In Reply Drs Himmelstein and Woolhandler suggest that pay-for-performance incentives led clinicians to falsify records. The program we evaluated was designed to align incentives with delivering high-quality care, rather than high-volume care, which is the incentive alignment under fee-for-service, and to pay clinicians more for doing what is difficult, such as controlling blood pressure in a patient with comorbidities.
We have no data specifically on falsification. However, prior literature on Healthcare Effectiveness Data and Information Set measures found that clinicians generally round to the nearest number on a 10 scale (ie, round to 130 mm Hg for a systolic blood pressure measurement of 129 mm Hg or 131 mm Hg).1 In reviewing data from our practices abstracted by a third party,2 we found this rounding behavior in both study groups. Of 1651 visits reviewed, among systolic blood pressure measurements ranging from 129 to 131 mm Hg, the control group (n = 141) percentages for readings were 4.2% at 129 mm Hg, 93.0% at 130 mm Hg, and 2.8% at 131 mm Hg, and in the incentive group (n = 116), they were 1.7%, 91.3%, and 6.9%, respectively.