Abstract Background Carpal tunnel syndrome (CTS) is the most common entrapment neuropathy. Ultrasound assessment of the median nerve inlet-to-outlet area ratio (IOR) has demonstrated diagnostic value; however, commonly used fixed thresholds are often arbitrary and may not reflect normal population variability. Aim To derive percentile-based normative reference values for median nerve IOR in healthy controls and evaluate the diagnostic performance of percentile-derived thresholds for CTS. Patients and methods This secondary exploratory cross-sectional analysis of a publicly available, de-identified dataset (Dryad; doi: https://doi.org/10.5061/dryad.jt2sm ) included 90 participants (44 healthy controls and 46 electrodiagnostically confirmed CTS cases). Median nerve cross-sectional area was measured at the carpal tunnel inlet (pisiform level) and outlet (hamate level), and IOR was calculated as inlet/outlet. Distribution-free percentiles (5th–99th) were generated from the control group to establish reference limits. Diagnostic performance of percentile-based thresholds was evaluated using receiver operating characteristic (ROC) analysis, with optimal discrimination assessed using Youden’s index. Results The 95th percentile control threshold (IOR > 1.318) provided the most balanced diagnostic performance, yielding sensitivity 91.3%, specificity 93.2%, positive predictive value 93.3%, negative predictive value 91.1%, and overall accuracy 92.2%. Discrimination was excellent (AUC 0.984; 95% CI 0.962–1.000). A percentile-anchored interpretation framework enabled graded risk stratification (normal 95th). Findings represent exploratory performance within this dataset and require external validation. Conclusion Percentile-anchored IOR thresholds provide a statistically grounded alternative to fixed cutoffs for ultrasound assessment of CTS, aligning interpretation with reference interval methodology and supporting structured risk stratification. External validation in larger, multi-center, and demographically diverse populations is required before universal clinical implementation.
Mohamed Safwat Hamza (Wed,) studied this question.