Screening for Adolescent Idiopathic Scoliosis is More Accurate When Performed by Healthcare Professionals Compared to Untrained Parents: A Diagnostic Accuracy Study

Background: Many countries ended their professional scoliosis screening due to ongoing controversies. Discontinuation of a screening program results in a shift of screening responsibility from trained healthcare professionals to untrained parents. It is unclear if parents are able to detect scoliosis. Therefore, this study aims to evaluate the diagnostic accuracy of screening when it was performed by untrained parents and compare this with healthcare professionals.  Methods: In this cross-sectional study, parents and healthcare professionals were asked to complete a survey with pictures of 28 children (20 AIS & 8 non-scoliosis). Each slide contained two photographs of one child (standing position & full-flexion during forward-bending test). Lumbar and thoracic curves were represented with a range in severity (10 to >40°). Respondents had to answer whether they detected an abnormality that ought to be referred to a specialist. Diagnostic accuracy measures of the screening were calculated for both groups and for various curve severities. Differences were statically tested with the McNemar test.  Findings: Digital surveys were completed by 100 untrained parents and 122 healthcare professionals. The sensitivity for recognizing scoliosis was significantly lower in untrained parents (63.8%, [95% CI 61.7-65.9%]) compared to healthcare professionals (73.4%, [95% CI 71.6-75.2%]; p 40°) lumbar curves (Δ18.0%), followed by curves between 30-40° (lumbar curves Δ12.7%; thoracic curves Δ10.5%).  Interpretation: The sensitivity of  screening was significantly lower when it was performed by parents, while the false positive rate was similar between both groups. This suggests that scoliosis is less frequently detected when parents have the responsibility of screening and that the window of opportunity for conservative treatment may be missed. Trial Registration: The protocol was registered in the Dutch register for clinical trials (NL5777, https://www.trialregister.nl/). Funding Statement: The authors received no specific funding for this work. Declaration of Interests: All authors declare they have no commercial associations that might pose a conflict of interest in relation to the submitted manuscript. Ethics Approval Statement: Approved by the local Medical Ethics Committee (WO 16.017) before initiation. All individuals provided informed consent before participation in the study; both for using the pictures in the survey and for participating in this survey.