Supine Versus Standing Radiographs for Detecting Ischiofemoral Impingement: A Propensity Score-Matched Analysis.

Abstract

The purpose of this study was to assess parameters of ischiofemoral impingement on supine and standing anteroposterior hip radiographs and to suggest optimal cutoff points for detection of ischiofemoral impingement. A retrospective study included patients with a clinical history of hip pain. All hip joints with evidence of quadratus femoris muscle edema on MR images were included in the ischiofemoral impingement patient group. An age- and sex-matched control group was derived from the same cohort by propensity score matching. On radiographs, two readers independently measured the following parameters: ischiofemoral space, quadratus femoris space, hamstring tendon area, ischiofemoral distance on supine radiograph, ischiofemoral distance on standing radiograph, and femoral neck-shaft angle. Group differences in parameters were assessed by Mann-Whitney U test. The intraclass correlation coefficient and the ROC AUC were obtained. Correlations between radiographic and MRI measures were assessed with Pearson correlation and Bland-Altman plot analyses. The Youden J index was used to select optimum cutoff points for each parameter. There were 30 patients (44 hip joints; mean age, 54.8 ± 11 years) in the ischiofemoral impingement group and 88 patients (88 hip joints; mean age, 51.8 ± 13.4 years) in the control group. There were significant group differences in ischiofemoral space, quadratus femoris space, ischiofemoral distance on supine radiograph, ischiofemoral distance on standing radiograph, and neck-shaft angle (p < 0.05). Ischiofemoral distance on supine and standing radiographs exhibited good discriminative ability (AUC > 0.80). The optimal cutoff points for ischiofemoral distances on supine and standing radiographs were 19.9 and 19.1 mm for reader 1 and 21.1 and 17.0 mm for reader 2. Ischiofemoral space, quadratus femoris space, ischiofemoral distance on supine radiograph, and ischiofemoral distance on standing radiograph exhibited nearly perfect interobserver agreement (r > 0.8). Ischiofemoral distances on supine and standing hip radiographs had good diagnostic performance and can be used as a screening tool, with optimal cutoff points.