Cut-off point of the Scoliometer in school scoliosis screening.

Abstract

A large-scale study on school screening for scoliosis was conducted to assess the referral rate, prevalence rate, and positive predictive value using different angles of trunk rotation as criteria for referral. To determine an ideal angle of trunk rotation cut-off point to be used for referral in school screening for scoliosis. When using the Scoliometer (Orthopedic Systems, Inc., Hayward, CA) in school scoliosis screening, 5 degrees and 7 degrees angles of trunk rotation have been recommended as criteria for referral. Low positive predictive values and over-referral at these levels have been reported. The Adams forward bend test and Scoliometer measurement were combined for school scoliosis screening in 33,596 girls from the fifth, sixth and seventh grades. Nurses were the primary screeners. Girls with trunk rotation angles of 5 degrees or more were referred for radiography. The referral rate was 5.2%. By selecting 6 degrees, 7 degrees, 8 degrees, 9 degrees or 10 degrees angles of trunk rotation as criteria for referral, the referral rate became 2.4%, 1.4%, 0.7%, 0.5%, or 0.3%, respectively. The prevalence rate for scoliosis equal to or larger than 10 degrees, 20 degrees, 30 degrees, or 40 degrees of the Cobb angle was 1.47%, 0.21%, 0.04% and 0.02%, respectively, by using a 5 degrees angle of trunk rotation as the criterion for radiography. The positive predictive value was 28.3% for scoliosis of 10 degrees or more, 4% for scoliosis of 20 degrees or more, 0.8% for scoliosis of 30 degrees or more, and 0.4% for scoliosis of 40 degrees or more with a 5 degrees angle of trunk rotation as the criterion for referral. By selecting angles of trunk rotation larger than 5 degrees as criteria for referral for radiography, the positive predictive value increased, but positive cases with larger Cobb angles also decreased markedly. The optimal cut-off point for referral when using the Scoliometer in school screening of scoliosis is still difficult to determine.