Peer assessment of professional performance by hospitalist physicians

Background: The clinical collaborations among hospitalist physicians create opportunities for peer evaluation. We conducted this study to generate validity evidence for a scale that allows for peer assessment of professional performance. Methods: All of the hospitalist physicians working for >1 year at our hospital were asked to assess each of their physician colleagues along eight domains and name three colleagues whom they would choose to care for a loved one needing hospitalization. A mean composite clinical performance score was generated for each provider. Statistical analyses using the Pearson coefficient were performed. Results: The 22 hospitalist physician participants were confident in their ability to assess their peers' clinical skills. There were strong correlations between the domains of clinical excellence (r > 0.5, P < 0.05). Being selected as a doctor whom colleagues would choose to take care of their loved ones was highly correlated with high scores in the domains of humanism, diagnostic acumen, signouts/handoffs, and passion for clinical medicine, and higher composite clinical performance scores (all r > 0.5, P < 0.05). High scores on the Press Ganey questions correlated with peer assessment of humanism (r = .78, P = 0.06). Conclusions: The correlation among scale items, the composite clinical performance score, and the variable "a doctor whom you would choose to care for a loved one" provides validity evidence to our assessment scale. Such measurements may allow hospitalist groups to identify top performers who could be recognized, rewarded, and held up as role models and weaker performers who may need focused training or remediation.