A California experiment has found that pay-for-performance can succeed if key elements -- including adequate incentives, uniform performance measures across all payment sources and public reporting of physician performance -- comprise the program.
California Pilot Demonstrates P4P Efficacy
By Leslie Champlin
• Los Angeles
2/14/2006
The program was described during a panel discussion at the National Pay for Performance Summit here Feb. 6-9. The summit was sponsored by the Integrated Healthcare Association.
The California pilot included several elements of AAFP's policy on P4P, including
The California pilot included several elements of AAFP's policy on P4P, including
- a focus on improved quality of care;
- use of evidence-based clinical guidelines;
- involvement of practicing physicians in program design;
- use of reliable, accurate and scientifically valid data;
- use of positive physician incentives; and
- voluntary physician participation.
California's three-year P4P project involved seven commercial health plans and 35,000 physicians. Participating physician groups contributed to development of evidence-based performance measures and then reported on those measures across their total patient populations; insurers paid incentives to practices rather than to individual physicians. Performance ratings for the groups were published online.
Early results showed improvement in clinical processes, patient satisfaction and adoption of electronic health records, said Stephen Shortell, Ph.D., M.P.H., dean and professor of health policy and management and professor of organization behavior at the University of California, Berkeley, School of Public Health and Haas School of Business.
"Clinical improvement was widespread," said Shortell. "Eighty-seven percent of physician groups improved their clinical average (score) by an average of 5.3 percentage points. And that's the average over all the clinical measures, including breast cancer screening, cervical cancer screening, overall asthma care, HbA1c screening and cholesterol screening."
Moreover, an average of 65.7 percent of participating physician groups demonstrated improvement in patients' experiences during those clinical encounters, he added, and adoption of EHRs grew from 74 participating physician groups in 2003 to 119 in 2004.
"And there was a positive, statistically significant relationship between clinical average scores and information technology scores," said Shortell. "So the assumption that IT is relevant to clinical outcome scores is borne out."
That doesn't mean the program didn't encounter some issues, said Cheryl Damberg, Ph.D., senior researcher for RAND Corp., Santa Monica, Calif., a nonprofit institution that evaluated the California experiment and that provides research and analysis to public policy-makers. A RAND study of physician reactions to the program identified specific programmatic necessities.
Damberg found that participating physicians "became more invested" in meeting performance measures and in prospectively gathering data on meeting those measures, but the physicians reported that 1 percent to 1.5 percent incentive payments were inadequate. Incentives should range from 5 percent to 10 percent, said Damberg.
Bart Asner, M.D., CEO of Monarch HealthCare, Irvine, Calif., agreed. "One and a half to 2 percent (incentive pay) does not do it," he said. "That won't meet their return on investment" for EHRs necessary to participate in P4P programs.
At the same time, insurance plans have to offer a sufficient patient base to make P4P participation viable for physicians, said Asner. "One of the key things is that plans have to have enough business to drive through the practices to get (those practices') attention," he said. "If you pay 5 percent as an incentive, but that's only 10 percent of a physician's practice, that's negligible."
Damberg also found that
Early results showed improvement in clinical processes, patient satisfaction and adoption of electronic health records, said Stephen Shortell, Ph.D., M.P.H., dean and professor of health policy and management and professor of organization behavior at the University of California, Berkeley, School of Public Health and Haas School of Business.
"Clinical improvement was widespread," said Shortell. "Eighty-seven percent of physician groups improved their clinical average (score) by an average of 5.3 percentage points. And that's the average over all the clinical measures, including breast cancer screening, cervical cancer screening, overall asthma care, HbA1c screening and cholesterol screening."
Moreover, an average of 65.7 percent of participating physician groups demonstrated improvement in patients' experiences during those clinical encounters, he added, and adoption of EHRs grew from 74 participating physician groups in 2003 to 119 in 2004.
"And there was a positive, statistically significant relationship between clinical average scores and information technology scores," said Shortell. "So the assumption that IT is relevant to clinical outcome scores is borne out."
That doesn't mean the program didn't encounter some issues, said Cheryl Damberg, Ph.D., senior researcher for RAND Corp., Santa Monica, Calif., a nonprofit institution that evaluated the California experiment and that provides research and analysis to public policy-makers. A RAND study of physician reactions to the program identified specific programmatic necessities.
Damberg found that participating physicians "became more invested" in meeting performance measures and in prospectively gathering data on meeting those measures, but the physicians reported that 1 percent to 1.5 percent incentive payments were inadequate. Incentives should range from 5 percent to 10 percent, said Damberg.
Bart Asner, M.D., CEO of Monarch HealthCare, Irvine, Calif., agreed. "One and a half to 2 percent (incentive pay) does not do it," he said. "That won't meet their return on investment" for EHRs necessary to participate in P4P programs.
At the same time, insurance plans have to offer a sufficient patient base to make P4P participation viable for physicians, said Asner. "One of the key things is that plans have to have enough business to drive through the practices to get (those practices') attention," he said. "If you pay 5 percent as an incentive, but that's only 10 percent of a physician's practice, that's negligible."
Damberg also found that
- although early results didn't show that patients migrated to practices with high P4P ratings, public scorecards of practices' performance spurred physicians to meet measures;
- P4P programs should acknowledge that a practice's case mix affects performance and adjust for groups that care predominantly for seriously or chronically ill patients;
- payment methods must be transparent so physicians understand the criteria to be met and the payment that corresponds to meeting each of those criteria; and
- performance measures should be uniform across all plans.
Equally important, said Michael Belman, M.D., staff vice president and medical director for quality management at Blue Cross of California, Woodland Hills, insurers must realize their outlays will increase as physicians comply with performance measures.
"I have to constantly remind people that resource use will increase as we address underutilization of some services," he said. "Once they meet those standards, we'll begin to look at overuse and inappropriate services for efficiency standards."
"I have to constantly remind people that resource use will increase as we address underutilization of some services," he said. "Once they meet those standards, we'll begin to look at overuse and inappropriate services for efficiency standards."
