Size-Up, Size-Down: Accuracy of Component Sizing with Computerized Tomography and Robotic-Assisted Total Knee Arthroplasty.

Templating prior to total knee arthroplasty (TKA) can help to improve surgical efficiency and potentially improve alignment and outcomes. The purpose of this article is to evaluate the ability of computed tomography (CT)-based preoperative templating to accurately predict implant sizes. A total of 724 Stryker MAKO robotic-assisted TKA cases were retrospectively evaluated from a prospectively collected database between January 2020 and October 2023. Cases were performed by one of three adult reconstruction fellowship-trained orthopaedic surgeons from a health system that includes an academic level one trauma center, an ambulatory surgery center, and a community hospital. Out of the 724 cases, 391 were preoperatively templated independently by the surgeon and the company representative (MAKO Product Specialist [MPS]). The remaining 333 cases were only templated prior to incision by the MPS. Final implant sizes of the tibial and femoral components were compared to preoperative templates. The MPS was able to preoperatively predict the final tibial and femoral implants within one size in 97.2 and 97.8% of cases, respectively. A surgeon and MPS combined preoperative templating increased accuracy to predict the final tibial and femoral implants within one size in 98.9 and 99.5% of cases, respectively. Height and weight were positively correlated with the final implant size ( < 0.001). Non-surgeons can reliably predict implanted components in CT-based preoperative templating in the majority of cases, which is further enhanced by surgeon review and adjustments. In no cases in our series were the final size components implanted greater than two sizes larger or smaller. Our findings suggest that there is opportunity to avoid waste by processing fewer trial implants and transporting fewer components. This would likely decrease overall case cost and improve efficiency in the operating room. Level of evidence: III (retrospective cohort).