PURPOSE Physicians’ interruptions have long been considered intrusive, masculine actions that inhibit patient participation, but a systematic analysis of interruptions in clinical interaction is lacking. This study aimed to examine when and how primary care physicians and patients interrupt each other during consultations.

METHODS We coded and quantitatively analyzed interruption type (cooperative vs intrusive) in 84 natural interactions between 17 primary care physicians and 84 patients with common somatic symptoms. Data were analyzed using a mixed-effects logistic regression model, with role, gender, and consultation phase as predictors.

RESULTS Of the 2,405 interruptions observed, 82.9% were cooperative. Among physicians, men were more likely to make an intrusive interruption than women (β = 0.43; SE, 0.21; odds ratio [OR] = 1.54; 95% CI, 1.03-2.31), whereas among patients, men were less likely to make an intrusive interruption than women (β = −0.35; SE, 0.17; OR = 0.70; 95% CI, 0.50-0.98). Patients’ interruptions were more likely to be intrusive than physicians’ interruptions in the phase of problem presentation (β = 0.71; SE, 0.23; OR = 2.03; 95% CI, 1.30-3.20), but not in the phase of diagnosis and/or treatment plan discussion (β = −0.17; SE, 0.15; OR = 0.85; 95% CI, 0.63-1.15).

CONCLUSIONS Most interruptions in clinical interaction are cooperative and may enhance the interaction. The nature of physicians’ and patients’ interruptions is the result of an interplay between role, gender, and consultation phase.