Development of a clinical prediction rule for the diagnosis of cubital tunnel syndrome in Thai wheelchair users.

Abstract

Purpose: To develop and internally validate a clinical prediction rule (CPR) for diagnosing cubital tunnel syndrome (CuTS) in wheelchair users. To the authors' knowledge, no prior diagnostic CPR for CuTS was developed.

Methods: Participants were wheelchair users with spinal cord injuries/lesions aged ≥20 years. All underwent clinical examinations and electrodiagnostic studies. Clinical endpoint was a confirmed CuTS diagnosis; clinical symptoms and positive electrodiagnostic criteria. The CPR was developed using multivariable logistic regression with backward elimination. Coefficients of the selected predictors were converted into scores by division of the lowest coefficient and then rounded off to the closest integer. Internal validation was performed using bootstrap technique. The model's discriminative ability and calibration performance were evaluated.

Results: Seventy-seven wheelchair users (142 arms) were included, 28(19.7%) arms had CuTS. Multivariable analysis identified three statistically significant predictors for the final diagnostic model: numbness or tingling in the fourth (ulnar half) and fifth fingers, grip weakness and a positive elbow flexion test; "The CuTS-3 Diagnostic Score" (CuTS-3). The CuTS-3 demonstrated good discriminative ability; area under the receiver operating characteristic (AuROC) = 0.88 (95%CI: 0.82-0.95) and calibration. The bootstrap performance adjusted for the estimated optimism for the clinical endpoint was 0.869 (95%CI 0.797-0.946). A cut-off score of ≥2 was suggested for diagnosis as it showed good sensitivity (89.3%) and specificity (78.1%).

Conclusions: The CuTS-3 is a simple and well-performing diagnostic CPR. General clinical use is encouraged to help detecting and providing timely CuTS management. In the future, the model would benefit from external validation in different population.