Beyond Pay‐for‐Performance: Can Different Feedback and Payment Distribution Methods Affect Team Performance?

Abstract

Research ObjectiveThis study examines how, under team‐based pay‐for‐performance (P4P), managers incentivize physicians for teamwork through internal feedback and payment distribution methods. We evaluate whether these incentive methods affect team performance, drawing on theory of management and behavioral economics in interpreting why some feedback methods may be superior to others in specific contexts.Study DesignThe present research was conducted against the backdrop of Taiwan’s Family Doctor Plan (FDP), a program resembling US patient‐centered medical home initiatives in terms of team‐based care and value‐based purchasing. FDP created primary care teams, each consisting of 5‐10 physician practices, and rewarded the entire team if it achieved certain quality metrics. We developed an instrument and conducted a national survey of managers of 450 teams (response rate 46.9%). In the survey, managers assessed the extents to which their teams: (a) provided continuous, coordinated and comprehensive (3C) care, and (b) improved patient health as compared with the previous year. We then examined whether changes in performance scores could be materially explained by changes in feedback and payment distribution methods between and within groups. Because the choice of internal incentive methods may be correlated with unobserved, group‐specific factors such as team cohesion, we applied multilevel mixed‐effects linear regressions that accommodate endogenous time‐varying covariates. Furthermore, to tackle manager‐related bias, we included manager characteristics in the model and applied the Hausman‐Taylor approach to further accommodate endogenous time‐constant covariates.Population StudiedAfter deleting missing data, 138 primary care teams were included in the final study sample, forming a balanced panel with 276 group‐year observations.Principal FindingsAfter controlling for manager characteristics and other variables, all models tested showed that within teams, when physicians knew their own performances and anonymous peer performances, the team performed better than when physicians knew only their own performances, only team performances, or their own performances plus performances of named peers. Moreover, the results showed that varying methods for distributing performance pay had little effect on team performances.ConclusionsThe results suggest that nonfinancial incentives embedded in feedback are critical in improving team performance especially when financial incentives fail to motivate physicians. However, caution should be taken when applying principles of social comparisons, where public disclosure of identifiable peer performance may have a demotivating effect on lower‐rated physicians when they work in small teams and do not have full control over delivery of 3C care and patient outcomes. In this context, allowing physicians to compare their own performance with anonymous peer performance may be more effective as it appears to trigger low performers’ sense of competitiveness without making them less respected by team members.Implications for Policy or PracticeAdvocacy of P4P or feedback as quality improvement interventions does not typically seem to explicitly address their applications to team‐based models. This paper identifies a cost‐effective and easy‐to‐implement feedback method that is instrumental to health care managers and policy makers for improving team‐based care.Primary Funding SourceMinistry of Science and Technology.