ObjectiveTo investigate the effect of the number of coaching sessions after communication skills training on the medical communicative performance of oncologists in clinical practice.Methods/DesignThe training, consisting of a workshop and one (control group) vs. four (intervention group) sessions of individual coaching, was evaluated in a randomized controlled trial. Eligible participants included physicians working in any setting where patients with oncological diseases were treated. Real medical consultations were video recorded at three time points: before the workshop (t0), after the workshop (t1) and after completion of coaching (t2). The 1.5-day workshop was based on role-playing in small groups; in the coaching sessions, the videos recorded at t1 were analyzed in detail by both the trainer and the physician. The coaching sessions were manualized and based on the physician’s learning goals. The primary hypothesis was that the intervention group would improve to a higher extent than the control group, as assessed by external raters using rating scales specially developed for this project. Physicians were stratified for sex and setting and randomized by an independent statistician. The group assignment was revealed for physicians and trainers at the end of the workshop, while the raters were blinded to group assignments and assessment points.ResultsA total of 72 physicians participated in one of 8 workshops and could be allocated to either the control or intervention group. The intervention group showed a statistically significant improvement (ES d = 0.41, p<.01) in the All items domain of the rating scales between t1 and t2 and showed a significant advantage compared with the CG (ES = .41, p = .04). The impact on diverse specified skills was heterogeneous; a larger sample is necessary for more detailed analysis.ConclusionsThe training achieved some observable and significant changes in the communicative behavior of oncologists in clinical practice. The four coaching sessions showed some significant advantages compared to the single coaching session. Considerable effort is necessary to achieve sustained changes in communication in clinical every-day practice. Thus, our coaching concept is a promising method for this purpose.