Summary of Talk:
Background: Pay-for-performance (P4P) has become a popular provider side incentive applied by purchasers or insurers to improve the performance of health providers. There are several published system reviews evaluating the effects of P4P on health providers’ performance, but few could get confirmed and generalization conclusion from the existing studies. It was mentioned by several authors that P4P requires very careful design, but they just discussed different incentives on theory. This study reviewed studies evaluating effects P4P for primary health providers and analyzed the P4P design elements which might affect the size of effects of P4P in order to explain the differences in the effects of different P4P payments.
Method: We applied Cochrane review methods and searched 11 databases. To be included, the target of payment methods must be primary health facilities, primary care physicians or physician groups; and the study had to report changes in health provision process outcome or patient outcomes. In order to analyze the impacts of P4P design elements, we grouped the P4P payments by the following factors: level of payment targeting, type of performance measures, type of performance target, size of incentive, frequency of monitoring and feedback, frequency of payment, resourcing, and duration of follow-up.
Result: 48 studies were finally included. We found that payment directly targeting individual health workers, use of process performance measures, payment for each instance of performance measure, and quarterly payment were more likely to result in significant effectiveness. Higher proportion of incentives compared with total income, and allocation of facilities' P4P payment based on individuals' performance may lead to little higher effect size of P4P incentives.
Conclusion: If policy makers intend to apply P4P incentive to pay primary health providers, our review found that this intervention will probably lead to the improvement in service process outcomes, but will lead to little or no improvement in patients' utilization or health outcomes. We also suggest that when policy makers intend to apply P4P, or if they found that their P4P program did not result in expected effectiveness, they could analyze each component of their P4P design, especially the performance measures choice, performance target design, payment frequency, if the payment level being sufficient to change the behaviors of health providers, and how the payment really relating to individual performance.
YUAN, Beibei, Associate Professor in Peking University China Center for Health Development Studies. She continued the research on evidence-based method in the areas of health system and policies, including research on appropriate methodology on evidence synthesis and systematic reviews for health system and policy researches; conducting of systematic reviews on specific health system and policy research topics, like strategies for expanding health insurance coverage, payment methods evaluation; and teaching the courses related to systematic review or evidence synthesis. In addition her research interests also include incentives, work behaviors and performance of health workers in primary health facilities.