Evaluation of Accuracy and Reliability of Free Templating Software for Total Hip Arthroplasty.

Abstract

Preoperative planning in total hip arthroplasty (THA) is crucial for ensuring the availability of appropriate implants and providing surgeons with accurate size expectations. Without accurate templating, surgeons risk making incorrect decisions about implant size, potentially leading to intra-operative fractures, instability, dislocation and early prosthesis failures. While commercial digital templating tools are available, their costs can be prohibitive, especially in resource-limited settings. Open-source alternatives like Detroit Bone Setter offer a potential solution, but their reliability requires rigorous validation. We retrospectively analysed 85 patients who underwent primary THA between 2021 and 2024. Preoperative digital templating was performed using Detroit Bone Setter by two blinded senior orthopaedic surgeons who independently predicted femoral stem and acetabular cup sizes. Templated sizes were compared to intraoperatively implanted components using paired Student's t-tests and Bland-Altman plots. Inter-observer reliability was assessed using intraclass correlation coefficients (ICC). Significant differences were found between templated and actual implant sizes, with a mean bias of + 0.71 mm (p = 0.0003) for femoral stems and + 3.20 mm (p < 0.0001) for acetabular cups. The templated size fell within ± 2 sizes of the actual size in 85% of cases for femoral stems and 65% of cases for acetabular cups. Inter-observer reliability was high (ICC = 0.89 for femoral stems and 0.83 for acetabular cups). These findings underscore the potential of Detroit Bone Setter to serve as a viable low-cost planning tool, especially in resource-constrained surgical environments. Detroit Bone Setter demonstrates reasonable accuracy for femoral stem templating. However, it tends to overestimate acetabular cup sizes. Surgeons should be aware of this tendency and use intraoperative verification techniques to complement templating results.