LAUSR

Introduction Using a navigation system in open-wedge high tibial osteotomy (OWHTO) has higher accuracy than using the conventional method. However, unintentional over- and under-correction still exist. This study aimed to compare various factors related to over- and under-correction and to assess their predictive factors in the preoperative radiographs. Materials and methods This study involved 96 knees. The difference in the hip–knee–ankle angle (HKA) between the intraoperative navigation system and postoperative radiograph was termed navigation correction loss (NCL). Knees with absolute values of NCL (|NCL|) ≦ 1.5° and |NCL| > 1.5° were categorised into acceptable ( n  = 46) and outlier ( n  = 50) groups, respectively. The differences in joint line convergence angle (JLCA) between varus and valgus radiographs, varus JLCA, valgus JLCA, standing JLCA and standing HKA were compared between the two groups. Clinical results were evaluated using the American Knee Society (AKS) scores. Results The mean intraoperative HKA in the navigation system was − 3.8 ± 1.8°, and that in the postoperative standing radiograph was − 4.2 ± 2.5° ( p  = 0.033). Preoperative varus, valgus and standing JLCA were higher in the outlier group ( p  = 0.018, p  = 0.020 and p  = 0.001, respectively). Logistic regression analyses for preoperative factors of |NCL| ≦ 1.5° showed that standing JLCA was a determining factor, with an odds ratio of 1.334 (confidence interval was 1.087–1.637, p  = 0.006). AKS score was higher in the acceptable group ( p  = 0.040) postoperatively. Conclusions Higher preoperative standing JLCA was the predictive factor of |NCL| > 1.5°. This factor reduced the rates of under- and over-correction and resulted in better AKS score in OWHTO.