Comparison of X-ray and Computed Tomography Scan in the Diagnosis and Classification of Tillaux-Chaput Fractures in Adults

Category: Trauma; Ankle Introduction/Purpose: Anterolateral distal tibial fractures, also known as Tillaux-Chaput fractures (TCF), have slowly gained attention, considering the risk of chronic pain and osteoarthritis if not appropriately managed. This study aimed to compare the diagnostic accuracy of ankle radiographs in diagnosing and classifying TCF with computed tomography (CT) scans in the adult population. Methods: A retrospective review of ankle fractures evaluated at a single institution between 2013-2023 was performed. Inclusion criteria were patients aged >18 years and evaluation with ankle radiographs and CT scan. Exclusion criteria were distal tibial/pilon fractures,or prior ankle surgery. A matching 1:2 was performed on patients with and without TCF. A radiologist evaluated the TCF under CT scan and classified the fractures according to Rammelt’s classification: type 1,extraarticular avulsion; type 2,fracture with involvement of the incisura; type 3, fracture with impaction of the anterolateral plafond. His answers were used as a comparison parameter. Three foot and ankle surgeons assessed the radiographs and described if the TCF was present (yes/no). If present, they classified the TCF by subtype. The sensitivity(Sn), specificity(Sp), positive predictive value(PPV), and negative predictive value(NPV) of the radiographs for detecting TCF were calculated. Fleiss's Kappa coefficient measured the interobserver agreement for each subtype. P-values < 0.05 were considered significant. Results: 1134 acute ankle fractures were initially evaluated, of which 481 met the eligibility criteria. The prevalence of TCF was 17.3% (83/481), with 37 (44.6%) type 1, 37 (44.6%) type 2, and 9 (10.8%) type 3. Using radiographs, the overall Sn, Sp, PPV, and NPV for the diagnosis of TCF were 50.2% (95% CI 45.8-54.6), 69.9% (95% CI 65.9-73.9), 62.5% (95% CI 58.3-66.8%), and 58.4% (95% CI 54.1-62.7), respectively. There was disagreement between surgeons using X-rays and the radiologist under CT scan for all subtypes (type 1: κ -0.54, type 2: -0.31, type 3: -0.27, p< 0.001). Under CT scan, the global interobserver agreement was moderate for the TCF classification (κ 0.62) and its subtypes. Conclusion: This study demonstrated that ankle radiographs have low diagnostic accuracy for TCF in adults. Therefore, a CT scan should be routinely used when a TCF is suspected.