1. Disease Management and the AHRQ Research Agenda David Atkins, MD, MPH Agency for Healthcare Research and Quality Disease Management Colloquium, 2006

2. Outline of Talk DM and AHRQ’s agenda in research and quality DM and AHRQ’s agenda in research and quality The potential of, and obstacles to, DM in bridging the “quality chasm” The potential of, and obstacles to, DM in bridging the “quality chasm” Thoughts on what do we still need to know about DM Thoughts on what do we still need to know about DM

3. AHRQ Mission Statement To improve the quality, safety, efficiency, and effectiveness of health care for all Americans To improve the quality, safety, efficiency, and effectiveness of health care for all Americans

4. AHRQ Strategic Direction Accelerating the Pace of Innovation Ensuring Value through More Informed Choice Ensuring Value through More Informed Choice Assessing Innovation Faster Assessing Innovation Faster Implementing Effective Interventions Sooner Implementing Effective Interventions Sooner

5. What Is Appropriate Role of Government? Monitor health care quality Monitor health care quality – National Healthcare Quality and Disparities Reports Inform health care decision-makers Inform health care decision-makers – Payers, providers, plans, patients Support development of health technologies and practices Support development of health technologies and practices – Tools, technical assistance Convene stakeholders Convene stakeholders Support acquisition of new knowledge Support acquisition of new knowledge – Primary research, syntheses

6. Changes that Will Increase Importance and Alter Role of DM Growing elderly population Growing elderly population – More surviving with chronic disease – Some conditions (e.g. diabetes) increasing on their own Medicare drug benefit Medicare drug benefit Medicare chronic care pilots and demonstrations Medicare chronic care pilots and demonstrations Pay for Performance Initiatives Pay for Performance Initiatives Consumer directed health plans Consumer directed health plans Electronic health records Electronic health records

7. 1. Monitoring Quality of Chronic Care: Improving but still variable 85% of patients with acute MI prescribed beta- blocker at discharge 85% of patients with acute MI prescribed beta- blocker at discharge 65% of patients with CHF and LV dysfunction prescribed ACE inhibitors 65% of patients with CHF and LV dysfunction prescribed ACE inhibitors 65% of depressed patients initiating drug treatment who get a continuous trial of drug therapy during acute phase 65% of depressed patients initiating drug treatment who get a continuous trial of drug therapy during acute phase 27% of patients with high blood pressure who have optimal control 27% of patients with high blood pressure who have optimal control AHRQ: National Healthcare Quality Report, 2005

8. Quality of Diabetes Care - 2005 2005 National Healthcare Quality Report (www.qualitytools.ahrq.gov)

9. Post-MI Care - 2005 2005 National Healthcare Quality Report (www.qualitytools.ahrq.gov)

10. 2. Informing Decision Makers “Best Practices” Series Systematic reviews of interventions to improve care in IOM’S High Priority Health Conditions Systematic reviews of interventions to improve care in IOM’S High Priority Health Conditions – Emphasis on highest quality designs Improving care of diabetes and hypertension Improving care of diabetes and hypertension – 2004, 2005 Health literacy - 2005 Health literacy - 2005 Improving asthma care – due this year Improving asthma care – due this year Care coordination – due this year Care coordination – due this year

11. Diabetes Interventions Studied Patient education Patient education Patient reminders Patient reminders Promotion of self- management Promotion of self- management Provider education Provider education Provider reminders Provider reminders Facilitated relay of clinical data Facilitated relay of clinical data Audit and feedback Audit and feedback Organizational change Organizational change Financial, regulatory, legislative incentives Financial, regulatory, legislative incentives

12. Effects of # of Intervention Strategies on HbA1c and Provider Adherence

13. Improving Hypertension Control 63 studies of various interventions 63 studies of various interventions – Patient reminders, identifying high-risk patients, nurse follow-up, etc. Median reduction of 4.5 mm (SBP), 2.1 mm (DBP) Median reduction of 4.5 mm (SBP), 2.1 mm (DBP) Greater effects of interventions emphasizing organizational change and patient education Greater effects of interventions emphasizing organizational change and patient education Lesser effects of those emphasizing provider adherence with guidelines Lesser effects of those emphasizing provider adherence with guidelines

