Aims The purpose of this study was to determine the effect of radiographic protocol on spinopelvic mobility classification in total hip arthroplasty (THA) and to quantify patient category migration across five classification systems. Methods We retrospectively analyzed 145 consecutive patients (73 female, 72 male; mean age 62.7 years; mean BMI 30.2 kg/m 2 ) undergoing primary THA. Lateral radiographs were obtained in four combinations of standing (unsupported and supported) and sitting (relaxed and forward flexed) positions. Sacral slope was measured using the anterior pelvic plane reference frame. Change in sacral slope (ΔSS) was calculated for each combination and patients were classified using the Dorr/Stefl, Kanawade, Batra, Ranawat, and Sharma systems. Results Mean ΔSS was 19.3° ± 11.3° for unsupported standing to relaxed sitting (USRS) versus −2.6° ± 16.0° for unsupported standing to forward flexed sitting (USFFS). Between USRS and USFFS, reclassification rates were 68.3% (Dorr/Stefl), 64.1% (Batra), 52.4% (Kanawade), 47.6% (Ranawat), and 42.8% (Sharma). McNemar's test confirmed significant asymmetric shifts for Kanawade (χ 2 = 10.4, p = 0.001) and Batra (χ 2 = 6.3, p = 0.012). Cohen's kappa demonstrated near-zero cross-position agreement (κ = −0.05 to 0.08) compared with moderate within-position agreement (κ = 0.40 to 0.64). The sitting position exerted two to three times more influence on classification than the standing position. Conclusion Spinopelvic mobility classification is critically dependent on the radiographic protocol utilized. Up to 68% of patients change category depending on technique. Forward flexed sitting systematically overclassifies patients as stiff through a hip flexion-driven biomechanical artifact. Unlike classifications incorporating global spinal alignment, ΔSS-based systems alone cannot distinguish true spinal pathology from protocol-dependent measurement variation. Standardized protocols utilizing unsupported standing with relaxed sitting are essential.
Cabell et al. (Thu,) studied this question.