Screening for scoliosis: do we have indications for effectiveness?

Objective The effectiveness of screening for scoliosis has not been established. This study investigated whether patients with adolescent idiopathic scoliosis detected by screening are detected in an earlier stage of the clinical course, and whether these patients have better outcome than otherwise detected patients. Setting The study is a retrospective follow-up study of patients with adolescent idiopathic scoliosis who had completed treatment with a brace, by surgery, or with a brace followed by surgery. Of the 143 patients (born on or after I January 1984) consecutively recruited from 12 hospitals in the Netherlands, 125 (87%) agreed to participate. Of these, 5 1 patients were treated with a brace only and 74 patients were operated on. Screening for scoliosis is carried out in 80% of Dutch children. Methods Data on being screen detected or otherwise detected and Cobb angle at diagnosis were collected using youth health-care files, medical files and interviews by telephone with the patients. Results About 55% of the patients were detected by screening (programme sensitivity). Screen-detected patients had a significantly smaller Cobb angle at diagnosis (28 versus 40 degrees; P < 0.01) and had a 73% lower chance of having had surgery (45% versus 75%; P < 0.01) than otherwise-detected patients. Conclusion In the present study, two essential prerequisites necessary for a screening programme for scoliosis to be effective have been met. However, definite proof of the effectiveness of screening still needs to be established because length bias and over-treatment bias cannot be ruled out using this design.