Validation of a web-based questionnaire for shoulder disorder diagnosis and staging

Background: Shoulder pain is one of the most common musculoskeletal disorders globally, often requiring prolonged care and resulting in high healthcare costs. Telemedicine and remote rehabilitation show promise in managing these conditions, but standardized and reliable assessment tools remain limited. This study aimed to validate a novel web-based, self-administered questionnaire for estimating disease phases and specific diagnoses of shoulder disorders. Methods: A cross-sectional study was conducted with 30 participants reporting shoulder pain or restricted shoulder movement. Each participant completed a web application consisting of 33 symptom- and movement-related questions. The application guided users through the assessment using written and video instructions. Two experienced physical therapists independently evaluated each participant. In cases of disagreement, they discussed until reaching a consensus. Agreement rates between the web application and the therapists' consensus diagnoses were calculated in 4 domains: (1) classifying cases as shoulder-related vs. non-shoulder-related conditions, (2) classifying disease phase (inflammatory vs. noninflammatory), (3) identifying specific disorders, and (4) agreement across all 3 criteria. Results: The application accurately distinguished shoulder disorders from other conditions in 100% of cases. Among 27 participants with shoulder disorders, the disease phase was correctly classified in 92.6%, and specific diagnoses in 74.1%. Overall agreement across all 3 domains was 70%. Rotator cuff injuries were the most common diagnosis (70%). Conclusion: The web application showed high diagnostic agreement with experienced physical therapists and may be a useful screening and diagnostic tool for shoulder disorders in both clinical and remote settings. Larger studies are needed to confirm its utility and improve the algorithm.