Interruption in the medical interaction.

OBJECTIVES To describe patterns of interruption in the physician-patient interaction; specifically, to determine who interrupts, to determine if a shift in control occurs as a result of interruption, and to characterize the information gained when patients interrupt physicians. DESIGN Observational. SETTING Community-based primary care practices in North Carolina. SUBJECTS Internists and family physicians in private practice in North Carolina (six men and two women) and their patients (13 men and 27 women). INTERVENTIONS None. MAIN OUTCOME MEASURE The obtainment of control of the interaction, at least momentarily, as a result of interruption. RESULTS Analysis of 40 audiotaped interactions revealed 833 interruptions (mean +/- SD, 20.8 +/- 12.2 per interaction): Patients initiated 55% of all interruptions. Physicians and patients each gained control of the conversation after 50% of interruptions. Patients gained control after 74% of patient-initiated interruptions, and physicians gained control after 79% of physician-initiated interruptions. Patients were more likely to gain control by interrupting late in the interaction, and 75% of patient-initiated interruptions resulted in new information (solicited and unsolicited) being contributed to the interaction. CONCLUSION Interruption by patients can be an informative event.