Physician reimbursement by salary or fee-for-service: effect on physician practice behavior in a randomized prospective study

Abstract

We used a resident continuity clinic to compare prospectively the impact of salary v fee-for-service reimbursement on physician practice behavior. This model allowed randomization of physicians into salary and fee-for-service groups and separation of the effects of reimbursement from patient behavior. Physicians reimbursed by fee-for-services scheduled more visits per patient than did salaried physicians (3.69 visits v 2.83 visits, P less than .01) and saw their patients more often (2.70 visits v 2.21 visits, P less than .05) during the 9-month study. Almost all of this difference was because fee-for-service physicians saw more well patients than salaried physicians (1.42 visits and .99 visits per enrolled patient, respectively, P less than .01). Evaluating visits by American Academy of Pediatrics' guidelines indicated that fee-for-service physicians saw more patients for well-childcare than salaried physicians because they missed fewer recommended visits and scheduled visits in excess of those recommendations. Fee-for-service physicians also provided better continuity of care than salaried physicians by attending a larger percentage of all visits made by their patients (86.6% of visits v 78.3% of visits, P less than .05), and by encouraging fewer emergency visits per enrolled patient (0.12 visits v 0.22 visits, P less than .01). Physicians' interest in private practice, as determined by their career plans, correlated significantly with total number of patients enrolled (r = .48, P less than .05) and total clinic patients seen by each resident during the study (r = .40, P less than .05).(ABSTRACT TRUNCATED AT 250 WORDS)