Does winning a pay-for-performance bonus improve subsequent quality performance? Evidence from the Hospital Quality Incentive Demonstration

Abstract

Objective: To test whether receiving a financial bonus for quality in the Premier Hospital Quality Incentive Demonstration (HQID) stimulated subsequent quality improvement.

Data: Hospital-level data on process-of-care quality from Hospital Compare for the treatment of acute myocardial infarction (AMI), heart failure, and pneumonia for 260 hospitals participating in the HQID from 2004 to 2006; receipt of quality bonuses in the first 3 years of HQID from the Premier Inc. website; and hospital characteristics from the 2005 American Hospital Association Annual Survey.

Study design: Under the HQID, hospitals received a 1 percent bonus on Medicare payments for scoring between the 80th and 90th percentiles on a composite quality measure, and a 2 percent bonus for scoring at the 90th percentile or above. We used a regression discontinuity design to evaluate whether hospitals with quality scores just above these payment thresholds improved more in the subsequent year than hospitals with quality scores just below the thresholds. In alternative specifications, we examined samples of hospitals scoring within 3, 5, and 10 percentage point "bandwidths" of the thresholds. We used a Generalized Linear Model to estimate whether the relationship between quality and lagged quality was discontinuous at the lagged thresholds required for quality bonuses.

Principal findings: There were no statistically significant associations between receipt of a bonus and subsequent quality performance, with the exception of the 2 percent bonus for AMI in 2006 using the 5 percentage point bandwidth (0.8 percentage point increase, p<.01), and the 1 percent bonus for pneumonia in 2005 using all bandwidths (3.7 percentage point increase using the 3 percentage point bandwidth, p<.05).

Conclusions: We found little evidence that hospitals' receipt of quality bonuses was associated with subsequent improvement in performance. This raises questions about whether winning in pay-for-performance programs, such as Hospital Value-Based Purchasing, will lead to subsequent quality improvement.

Keywords: Pay-for-performance; hospitals; incentive; quality improvement.

Figure 1

Receipt of Quality Bonuses by Quality Performance for Incentivized Diagnoses, 2004–2005 Note. Dotted blue vertical line denotes threshold for 1 percent bonus and dashed red vertical line denotes threshold for 2 percent bonus.

Figure 2

Quality Performance around Payout Thresholds for Incentivized Diagnoses, 2005–2006 Note. Dotted blue vertical line denotes threshold for 1 percent bonus and dashed red vertical line denotes threshold for 2 percent bonus. Ninety-five percent confidence intervals shown for observed data.