Objectives: To determine what behaviors patients perceive as reflecting a physician's humanistic qualities, to develop an instrument for patients to use to assess the humanistic behaviors of their own physicians, and to compare patient assessment of residents' humanistic behaviors with patient satisfaction and the assessment of attending physicians.
Design: Cross-sectional descriptive study, using patient interviews and questionnaires, and evaluations of residents by attending physicians.
Setting: Inpatient medical service in a tertiary care teaching hospital and in a primary care internal medicine clinic.
Participants: Six medical interns and six medical residents, 119 medical patients in the hospital, and 111 patients in the internal medicine clinic.
Measurements and main results: The 25-item Physicians' Humanistic Behaviors Questionnaire (PHBQ) was developed from patients' statements about important humanistic behaviors. The mean PHBQ scores were 4.46 +/- 0.22 (mean +/- SD, on a scale of 1 to 5) in the clinic and 4.18 +/- 18 in the hospital (p = 0.003). The Spearman's rank correlations between the PHBQ and the Medical Interview Satisfaction Scale (MISS) were r = 0.8741 (p < 0.001) in the hospital and r = 0.8751 (p < 0.001) in the internal medicine clinic. The Spearman's rank correlation between the hospital PHBQ and the attending physician evaluations (for the six residents for whom the authors had complete data) was r = 0.5753 (p = 0.232).
Conclusions: Patients can evaluate the humanistic behaviors of their physicians using the PHBQ. There is good correlation between the PHBQ and patient satisfaction, which supports the validity of the PHBQ. The relative lack of agreement between patients and attending physicians suggests different observations, criteria, or standards. The higher ratings from patients in the clinic compared with those from patients in the hospital suggest that residents' behaviors are different or that patients have different observations, criteria, or standards in the two settings. Therefore, a complete assessment of residents' humanistic behaviors may require sampling in both settings.