To verify the accuracy of supine nonweight-bearing radiography versus standing radiography in preoperative planning and to determine the predictors of unintended limb alignment correction in medial open-wedge high tibial osteotomy (OWHTO).

Consecutive patients who underwent medial OWHTO for medial osteoarthritis of the knee with varus alignment were retrospectively reviewed. The analyzed pre- and postoperative radiologic measurements included postoperative mechanical axis deviation (MAD) on standing whole-leg radiographs (WLRs), the predicted value of the postoperative MAD on the preoperative supine (predicted MAD_supine) radiograph, and standing WLRs (predicted MAD_stand). Multiple linear regression analysis was used to identify variables predicting the postoperative MAD and unintended MAD_stand correction, defined as the difference between predicted MAD_stand and postoperative MAD.

Predicted MAD_supine showed statistically greater reliability in predicting postoperative MAD than predicted MAD_stand (intraclass correlation coefficient, 0.82 vs 0.45). Postoperative MAD was correlated with the predicted MAD_supine and the difference in hip–knee–ankle angle between preoperative standing and supine WLRs (ΔHKA angle_stand-supine) (R =763, R² = 0.582, adjusted R² = 0.569, P < .001) and did not differ significantly from the predicted MAD_supine, with a mean difference of 0.28% ± 5.11% (P = .656). The mean unintended MAD_stand correction was 6.52% ± 8.66%. The difference in preoperative MAD between standing and supine WLRs was a significant predictor for unintended MAD_stand correction (β = –0.350, P = .004).

Preoperative planning with supine WLRs can predict postoperative limb alignment in medial OWHTO more accurately than standing radiographs. The clinical significance of the ΔHKA angle_stand-supine for the risk of overestimation of postoperative limb alignment might be low because of the low power in the prediction model and small value of the ΔHKA angle_stand-supine. Preoperative soft-tissue laxity was significantly correlated with unintended correction of postoperative limb alignment resulting from preoperative planning with standing radiographs.

Level IV, retrospective comparative study