P4P: One review to rule them all

The body of literature remains mixed on P4P programs. Evidence that it is overtly effective remains elusive but would be so…precious…to policy makers.

Pay for performance (P4P) continues to expand as a way to incentivize health care quality and efficiency in the US, the UK, and select countries around the world.  P4P providers receive financial incentives based on measures of clinical quality, resource use, and patient outcomes.  The growth of P4P programs has fed the body of literature evaluating its effectiveness, but variations in financial incentives, program design, and outcome measurement continue to challenge the drawing of meaningful conclusions.

A recent paper reviews  and attempts to draw conclusions regarding efficacy, cost effectiveness, unintended consequences, effect on inequality, financial incentives, and design features driving outcomes. Most studies were conducted in the primary care setting; some focused on acute inpatient care.

Outcome measures on effectiveness focused on prevention and chronic care. Results from strong randomized controlled studies were mixed and provided insufficient evidence to support conclusions for or against the use of P4P.  Non-randomized studies showed small efficacy improvements.

P4P may improve cost-effectiveness, though most studies were of relatively small programs or failed to account for the added administrative costs associated with P4P.

A number of studies found various examples of unintended consequences related to P4P.  These spillover effects include risk-selection by providers, deterring non-compliant patients, referring severely ill patients to other facilities, excluding the severely ill, reductions in continuity of care, and lack of improvement for non-incentivized conditions.

P4P narrowed socioeconomic inequalities in the UK, though inequalities in age, sex, and ethnicity have largely persisted.

Five high quality studies examined the effect of financial incentives on quality improvement in primary care.  These studies revealed financial incentives to be of little benefit unless coupled with other feedback.  These results were further supported by data that combined P4P with public reporting and saw a 2%-4% percent improvement over controls.  While the effects of P4P and public reporting were difficult to discern individually, public reporting likely played an equal, if not greater, role.

Commentary

Despite more research, evidence for P4P programs remains dilute and prohibits definitive conclusions about efficacy, cost effectiveness, unintended consequences, inequalities, financial incentive use, and effective design features.  Most programs are designed for overall quality improvement rather than empirically testing financial incentives.  They have placed priority on “quality now” rather than discovering provider-specific drivers of quality.  They define their program as P4P, but include quality improvement efforts beyond financial incentives.  They measure the cumulative effect of these programs while ignoring confounding and repeatable results. Evidence from these programs provides little instruction to providers seeking to duplicate results and serves little purpose in determining whether financial incentives change provider behavior or drive quality improvement.

What we learn from these studies is that we cannot focus on financial incentives as the sole driver of quality, nor should we.  In fact, performance based compensation may not be necessary at all.  If the end goal is to reduce harm and maximize quality of care, a focus on one aspect of an organization can be detrimental to the whole.  Quality requires a holistic approach.

Eijkenaar, F, et al. Health Policy. 2013 (in press).

by

Patrick Fitzgerald, MPH