Does performance-based remuneration for individual health care practitioners affect patient care?: a systematic review

Abstract

Background: Pay-for-performance (P4P) is increasingly touted as a means to improve health care quality.

Purpose: To evaluate the effect of P4P remuneration targeting individual health care providers.

Data sources: MEDLINE, EMBASE, Cochrane Library, OpenSIGLE, Canadian Evaluation Society Unpublished Literature Bank, New York Academy of Medicine Library Grey Literature Collection, and reference lists were searched up until June 2012.

Study selection: Two reviewers independently identified original research papers (randomized, controlled trials; interrupted time series; uncontrolled and controlled before-after studies; and cohort comparisons).

Data extraction: Two reviewers independently extracted the data.

Data synthesis: The literature search identified 4 randomized, controlled trials; 5 interrupted time series; 3 controlled before-after studies; 1 nonrandomized, controlled study; 15 uncontrolled before-after studies; and 2 uncontrolled cohort studies. The variation in study quality, target conditions, and reported outcomes precluded meta-analysis. Uncontrolled studies (15 before-after studies, 2 cohort comparisons) suggested that P4P improves quality of care, but higher-quality studies with contemporaneous controls failed to confirm these findings. Two of the 4 randomized trials were negative, and the 2 statistically significant trials reported small incremental improvements in vaccination rates over usual care (absolute differences, 8.4 and 7.8 percentage points). Of the 5 interrupted time series, 2 did not detect any improvements in processes of care or clinical outcomes after P4P implementation, 1 reported initial statistically significant improvements in guideline adherence that dissipated over time, and 2 reported statistically significant improvements in blood pressure control in patients with diabetes balanced against statistically significant declines in hemoglobin A1c control.

Limitation: Few methodologically robust studies compare P4P with other payment models for individual practitioners; most are small observational studies of variable quality.

Conclusion: The effect of P4P targeting individual practitioners on quality of care and outcomes remains largely uncertain. Implementation of P4P models should be accompanied by robust evaluation plans.

Primary funding source: None.