Ultrasound curve angle is more accurate than spinous process angle and can reduce unnecessary radiography in school scoliosis screening

Aims The screening programme for adolescent idiopathic scoliosis (AIS) in Hong Kong results in more than 3,000 children undergoing radiographs of their spines annually. Some of those referred for radiography will have mild curves of Cobb angle < 20° that do not require specialist referral. A previous study suggested that augmenting the current screening programme with ultrasound spinous process angle (SPA) measurement could reduce unnecessary radiography by 51%. The ultrasound curve angle (UCA) was developed as a better method of measuring Cobb angle than the SPA. This study compares the measurement properties of UCA with SPA in determining referral status (referred when Cobb angle ≥ 20°) for specialist consultation in skeletally immature females with AIS. Methods A total of 413 females recommended for radiography by scoliosis screening were recruited from August 2018 to January 2020. Another cohort of 94 females was recruited for validation from January 2020 to January 2023. Ultrasound scanning was done according to a standard procedure at the same setting as biplanar radiography (EOS system). Logistic regression models were used to assess the accuracy of ultrasound in determining referral status. Results In the primary cohort, logistic regression model with UCA achieved a sensitivity of 0.94 and specificity of 0.59 to predict positive referral status. The SPA model achieved a sensitivity of 0.91 and specificity of 0.31. In the validation cohort, the logistic regression model with UCA achieved a sensitivity of 0.95 and specificity of 0.56. The model with SPA resulted in a sensitivity of 0.95 and specificity of 0.19. Combining both cohorts, six cases (1.2%) were screened as false-negative. Conclusion Incorporation of ultrasonography into the scoliosis screening programme could potentially reduce unnecessary radiography by 56% to 59% in skeletally immature female patients with AIS. UCA was shown to be more accurate and have a lower false-negative rate than SPA, indicating the potential for its incorporation into the regular scoliosis screening workflow to reduce unnecessary radiation. Cite this article: Bone Joint J 2025;107-B(11):1220–1228.