COMMENT & RESPONSEIn Reply We read with interest the letters from Manoj and Singh and from Rehm and colleagues regarding our recent work. 1 It is clear from responses to this study, as well as previous publications, 2 that this is a topic of great interest not just among surgeons but across the profession of medicine and indeed more broadly in society. Many of the comments arising in regards to this article bear similarities to those relating to prior publications that we have addressed previously. 3 We want to emphasize a point made obliquely by both Manoj and Singh and Rehm and colleagues: women are underrepresented in surgery. Almost certainly as a causative factor, and potentially as a result, women in surgery face many training and professional challenges 4 that are not shared by their male colleagues. However, we would not expect that a lower number of women in surgery would lead to disproportionately lower complication rates as suggested by Manoj and Singh.A number of methodologic considerations bear comment. First, in our analysis, we explicitly accounted for fixed procedure-level effects using procedural fee codes. This should abate concerns of Rehm et al regarding differential case mix between male and female surgeons.Before doing so, it is worth reiterating the nature of the dataset; using health administrative data in Ontario, Canada, we are able to capture all interactions between patients and the health care system, including surgical procedures, hospitalizations, emergency department visits, and outpatient care, in addition to capturing patient-and physician-level information in rich detail. This has a number of important implications. First, the use of data from a single-payer health care system provides a comprehensive dataset. Second, the rich data included allow for detailed data regarding patients' acute and chronic health information, health care utilization, and outcomes. While all observational studies are affected by residual confounding, the granularity of data available here exceeds that of other datasets, such as Medicare.Rehm and colleagues note that male surgeons performed a greater proportion of cases in orthopedic and neurosurgery than would be anticipated given the proportion of male surgeons in these subspecialties. Indeed, this reflects data that male surgeons were more likely to be high volume. Given that surgical volume is typically associated with improved patient outcomes, 5 it is our view that residual confounding on account of volume should favor male, not female, surgeons. These authors further note that patients treated by male surgeons were somewhat older and had more comorbidities; this is true and reflects in part differences in the surgical specialties of male and female surgeons. Thus, not only did we account for this in multivariable regression models in keeping with standard methods for case-mix risk adjustment, but we also further accounted for this in our analysis, which used procedure-specific fixed effects.We encourage our colleagues to be open and inquisiti...