LAUSR

To the Editor: Patient satisfaction is a high priority of health systems and is increasingly used as a metric for performance by payers. Physicians are typically ranked on a percentile basis, where minimal changes in absolute patient satisfaction can lead to large changes in physician ranking. We sought to determine physician characteristics associated with higher patient satisfaction in an academic setting. We explored the relationship between top-box rating (5 out of 5), physician sex, location of physician training, seniority, academic rank, faculty line, and time spent doing clinical work in a department of medicine. A total of 19 149 satisfaction scores were obtained for 231 physicians. We also linked data from the Professional Fulfillment Index for a subset of 82 physicians. Professional fulfillment is obtained from the mean of 6 items on a similar Likert-type scale (happiness at work, feeling worthwhile at work, work satisfaction, feeling of control, meaning in work, and professional contribution). To determine those physician characteristics independently associated with higher patient ratings, we performed a backward stepwise selection procedure with the fraction (0% to 100%) of ratings as top box as the dependent variable. Among the 231 physicians, 105 (45%) were female, 58 (25%) were primary care providers, 110 (48%) were clinician educators, 112 (48%) had finished residency training before 2000, 115 (50%) had completed their last clinical training (fellowship or residency) on-site, and 102 (44%) were from a top 20 medical school as defined by US News and World Report. Overall, 84.6% of physician–patient encounters received a top-box rating. In multivariate analysis, physician subspecialty (P = .003), lower percent of time devoted to clinical work (P < .001), and most recent residency on-site (P = .006) were significantly associated with a higher top-box rating. For example, blood and marrow transplant physicians had more top-box ratings by 9.4% than those in immunology/rheumatology. Those doing more than 70% clinical work had lower top-box ratings (−3.8%) compared with those doing less clinical work, and those who underwent residency training on-site had higher top-box ratings by 2.9% than those trained elsewhere. Overall, 20% of the variation in physician topbox ratings was explained by these 3 variables. Sex, seniority, academic rank, faculty line, and location of medical school training were not associated with patient satisfaction after adjustment. Adjusting for professional fulfillment decreased the negative clinical workload effect on patient satisfaction from −3.3% (P = .01) to −2.8%, though it remained borderline significant (P = .04). Our study identified several physician characteristics associated with patient experience. Of the physician characteristics associated with worse patient satisfaction, the one that is most modifiable is a high clinical workload. A high clinical workload may increase pressure to limit time with each patient, which may lead to reduced patient satisfaction. In addition, a high clinical workload may reduce control over one’s time, leading to more “burnout” and reduced “professional fulfillment” and, in turn, lower patient satisfaction. Although our study was not designed to answer this question, our analysis suggests that fulfillment explains part of the relationship between workload and patient satisfaction.