Issues in the Design and Implementation of Pay-for-Performance Program
Issues in the Design and Implementation of Pay-for-Performance Programs
Gary J. Young, J.D., Ph.D.
Professor and Director
Northeastern University Center for Health Policy and Healthcare Research
Presentation for Agency for Healthcare Research and Quality, Annual Meeting
September 20, 2011
Financial support from the Agency for Healthcare Research and Quality and the Robert Wood Johnson Foundation Investigator Award for Health Policy Research.
Design and Implementation Issues
Multi-year study of over 70 provider organizations with quality-related incentive arrangements.
- Surveys of Physicians.
- Interviews with senior leaders from physician organizations and hospitals.
- Analysis of Clinical Data.
Design and Implementation Issues
- Selecting unit of accountability.
- Managing provider attitudes toward pay-for-performance program.
Unit of Accountability
- Teams (within or across organizations—e.g., ACOs).
Interviews with Senior Leaders from Physician Organizations and Hospitals
- Telephone Interviews and Site Visits.
|Setting||# Senior Leaders|
Interviews with Senior Leaders
General attitudes and issues:
- Quality incentives (better than utilization).
- Adequacy of dollars (new or old money).
- Awareness and involvement of physicians (grass roots vs. system engineering).
- Internal distribution of financial rewards (where individual physicians were not the unit of accountability):
- $ individual performance on pay-for-performance (P4P) criteria.
- $ individual performance on non-P4P criteria.
- $ equally independent of performance.
- $ retained at group level for investment (unit of accountability issue).
Interviews with Group Practice Executives
"We have a point system, but I would not classify that under the heading of necessarily a quality system. I'd call it more of a participation system. I think the outcome spills over a little into quality because again, the camaraderie and the communication improve and that's always a good thing when PCPs are talking to specialists, interfacing more...."
Incentives and Unit of Accountability
Image: A chart contrasting Efficiency of Incentive vs. Investment in infrastructure is shown.
- Self Determination Theory.
- Professional Control.
Interviews with Senior leaders
"Plans just throw some money in our way and think we will notice and pay attention. They do not seem to understand that our physicians have deep concerns about what strings are attached. We are always worried about the hidden agenda and what a particular incentive opportunity means for our future."
Attitudes and Responsiveness to Financial Incentives
- Study Setting: Physician network (IPA) in Rochester NY.
- Implemented tournament-style P4P program for diabetes care.
- > 300 PCPs.
- Quality measure: Percentage of expected number of diabetic exams/screens (LDL, 2 HbA1c, urinanalysis, eye exam) conducted.
- Financial incentive: 50 to 150% of withhold payment.
- Potential payout up to about $3,000 for diabetic component.
- Survey of physicians at Baseline.
- Approximately 335 physicians surveyed.
- Approximately 48% response rate.
- No performance differences between respondents and non-respondents.
Measurement of Attitudes
Five-point, multi-item Likert scales
- Autonomy: "The incentive system interferes with my autonomy for how I care for patients." (reverse scaled)
- Goal importance: "This financial incentive is tied to a quality target that is clinically meaningful for diabetic patients."
Image: A line graph labeled "Overview: Six-Year Trends in RIPA Diabetes Care" is shown.
Image: A line graph labeled "Physician Performance Score for Diabetes Pay-for-Performance, 1999-2004" is shown.
Image: A line graph labeled "Autonomy and Physician Performance" is shown.
Image: A line graph labeled "Goal Importance and Physician Performance" is shown.
- The unit of accountability carries possible tradeoffs between infrastructure investment and power of incentives.
- Provider attitudes toward incentive programs may be an important moderator of an incentive program's success. Attitudes among providers toward same incentive program may vary markedly.
- Identify providers with negative attitudes.
- Create opportunities for providers to have input into program design/implementation.