Abstract
ICLB15 - The Preservation Scorecard: Fundamentals of Tracking Outcomes and Standardizing Recovery, Session Room 12 (BMCC Level 3), October 9, 2026, 14:00 - 16:00
Introduction
Computed tomography (CT) is commonly used to assess tibial torsion in patients with lower-limb rotational abnormalities. However, several measurement techniques have been described, using different proximal and distal reference axes. This methodological variability may influence absolute torsion values and may alter whether patients are classified as having excessive external tibial torsion. The purpose of this study was to compare 4 CT-based methods for measuring tibial torsion and to evaluate their impact on classification using a 40° threshold.
Methods
This retrospective imaging study included 120 lower limbs in 102 patients who underwent lower-limb CT rotational assessment. Four measurement methods were evaluated by combining 2 proximal reference axes—the posterior femoral condylar axis (FCA) and the posterior tibial condylar axis (TCA)—with 2 distal reference axes—the distal tibial plafond axis and the bimalleolar axis. The 4 methods were FCA–Plafond, FCA–Bimalleolar, TCA–Plafond, and TCA–Bimalleolar. Pairwise comparisons, linear regression, Bland–Altman analysis, and intraobserver/interobserver reliability analyses were performed. Excessive external tibial torsion was defined as ≥40°.
Results
Mean tibial torsion differed significantly according to the measurement method used: 42.3° ± 6.2° for FCA–Plafond, 36.5° ± 6.6° for FCA–Bimalleolar, 38.8° ± 6.3° for TCA–Plafond, and 33.2° ± 6.5° for TCA–Bimalleolar. All pairwise comparisons were significant (P < .001). The largest difference was observed between FCA–Plafond and TCA–Bimalleolar, with FCA–Plafond yielding values 9.15° higher on average. Using a ≥40° threshold, the proportion of limbs classified as having excessive external tibial torsion varied markedly by method: 70.7% for FCA–Plafond, 44.0% for TCA–Plafond, 32.5% for FCA–Bimalleolar, and 20.5% for TCA–Bimalleolar (Cochran Q test, P < .001). Interobserver reliability was good for all methods, with ICCs ranging from 0.808 to 0.875.
Conclusion
CT-based tibial torsion measurements are highly method-dependent. The choice of proximal and distal reference axes substantially affects absolute torsion values and the proportion of patients classified as having excessive external tibial torsion. Measurement methods should therefore be explicitly reported, and threshold values should not be considered interchangeable across techniques.