the detail line item costs for that visit Clinical information such as site of service and diagnosis or surgical procedure and some outcomes measures are shown for each visit for the selected patient @ r h e s s m d

DAILY CASE MANAGEMENT REPORTS @ r h e s s m d

DEPRESSION SCREENING PHQ 9 PROMIS Depression (only) (only) More information at Poster Session C, # total visits 408,926 94,488 Monday 6:30-8:00 # of surveys 11,814 (2.9%) 22,042 (23.3%) # of unique pts 7,289 16,046 # At “risk” 2,926 (0.7%) 1,297 (1.4%) @ r h e s s m d

PaTH's experiences with EPIC/PRO development and implementation Cecilia Dobi, MS, Project Manager, Temple University School of Medicine Diana Gumas, MS, Senior IT Director, Johns Hopkins Medicine Albert Wu, MD, MPH, Professor, Johns Hopkins Bloomberg School of Public Health Rachel Hess, MD, MS, PaTH Principal Investigator, UPMC, UUHC

THREE INITIAL USE CASES: • Idiopathic Pulmonary Fibrosis • Atrial Fibrillation • Obesity @ r h e s s m d

CHOOSING PROS: TEAM WORK! • Clinician • Researcher • Patient-partner • Informatician • System Analyst • PRO experts @ r h e s s m d

PRO COLLECTION (RESEARCH, CLINICIAN, STAFF) • Epic MyChart • RedCap @ r h e s s m d

SENDING QUESTIONNAIRES Report Multi-patient messaging Series @ r h e s s m d

Our hope is that, at the end of the day, we’ve enhanced each person’s health and quality of life @ r h e s s m d

University of Utah Population Health Sciences Booth 116