EOS Imaging is Accurate and Reproducible for Preoperative Total Hip Arthroplasty Templating

Background: Templating is a critical part of preoperative planning for total hip arthroplasty (THA). The accuracy of templating on images acquired with EOS is unknown. This study sought to compare the accuracy and reproducibility of templating for THA using EOS imaging to conventional digital radiographs. Methods: Forty-three consecutive primary unilateral THAs were retrospectively templated, six months postoperatively, using preoperative 2D EOS imaging and conventional radiographs. Two blinded observers templated each case for acetabular and femoral component size and femoral offset. The retrospectively templated sizes were compared to the sizes selected during surgery. Interobserver agreement was calculated, and the influence of demographic variables was explored. Results: EOS templating predicted the exact acetabular and femoral size in 71% and 66% of cases, respectively, and to within one size in 98% of cases. The acetabular and femoral component size was more likely to be templated to the exact size using EOS compared to conventional imaging (P < .05). The femoral component offset choice was accurately predicted in 83% of EOS cases compared to 80% of conventional templates (P = .341). Component size and offset were not influenced by patient age, gender, laterality, or BMI. Interobserver agreement was excellent for acetabular (Cronbach's alpha = 0.94) and femoral (Cronbach's alpha = 0.96) component size. Conclusions: Preoperative templating for THA using EOS imaging is accurate, with an excellent inter-observer agreement. EOS exposes patients to less radiation than traditional radiographs, and its three-dimensional applications should be explored as they may further enhance preoperative plans. (C) 2020 Elsevier Inc. All rights reserved.