1. Integrated Care Coordination Information System: primary care redesign through care coordination and population management David A. Dorr, MD, MS Associate Professor Department of Medical Informatics & Clinical Epidemiology General Internal Medicine & Geriatrics OHSU Funding for this research from The John A. Hartford Foundation, AHRQ, Intermountain Healthcare, and the National Library of Medicine More information at caremanagementplus.org

2. Overview Care Management Plus: ICCIS need and trial Prioritized functions Unintended consequences Sustainability: Free take one vs. thoughtful partnership A.k.a – How to build a better system of care for your most at-risk primary care patients

3. Becomes

4. 4 Arm 1: Coordination of Care -Complete assessment/care plan -Education -Goal setting and follow up -Communication -Motivation/coaching -Completing CM services Evaluation (Aim 4): Outcomes (health/satisfaction) and their relationship to implementation and use of IT Needs assessment / Build system (1 year + ) Train clinics and care managers Randomly assigned goals for IT use Arm 2: Quality -choose 5 of 20 quality measures: prevention, diabetes, vulnerable elderly, asthma, congestive hearth failure Data from ICCIS, Payers

5. Needs assessment Behkami, Proc AMIA, 2009

6. ICCIS Care Coordinatio n Workflow A centralized reminder list of tasks and communications that were proactively planned but incomplete allows population-based tasks to be merged with individual encounter tasks.

7. Quality measure dashboard Dashboard can be run by clinic, team, or individual PCP

8. The abilities to document exclusions at multiple levels and generate targeted population-based review cycles avoid the problems caused by static quality reports and allow providers to efficiently focus outreach efforts on high risk populations. ICCIS Interactive Quality Reports

9. (1) Wilcox, Proc of AMIA Symp, 2005 Patient Worksheet When working with persons with multiple illnesses or complex illness, a clinical summary that captures a core set of information improves patient outcomes ( 1). Care coordination and behavioral modification (goal setting) elements often require special effort and the quality summary requires more advanced monitoring and implementation than most standard EHRs provide.

10. It worked! (see our poster) Figure 3: Absolute adherence change for Arms and Clinics Arms reimbursed Table 1. Care coordination activities

11. Unintended consequences: Errors / fixes 278 fixes of systemic errors in first 6 months of study Sources : – data (multiple EHRs, minimal standards); – workflow/usability; – Understanding/naming to reduce confusion

12. Sustainability - ‘Free, take one’ – dissemination to 208 teams SFDPH (12 sites) Intermountain (16 teams) OHSU (9 teams) PeaceHealth (20 teams) Daughters of Charity (5 teams) Colorado Access (16 teams) HealthCare Partners (2 sites)

13. Sustainability: Thoughtful partnership Readiness assessment : define benefit up front – E.g., Medical Home care coordination; ACO reduction in hospitalizations and shared savings Partner on achievement of goals Share savings or benefit together – Example: intensive care management demonstrations; SNP plans

14. Oregon Health & Science University – David Dorr, PI – Kelli Radican – Susan Butterworth – Nima Behkami – Marsha Pierre-Jacques Williams – Gwenivere Olsen – Molly King – Kristin Dahlgren Columbia University – Adam Wilcox Intermountain Healthcare – Cherie Brunker, Co-PI (UU) – Liza Widmier – Mary Carpenter – Bryan Gardner – Ann Larsen Advisory Board – K. John McConnell – Tom Bodenheimer – Eric Coleman – Cheryl Schraeder – Heather Young – Steven Counsell – Larry Casalino

15. Thank you & Main lessons TopicToolAssessment WorkflowTickler as CDSS and single workflow Needs assessment and requirements; usability Patient-centered Care Patient WorksheetAccuracy, usefulness from clinical staff Unintended consequences Error tracking with clinical consequences Fixes needed Sustainability“Free, take one” v. Thoughtful partnership Need has to be clearly assessed and targeted dorrd@ohsu.edu http://www.caremanagementplus.org

16. Thank you! dorrd@ohsu.edu http://www.caremanagementplus.org

17. Sticky problems ProblemExplanationICCIS Result Data in many different EHRs EHRs have different data structures Extracted data from 4 different EHRs Functions in HIT systems siloed Many functions are in separate EHR settings Create universal workflows in separate application EHRs have variable standard implementations Although a standard vocabulary is available, it isn’t used NOT EASY – manual mappings, many errors until it is solved Population management is an analytic, not transactional issue Reports take a long time to run and are static Create interactive views of the reports (e.g., quality measure performance) with associated tables

18. Usability Log metrics: click throughs (<5 seconds on page): 62% ; loops/ repeated actions Interviews: Use / workflow / challenges / errors