Fiscal Fitness: How States Cope
In California, incentive programs are helping programs make pay-for-performance work
With more than 100 provider pay-for-performance (P4P) programs operating nationally, analysts are beginning to identify fundamental lessons that can be used to grow the effort and make it even more effective. Although far more prevalent in the commercial sector and Medicare, incentive programs also are emerging in Medicaid as a way to improve health care services and outcomes and the Center for Health Care Strategies (CHCS) recently looked at the experiences of seven California Medicaid managed care plans that are implementing incentive programs as the Local Initiative Rewarding Results (LIRR) project funded by the California HealthCare Foundation, one of seven initiatives within the Rewarding Results Program, a national initiative of the Robert Wood Johnson Foundation and the California HealthCare Foundation to test provider performance incentive strategies.
Over the past three years, some 3,300 physicians have been involved in the LIRR project, touching the lives of 350,000 babies, teenagers, and parents.
"While much still needs to be evaluated, pay-for-performance programs represent an opportunity to test whether incentive-based reimbursement can improve the delivery of care for those who need it most," says CHCS senior vice president for program Nikki Highsmith. "Preliminary results from the project show that simple, targeted financial incentives can make a difference."
The key lessons learned to date from LIRR and other health plan incentive programs include:
1. Promoting access and preventive care. Recognizing the first step in delivering health care to Medicaid beneficiaries is simply getting them "in the door," incentives in Medicaid often focus first on access measures such as getting mothers in for pre-natal care and well-baby visits, rather than on clinical measures such as the percentage of members with asthma who have been prescribed a controller medication. LIRR providers targeted incentives for well-baby and adolescent well-care visits, noting the incentive dollars allowed them to do more outreach to patients most in need of care. As incentive programs that initially focus on access and prevention become more sophisticated, rewards for improvements in chronic care and specific clinical outcomes can be added.
2. Engaging providers. Demonstrating that a performance gap exists between actual (what is provided) and ideal (what should be provided) care can be effective in getting providers to participate in P4P projects. With one study indicating that only 33% of physicians receive data on the quality of care they provide, LIRR participant San Francisco Health Plan sends provider payment checks based on goals met, but shows the opportunity for further improvement by also sending a voided check showing the amount the provider could have received had performance been better. Ms. Highsmith reports that for almost all the LIRR health plans, getting the word out about incentive programs and gaining providers' attention required creativity and sustained effort throughout the project.
3. Selecting clear measures. Measures used to evaluate physician performance should be based on solid clinical or practice-based evidence that is accepted by the provider community and for which a change in behavior or practice will result in measurable change, the report says. Standardized measures such as the Health Plan Employer Data and Information Set (HEDIS) often are adopted in P4Pprograms because they are widely used and such measurements make it possible to compare the performance of various organizations. For the LIRR project, HEDIS rates for well-baby and well-adolescent visits were chosen to measure improvement for all the plans. Ms. Highsmith said the downside of using this type of standardized measure is that criterion may not be available to gauge performance on select services or processes. This is particularly true for Medicaid because not all clinical areas and populations are represented in national measurement sets.
4. Paying attention to incentive program structure. Incentive programs should reflect each plan's specific goals and objectives. "In designing an incentive program," the report said, "health plans must decide how to target the clinical outcomes or processes, how to measure improvement, and how to structure payment. The health plans in the LIRR project collectively decided to emphasize HEDIS rates for well-baby and well-adolescent visits, as well as the submission of encounter data." Plans can choose and customize incentives based on a number of criteria including administrative burden to the plan and providers, the plan's ability to estimate payout, whether the plan wants to emphasize relative improvement vs. hard targets, and the degree of control that a physician or practice has in reaching goals. CHCS identified options as a per-service bonus, a tiering bonus, a risk (quality) pool distribution, or a threshold bonus.
5. Being mindful of data challenges. While timely and accurate data on process and outcome improvements are the basis for any incentive program designed to reward quality, collecting such information is not always simple. Health plans implementing incentive programs need to provide enough lead time to test the data collection methodology and also should recognize the cost of collecting data to both providers and the plan.
6. Remembering that money isn't everything. Those implementing incentive programs for Medicaid providers must consider that many publicly financed providers such as county hospitals and clinics may not legally be allowed to accept monetary incentive payments. But that doesn't mean the notion of incentives has to be abandoned entirely. San Francisco Health Plan donated incentive payments to a foundation from which the publicly funded provider could draw. Nonmonetary incentives, including in-kind staff for specific projects, technological equipment, or training, and the referral of new members can be used in lieu of cash. Several LIRR plans offered nonfinancial provider supports while also rolling out financial incentives. Five plans provided feedback to providers on performance, two plans provided in-person consultation to low-performing providers, and three plans provided training on how to submit encounter data. A number of plans assisted providers by notifying them when a member was due for needed care or by directly contacting the member to suggest making an appointment.
7. Considering member incentives. The report said member incentives can play a vital role by encouraging Medicaid consumers to seek necessary care. Member incentives can engage consumers in their own health care and can help address provider concerns about hard to reach and noncompliant members. However, for Medicaid beneficiaries, monetary incentives may be counted as income that could potentially disqualify someone from eligibility. "Therefore," the report said, "Plans need to be aware that even a small incentive could push someone over the income limit." L.A. Care Health Plan, one of the LIRR projects, paid $10 for each series of three well-baby visits and an additional $25 if all six visits were completed, and one movie ticket for each annual well-adolescent visit. The plan was surprised at the low rate of members redeeming the rewards, although there still was a noticeable increase in both well-baby and well-adolescent visits. Plan officials speculate members may have intended to complete the paperwork needed to redeem their reward but forgot to do so. San Francisco Health Plan coupled movie tickets with intensive outreach through advertising in high schools, telephone calling campaigns, etc., to encourage adolescents to visit their doctors. Between 2002 and 2005, the plan saw HEDIS rates for adolescent well-care go to 45.1% from 29.4%.
8. Coordinating with other payers. One major sticking point right now is that many physicians receiving performance reimbursements in California are eligible for quality incentives from more than one health plan or payer. And the various plans and payers often have different measures and different administrative requirements that can make it more difficult for plans to follow and comply with the various requirements.
Ms. Highsmith tells State Health Watch it appears the most crucial of the eight lessons learned is engaging the provider community in the work of developing and operating an incentive program.
"This is something that has to be done up front," she says. "You need to get buy-in on the perspective so providers don't see it as something being done to them." Also especially important are selecting relevant and meaningful measures on which to base incentive payments.
Effective structuring of an incentive program, she says, depends on the plan's goals for the program.
"Goals should be set up front and the program built relative to those goals," Ms. Highsmith says.
As incentive programs continue to develop, Ms. Highsmith says she hopes a consensus can develop on a structure to eliminate some of the disparities that now exist in measures and methods of payment. She recognizes it may not be possible to have a single national program, but thinks regional models may be achievable.
A major challenge facing the pay-for-performance effort in the future is how to deal with the fact that it is the providers at the bottom of the measurements who need the most help. If such providers are penalized under P4P, they are not likely to be able to build the infrastructure needed to improve quality. A way must be found in an incentive structure, she says, to be able to work with the bottom tier providers as well as those in the top tier.
One possible approach, Ms. Highsmith says, is to pay for participation, rewarding infrastructure development or participation in quality improvement initiatives.
"You can incent the fundamental building blocks and then incent outcomes," she says. "Pay-for-performance can be a powerful tool to motivate a focus on quality improvement."
[Download information on the LIRR project from www.chcs.org. Contact Ms. Highsmith at (609) 528-8400.]