Patient Complications After Total Joint Arthroplasty: Does Surgeon Gender Matter?

Talia Ruth Chapman, MD; Benjamin Zmistowski, MD; Kaitlyn Votta, BS; Ayesha Abdeen, MD; James J. Purtill, MD; Antonia F. Chen, MD, MBA


J Am Acad Orthop Surg. 2020;28(22):937-944. 

In This Article


This national database provided 9,284 surgeons who performed at least 20 TKA or THA over the study period. These surgeons combined performed 1,109,370 TKAs and 432,948 THAs. The mean ACR for all surgeons was 2.84% (range: 1.4 to 5.6) and 2.25% (range: 1.1 to 5.7) for total hip and knee arthroplasty, respectively (P < 0.001).

Of the 8,965 surgeons with identified gender, 187 (2.0%; 187 of 8,965) were identified as women and performed 21,216 arthroplasties (1.4%; 21,216 of 1,518,419). Female surgeons were earlier in their career, performed fewer arthroplasties, were less likely to perform THA, and were more likely to practice in New England and Mountain geographic divisions (Table 2).

For THA, women had a mean ACR of 2.78% (range: 2.1 to 3.7) compared with 2.84% (range: 1.4 to 5.6) for men (P = 0.16). For TKA, women also demonstrated a similar mean ACR (2.24%; range: 1.3 to 3.3) compared with men (2.26%; range: 1.1 to 5.7; P = 0.37). When isolating factors that predicted the reported ACR for hips, notable associations between the higher number of THA performed, better medical school rank, fewer years in practice, and geographic region were associated with a lower ACR (Table 3). Similarly, the higher number of TKA performed, better medical school rank, and geographic region predicted lower ACR for TKA. However, in addition to higher procedural volume and geographic region, female surgeons demonstrated a statistically significant lower mean ACR when considering TKA and THA in aggregate (2.41% versus 2.48%; P = 0.005).

Owing to the potential for confounding variables, multivariate analysis was performed. This demonstrated that statistically significant predictors of a lower ACR after lower extremity arthroplasty were independently associated with a higher arthroplasty volume (P < 0.001), total knee versus THA (P < 0.001), fewer years in practice (P = 0.05), and practicing in New England (P = 0.002), Pacific (P < 0.001), and West North Central (P = 0.002) regions compared with East North Central (Table 4). After multivariate analysis, surgeon gender was not a clinically significant variable in predicting patient outcomes (standardized coefficient = 0.04; P = 0.11).