14. Improving Asthma Care 53 RCTS and 17 controlled before –after 53 RCTS and 17 controlled before –after Children: Educational interventions aimed at parents most important Children: Educational interventions aimed at parents most important – 4 studies: 8+ hours of educations – 2 studies – single individual session with specialist Adults: Education combined with system change or multidisciplinary approach more effective Adults: Education combined with system change or multidisciplinary approach more effective Adolescents: Limited research, little impact Adolescents: Limited research, little impact Patient self-management review in progress Patient self-management review in progress

15. General conclusions and limitations of Both DM and system approaches effective Both DM and system approaches effective Literature limited by poor reporting of specific details of interventions Literature limited by poor reporting of specific details of interventions Secular improvements, reporting bias, and weaker study designs may exaggerate effects. Secular improvements, reporting bias, and weaker study designs may exaggerate effects. Combination approaches needed to affect outcomes Combination approaches needed to affect outcomes Limited studies of commercial DM programs with good outcomes data Limited studies of commercial DM programs with good outcomes data Difficult to generalize findingsa across settings and populations Difficult to generalize findingsa across settings and populations

16. Care Coordination Overview of interventions and concepts Overview of interventions and concepts 53 systematic reviews 53 systematic reviews 17 different interventions in 7 different populations 17 different interventions in 7 different populations – E.g. multidisciplinary teams for diabetes care – Case management for depression 4 conceptual frameworks 4 conceptual frameworks

17. Effects of DM on overall health care costs Debates over appropriate methodology Debates over appropriate methodology CMS Pilots with RCT design may provide more definitive answer CMS Pilots with RCT design may provide more definitive answer – RCT of DM for diabetes and CHF in Indiana Medicaid 2006 DMAA initiative to standardize methods 2006 DMAA initiative to standardize methods Problems in: Problems in: – Accounting for administrative costs of programs – Controlling for secular trends in costs – Regression to mean and selection bias

18. Challenge for Research: How do we balance concerns about “internal validity” (does it really work?) with “external validity” (is it relevant to the real world?) How do we balance concerns about “internal validity” (does it really work?) with “external validity” (is it relevant to the real world?) Need to understand and reduce sources of bias in non-randomized studies of DM Need to understand and reduce sources of bias in non-randomized studies of DM Need combination of clinical and economic outcomes to validate effects Need combination of clinical and economic outcomes to validate effects

19. 3. Helping Develop Effective Practices in Disease Management Working with Partners Working with Partners – Health plans - disparities – Medicaid programs HIT demonstrations HIT demonstrations Developing Tools Developing Tools

20. Health Disparities Health Plan Collaborative Partnership between RWJ, AHRQ, 9 National Health Plans Partnership between RWJ, AHRQ, 9 National Health Plans 76 million covered lives 76 million covered lives Focus on reducing disparities in diabetes Focus on reducing disparities in diabetes Center for Health Care Strategies/ Rand/ Institute for Healthcare Improvement providing training and technical assistance Center for Health Care Strategies/ Rand/ Institute for Healthcare Improvement providing training and technical assistance

21. Working with Medicaid 2 year project beginning 2005 2 year project beginning 2005 Working through “knowledge translation” contractors with 6 states that have implemented DM in their Medicaid fee-for- service plans Working through “knowledge translation” contractors with 6 states that have implemented DM in their Medicaid fee-for- service plans Establishing “learning network” to promote sharing knowledge about developing, running and evaluating disease management Establishing “learning network” to promote sharing knowledge about developing, running and evaluating disease management Improve ability to use data to measure quality Improve ability to use data to measure quality Improve decisions in DM contracting Improve decisions in DM contracting

22. Health Information Technology Regional Projects – “RIOs” Promoting regional collaborations to share data Promoting regional collaborations to share data Emphasis on chronic diseases Emphasis on chronic diseases Community-based disease registries Community-based disease registries

23. Promoting Tools National Guideline Clearinghouse National Guideline Clearinghouse National Quality Measures Clearinghouse National Quality Measures Clearinghouse Quality Tools Quality Tools Estimating Costs of Chronic Disease Estimating Costs of Chronic Disease – AHRQ/CDC collaboration using Medical Expenditure Panel Survey Consumer satisfaction (CAHPS) Consumer satisfaction (CAHPS) – Piloting measures of self-management support

24. Barriers to the “Business Case” for Quality Not paying for quality, paying for defects Not paying for quality, paying for defects Inability to market quality to consumers Inability to market quality to consumers Payoffs removed in time and place Payoffs removed in time and place Disconnection between consumers and payers Disconnection between consumers and payers – Patients can’t pay for what they value Clinicians lack access to relevant information Clinicians lack access to relevant information –Leatherman, Berwick wt al. Health Affairs 2003

