Reliability Assessment For Idiopathic Scoliosis Cobb Angle Manual Measurements

Objective:To determine the reliability and reproducibility of Cobb angle manual measurements for idiopathic scoliosis using radiographs that had not been premeasured. Methods:Preoperative radiographs (standing full-length posteroanterior, lateral radiogra-phs) of 53 patients were evaluated by four orthopedic surgeons independently on two separate occasions. Reliability and reproducibility was quantified using intraclass correlation coefficient (ICC). Results:Coronal Cobb Angle Measurement of proximal thoracic (PT) demonstrated low to fair intraobserver reproducibility (r=0.592-0.714) and interobserver reliability (r=0.538-0.735). Main thoracic (MT) curves demonstrated good to excellent intraobserver reproducibility (r=0.879-0.942) and interobserver reliability (r=0.919-0.964), while thoracolumbar/lumbar (TL/L) demonstrated fair to excellent intraobserver reproducibility (r=0.649-0.915) and fair interobserver reliability (1=0.915-0.945). Siggital Cobb Angle Measurement of T5-12 (intraobserver r=0.787-0.909; interobserver r=0.911-0.943), T10-L2(intraobserver r=0.857-0.964; interobserver 1=0.915-0.968) and T12-L1 (intraobserver r=0.89-0.983; interobserver r=0.937-0.981) demonstrated good to excellent intraobserver reproducibility and interobserver reliability, and T2-5 (intraobserver r=0.373-0.548, low to fair;interobserver r=0.433-0.527, low) demonstrated low to lesser reliability. Conclusion:The majority of Cobb angle measurements demonstrated good or excellent reliability and reproducibility and can be recommended for routine clinical and academic use, but only fair to good for thoracic kyphosis from T2 to T5 and PT curve,in which some landmarks were not clearly visible. Coronal measures may be more reliable than sagittal measurements.