LAUSR

Editor We would like to thank our colleagues Jia et al. for their thoughtful comments on our study of conversion in robotic versus laparoscopic rectal cancer surgery1. Risk factors for conversion depicted by our multivariate analysis were male gender, BMI above 30 kg/m2, stage II and III cancer, and previous abdominal surgery, which can be translated into clinical practice as narrow pelvis, thicker mesentery, advanced tumours and presence of adhesions. We agree that improved visualization and navigation with robotic surgery may overcome some limitations of laparoscopy, and early independent experiences and the latest meta-analysis are concordant with our findings2–4. Baseline characteristics of robotic and laparoscopic groups in our study were not different except for the higher number of stage III patients receiving neoadjuvant radiotherapy in the robotic group, thus potentially representing, if any, a negative selection bias of the robotic group. The non-significantly increased bleeding rate in the robotic group (1⋅6 versus 0⋅3 per cent) can be attributed to the early experience of robotic surgery (inclusion of learning curve cases) with a lower threshold to reactive conversion in that setting. By contrast, the higher number of conversions for advanced tumour stage and adhesions despite advanced experience in the laparoscopic group may reflect the limits of the technique when dealing with complex disease presentations and restricted mobility, demonstrating the advantages of improved exposure with the robotic platform. We acknowledge that patient selection could have affected surgical approach decision-making. However, this is highly doubtful in such a standard practice as ours, with surgeons using either minimally invasive approach as standard for all their cases, thus limiting the risk of selection bias. For these different reasons, choosing propensity score matching over multivariable regression analysis was not deemed necessary to truly reflect the experience of a standardized practice with minimal risk of patient selection. J. Crippa , F. Grass and D. W. Larson Division of Colon and Rectal Surgery, Mayo Clinic, Rochester, Minnesota, USA [email protected] [email protected] DOI: 10.1002/bjs.11599