25. Breaking Down Barriers to Business Case Patients: Patients: – Better information on quality – Greater choice (e.g. Consumer directed plans) Clinicians: Clinicians: – Health information technology, registries – Ability to market, incentives for quality – Innovate in approaches to care Payers: Payers: – Pay for performance – Differential pay for sicker patients – Pay for alternative delivery modes (group visits, e-mail) – Support IT and greater choice

26. 4. Convening Stakeholders in DM Link clinicians, plans, payers, patients, policy makers, vendors Link clinicians, plans, payers, patients, policy makers, vendors Look across conditions Look across conditions Improve our ability to measure progress Improve our ability to measure progress Identify partnerships to advance implementation Identify partnerships to advance implementation Emphasize importance of disparities Emphasize importance of disparities

27. Input From Research and QI Community Help transfer knowledge Help transfer knowledge – Disseminate models of success – Connect partners, establish learning networks Bridge gap between Research/QI community Bridge gap between Research/QI community – Help promote better reporting – Improve research methods, synthesis Research and Evaluation Research and Evaluation – Patient self-management

28. Input from Employer Purchasers Improve models for predicting costs of chronic diseases Improve models for predicting costs of chronic diseases – Including productivity Improve and standardize methods for calculating ROI Improve and standardize methods for calculating ROI – Provide objective standards to validate vendor analyses – Promote greater transparency of methods Identify best methods for self-management support and valid measures to gauge success Identify best methods for self-management support and valid measures to gauge success

29. Improving Methods to Assess Economic Impact

30. 5. Generating New Knowledge: Challenges in DM Research Rapid pace of change Rapid pace of change RCTs difficult, less applicable to real world RCTs difficult, less applicable to real world Growth of private sector activity Growth of private sector activity – Proprietary data Disease-specific research silos Disease-specific research silos Importance of system interventions Importance of system interventions

32. Learning from what doesn’t work Not all approaches to DM are effective Not all approaches to DM are effective Telephonic support for CHF in Kaiser Telephonic support for CHF in Kaiser – Frank et al., Ann Intern Med 2004 Possible reasons: Possible reasons: – Less effective in low-risk patients – Telephone-only DM lacked other components – Better baseline of care We need to do a better job of determining: We need to do a better job of determining: – Essential components – Applicable populations – Effect of settings

33. 3 Critical Areas for Research and Action Standardizing methods and evaluation Standardizing methods and evaluation Patient self-management Patient self-management Incorporating DM into system redesign Incorporating DM into system redesign

34. Standardizing Evaluations DMAA approach to standardizing methods DMAA approach to standardizing methods Project to develop decision guide for Medicaid programs on economic evaluations of DM Project to develop decision guide for Medicaid programs on economic evaluations of DM Institute of Health Policy/Brandeis project to develop guidance for health plans Institute of Health Policy/Brandeis project to develop guidance for health plans Can we promote greater transparency while protecting proprietary methods? Can we promote greater transparency while protecting proprietary methods?

35. Patient Self-Management RAND review of patient self-management RAND review of patient self-management – Literature review – Informant interviews with industry, health plans, researchers, purchasers Describe range of approaches Describe range of approaches Describe methods for evaluating effectiveness of self-management support Describe methods for evaluating effectiveness of self-management support – Short term measures Examine specific issues: Examine specific issues: – What approaches work in hard to reach groups (e.g. low literacy, non-English speaking)?

36. Informed, Activated Patient Productive Interactions Prepared, Proactive Practice Team Improved Outcomes Delivery System Design Decision Support Clinical Information Systems Self- Management Support Health System Resources and Policies Community Health Care Organization Care Model

37. Incorporating DM Into Efforts to Redesign the Care System How can DM be better integrated into primary care? How can DM be better integrated into primary care? – Does it make a difference? Can we promote more effective practice teams in a fragmented healthcare system? Can we promote more effective practice teams in a fragmented healthcare system? Which organizational/delivery system interventions are most effective? Which organizational/delivery system interventions are most effective? How can we promote and measure their use in HIT innovations? How can we promote and measure their use in HIT innovations?

38. Conclusion Disease management models will continue to evolve Disease management models will continue to evolve Effective integration into clinical practice remains major issue Effective integration into clinical practice remains major issue Cost-saving vs. “improving value” Cost-saving vs. “improving value” DM as a component of (not alternative to) of system redesign DM as a component of (not alternative to) of system redesign