Using Physician Payment to Improve Health System Performance Erin Strumpf, PhD McGill University Levers of change to act on health system performance workshop Québec Ministère de la Santé et des Services sociaux Seminar on Performance Measurement March 24, 2011
Health System Performance Efficiency value for money Getting the right care to the right patients Health care services whose benefit exceeds their cost This means focusing on Costs and benefits of treatment decisions Quality of care and outcomes Prevention and effective management of chronic conditions Appropriate mix of primary, specialist and tertiary services
Not All Health Care Services Are Created Equal Highly effective and inexpensive innovations Antibiotics, or aspirin and beta blockers for cardiac care More expensive yet effective treatments for appropriate patients Hip and knee replacements, surgical interventions for heart attack patients Gray area treatments with uncertain clinical value ICU days among chronically ill patients Chandra and Skinner, Technology and Expenditure Growth in Health Care 2008
Available Levers and Incentives Other systems use financial incentives targeted at patients Value-based insurance design, co-payments, etc Levers available in the Canadian context largely target providers
Why Focus on Physician Payment? 13% of health care spending on physician services Smaller share than hospitals and drugs, but growing faster Physicians decisions have a large impact on patients outcomes and system efficiency What services to provide Outpatient and inpatient Prescription drugs Referrals to specialists and social services
Fee-for-Service (FFS) Providers are paid a negotiated price for each service provided Rewards volume of services Not value, outcomes, or quality 48% Canadian physicians earn 90%+ of income from FFS* 56% in AB, 30% PEI, 43% in QC Perverse incentives Over-provision of (paid) services Under-provision of (unpaid) tasks Little ability for insurer to control costs or improve efficiency *2007 National Physician Survey
Capitation and Fundholding Providers are paid a fixed amount per year for each patient enrolled in their practice Primary care services without additional payments All medical services without additional payments Rewards keeping patients healthy using a range of services Perverse incentives Selection of healthier patients Excessive use of specialist or hospital care (capitation) Under-provision of services and quality, especially problematic with limited supply of health providers Strong incentives for cost control
Mixed Payment Blending FFS and capitation can provide strong incentives for efficiency while moderating the perverse incentives inherent in each Balance of consideration of costs, benefits, and quality 31% of Canadian physicians QC: mixed compensation for specialists (per diem and pro-rated FFS) ON: mixed payment for GPs in FHNs and FHOs (capitation and FFS/incentives) Evidence of physicians changing their behavior in response*, still need to understand the impacts on patient health outcomes *Dumont et al, Journal of Health Economics 2008; Kantarevic et al, Journal of Health Economics 2011
Pay-for-Performance (P4P) Providers are paid based on desired outcomes: Structure: the resources assembled to deliver care, including personnel, facilities, and materials Process: the completion of specific tasks or recommended treatments Outcome: patients experience and health status Computerized physician order entry (CPOE), ICU staffing, information systems to track chronically ill patients Cervical cancer screening, mammography, immunizations, chronic disease management In-hospital mortality, complication, and readmission rates Rosenthal et al, Health Affairs 2004
Pay-for-Performance (P4P) Can be combined with many other payment systems Perverse incentives Teaching to the test Selection of patients to improve ratings Cost control potential depends on the extent to which rewarded processes and outcomes reduce costs or improve efficiency
Pay for Performance Challenges The details of such a system are not obvious and they do matter* Do incentives apply to individuals or groups? What s the right amount to pay? Which performance measures to use? Do both achievement and improvement get rewarded? Adjustment of measures and payments to moderate perverse incentives Potential for impact is greater in a single-payer environment *Rosenthal et al, JAMA 2007
Challenges to Using Physician Payment to Improve Health System Performance Political barriers to change Data to measure impacts on ultimate outcomes of interest Physician behavior and utilization patterns are not sufficient to inform questions about system efficiency We have to measure quality of care and health outcomes