Published in:
01-12-2019 | Care | Original Paper
Effects of pay-for-performance for primary care physicians on diabetes outcomes in single-payer health systems: a systematic review
Published in: The European Journal of Health Economics | Issue 9/2019
Login to get accessAbstract
Background
Although pay-for-performance (P4P) for diabetes care is increasingly common, evidence of its effectiveness in improving population health and health system sustainability is deficient. This information gap is attributable in part to the heterogeneity of healthcare financing, covered medical conditions, care settings, and provider remuneration arrangements within and across countries. We systematically reviewed the literature concentrating on whether P4P for physicians in primary and community care leads to better diabetes outcomes in single-payer national health insurance systems.
Methods
Studies were identified by searching ten databases (01/2000–04/2018) and scanning the reference lists of review articles and other global health literature. We included primary studies evaluating the effects of introducing P4P for diabetes care among primary care physicians in countries of universal health coverage. Outcomes of interest included patient morbidity, avoidable hospitalization, premature death, and healthcare costs.
Results
We identified 2218 reports; after exclusions, 10 articles covering 8 P4P interventions in 7 countries were eligible for analysis. Five studies, capturing records from 717,166 patients with diabetes, were graded as high-quality evaluations of P4P on health outcomes. Based on three quality studies, P4P can result in reduced risk of mortality over the longer term—when linked to performance metrics. However, studies from other jurisdictions, where P4P was not linked to specific patient-oriented objectives, yielded little or mixed evidence of positive health impacts.
Conclusion
Evidence of the effectiveness of P4P depends on whether physicians’ incentive payments are explicitly tied to performance metrics. However, the most appropriate indicators for performance monitoring remain in question. More research with rigorous evaluation in different settings is needed.