Purpose or Case Report: Children with suspected developmental dysplasia of the hip (DDH) often undergo both anteroposterior (AP) and frog-leg lateral (FL) radiographs of the hips to derive their associated acetabular indices (AI) in gauging the maturity and development of their hips. Unfortunately, when the AP view is suboptimally positioned, repeat imaging is often necessary, resulting in additional radiation exposure. This study aims to determine whether FL radiographs are equivalent to AP radiographs for measuring AI in children being evaluated for suspected DDH.
Methods & Materials: This Institutional Review Board–approved retrospective study included children aged 6–24 months who underwent bilateral hip AP and FL radiographs at a large tertiary pediatric hospital (01/2024-02/2024). Patients with a history of hip surgery or skeletal dysplasia were excluded. We also excluded radiographs considered excessively tilted or rotated (defined as not meeting a rotation index of 0.5–2.0 and a tilt index of 0.9–1.4 on AP view). AP and FL views were separated and randomized, and the AI were measured independently in a blinded fashion by three attending pediatric radiologists and one attending pediatric orthopedic surgeon. A linear mixed-effects model was used to estimate the mean difference in the AI made between the AP and FL radiographs. The AI made based on the AP and FL radiographs were considered equivalent when the 90% Wald confidence interval (CI) for their difference lay within a prespecified ±2° equivalence margin. Inter-reader agreement within each view was assessed using a two-way random-effects, absolute-agreement, and single-measure intraclass correlation coefficient (ICC) with 95% CIs.
Results: Thirty patients (age range=6.1–22.9 months, mean=12.2±4.6 months; 17 females) with 60 hips, all with paired AP and FL radiographs, comprised the study population. Four readers generated a total of 480 measurements of AI. In the primary mixed-effects analysis, the mean difference of the AI (AP-FL) was 0.20° (90% CI, −0.12°–0.52°), and the entire 90% CI was within the prespecified ±2° equivalence margin. Inter-reader agreement within each view was moderate to good: the ICC was 0.73 (95% CI, 0.61–0.82) for AP and 0.69 (95% CI, 0.57–0.78) for FL.
Conclusions: AP and FL radiographs provide equivalent measurements of AI in children aged 6–24 months. This has important clinical implications as repeat AP imaging in the setting of a suboptimal AP view may be unnecessary when a FL view is available.