BACKGROUND: A vast amount of literature on effects of pay-for-performance (P4P) in health care has been published. However, the evidence has become fragmented and it has become challenging to grasp the information included in it. OBJECTIVES: To provide a comprehensive overview of effects of P4P in a broad sense by synthesizing findings from published systematic reviews. METHODS: Systematic literature search in five electronic databases for English, Spanish, and German language literature published between January 2000 and June 2011, supplemented by reference tracking and Internet searches. Two authors independently reviewed all titles, assessed articles' eligibility for inclusion, determined a methodological quality score for each included article, and extracted relevant data. RESULTS: Twenty-two reviews contain evidence on a wide variety of effects. Findings suggest that P4P can potentially be (cost-)effective, but the evidence is not convincing; many studies failed to find an effect and there are still few studies that convincingly disentangled the P4P effect from the effect of other improvement initiatives. Inequalities among socioeconomic groups have been attenuated, but other inequalities have largely persisted. There is some evidence of unintended consequences, including spillover effects on unincentivized care. Several design features appear important in reaching desired effects. CONCLUSION: Although data is available on a wide variety of effects, strong conclusions cannot be drawn due to a limited number of studies with strong designs. In addition, relevant evidence on particular effects may have been missed because no review has explicitly focused on these effects. More research is necessary on the relative merits of P4P and other types of incentives, as well as on the long-term impact on patient health and costs.
There is now a plethora of different quality improvement strategies (QIS) for optimizing health care, some clinician/patient driven, others manager/policy-maker driven. Which of these are most effective remains unclear despite expressed concerns about potential for QIS-related patient harm and wasting of resources. The objective of this study was to review published literature assessing the relative effectiveness of different QIS. Data sources comprising PubMed Clinical Queries, Cochrane Library and its Effective Practice and Organization of Care database, and HealthStar were searched for studies of QIS between January 1985 and February 2008 using search terms based on an a priori QIS classification suggested by experts. Systematic reviews of controlled trials were selected in determining effect sizes for specific QIS, which were compared as a narrative meta-review. Clinician/patient driven QIS were associated with stronger evidence of efficacy and larger effect sizes than manager/policy-maker driven QIS. The most effective strategies (>10% absolute increase in appropriate care or equivalent measure) included clinician-directed audit and feedback cycles, clinical decision support systems, specialty outreach programmes, chronic disease management programmes, continuing professional education based on interactive small-group case discussions, and patient-mediated clinician reminders. Pay-for-performance schemes directed to clinician groups and organizational process redesign were modestly effective. Other manager/policy-maker driven QIS including continuous quality improvement programmes, risk and safety management systems, public scorecards and performance reports, external accreditation, and clinical governance arrangements have not been adequately evaluated with regard to effectiveness. QIS are heterogeneous and methodological flaws in much of the evaluative literature limit validity and generalizability of results. Based on current best available evidence, clinician/patient driven QIS appear to be more effective than manager/policy-maker driven QIS although the latter have, in many instances, attracted insufficient robust evaluations to accurately determine their comparative effectiveness.
A vast amount of literature on effects of pay-for-performance (P4P) in health care has been published. However, the evidence has become fragmented and it has become challenging to grasp the information included in it.
OBJECTIVES:
To provide a comprehensive overview of effects of P4P in a broad sense by synthesizing findings from published systematic reviews.
METHODS:
Systematic literature search in five electronic databases for English, Spanish, and German language literature published between January 2000 and June 2011, supplemented by reference tracking and Internet searches. Two authors independently reviewed all titles, assessed articles' eligibility for inclusion, determined a methodological quality score for each included article, and extracted relevant data.
RESULTS:
Twenty-two reviews contain evidence on a wide variety of effects. Findings suggest that P4P can potentially be (cost-)effective, but the evidence is not convincing; many studies failed to find an effect and there are still few studies that convincingly disentangled the P4P effect from the effect of other improvement initiatives. Inequalities among socioeconomic groups have been attenuated, but other inequalities have largely persisted. There is some evidence of unintended consequences, including spillover effects on unincentivized care. Several design features appear important in reaching desired effects.
CONCLUSION:
Although data is available on a wide variety of effects, strong conclusions cannot be drawn due to a limited number of studies with strong designs. In addition, relevant evidence on particular effects may have been missed because no review has explicitly focused on these effects. More research is necessary on the relative merits of P4P and other types of incentives, as well as on the long-term impact on patient health and costs.
