Double-level osteotomy versus high tibial osteotomy for varus malalignment of the knee: A systematic review and meta-analysis.

Purpose

To compare double-level osteotomy (DLO) with high tibial osteotomy (HTO) for varus knee malalignment across patient-reported, radiographic, operative and complication outcomes.

Methods

PubMed, Scopus and Web of Science were searched to 15 May 2026 for comparative DLO versus HTO studies. Outcomes were pooled with a random-effects model as mean differences (MD) or risk ratios (RR) with 95% confidence intervals (CI), with absolute values for each arm. The primary outcome was joint-line obliquity (JLO). Sensitivity analyses (leave-one-out, one cohort per centre, comparable indications) were post hoc.

Results

Nine observational studies (805 knees; 306 DLO, 499 HTO) were included. At a mean follow-up of 3 years, postoperative JLO was lower after DLO than after HTO (four studies; 1.6° vs. 5.9°; MD: -4.14°, 95%CI: -5.22, -3.06; I2 = 76%). DLO produced a lower MPTA (seven studies; 91.3° vs. 95.4°; MD: -4.10°, 95%CI: -5.30, -2.90), LDFA (four studies; 86.0° vs. 90.2°; MD: -4.04°, 95%CI: -4.49, -3.58) and mLDFA (two studies; 85.6° vs. 89.0°; MD: -3.37°, 95%CI: -4.16, -2.58). The hip-knee-ankle angle difference was small (five studies; MD: -0.96°, 95%CI: -1.83, -0.09) and not robust in sensitivity analyses. No difference was demonstrated for joint-line convergence angle, posterior tibial slope, any KOOS subscale, KSS scores or time to return to sport. Operative time was longer with DLO (two studies; 137 vs. 91 min; MD: 42.9, 95%CI: 25.7, 60.2). Intraoperative hinge fracture was less frequent with DLO (three studies; 14.0% vs. 34.6%; RR: 0.41, 95%CI: 0.23, 0.73). No nonunion was reported.

Conclusions

At 3 years, no significant difference in patient-reported or functional outcomes was demonstrated between DLO and HTO. DLO achieved correction with less JLO and fewer hinge fractures, at the cost of a longer operation. Because DLO was often selected for bifocal or larger deformities, the radiographic differences partly reflect the indication. These hypothesis-generating findings require prospective confirmation. LEVEL OF EVIDENCE Level I.