RESUMO
OBJECTIVE: To assess the association between the frequency of empathic physician responses with patient anxiety, ratings of communication, and encounter length during hospital admission encounters. DESIGN: Analysis of coded audio-recorded hospital admission encounters and pre- and postencounter patient survey data. SETTING: Two academic hospitals. PATIENTS: Seventy-six patients admitted by 27 attending hospitalist physicians. MEASUREMENTS: Recordings were transcribed and analyzed by trained coders, who counted the number of empathic, neutral, and nonempathic verbal responses by hospitalists to their patients' expressions of negative emotion. We developed multivariable linear regression models to test the association between the number of these responses and the change in patients' State Anxiety Scale (STAI-S) score pre- and postencounter and encounter length. We used Poisson regression models to examine the association between empathic response frequency and patient ratings of the encounter. RESULTS: Each additional empathic response from a physician was associated with a 1.65-point decline in the STAI-S anxiety scale (95% confidence interval [CI], 0.48-2.82). Frequency of empathic responses was associated with improved patient ratings for covering points of interest, feeling listened to and cared about, and trusting the doctor. The number of empathic responses was not associated with encounter length (percent change in encounter length per response 1%; 95% CI, -8%-10%). CONCLUSIONS: Responding empathically when patients express negative emotion was associated with less patient anxiety and higher ratings of communication but not longer encounter length.