A comparison of traditional protractor versus Oxford Cobbometer radiographic measurement: Intraobserver measurement variability for Cobb angles

• Published in: Spine (Philadelphia, Pa. 1976) Pa. 1976), 2005-02, Vol.30 (4), p.440-443
• Main Authors: ROSENFELDT, Michael P, HARDING, Ian J, HAUPTFLEISCH, Jennifer T, FAIRBANK, Jeremy T
• Format: Article
• Language: English
• Subjects: Adolescent; Biological and medical sciences; Cerebrospinal fluid. Meninges. Spinal cord; Confidence Intervals; Female; Fundamental and applied biological sciences. Psychology; Humans; Investigative techniques, diagnostic techniques (general aspects); Lumbar Vertebrae - diagnostic imaging; Male; Medical sciences; Nervous system; Nervous system (semeiology, syndromes); Neurology; Observer Variation; Radiodiagnosis. Nmr imagery. Nmr spectrometry; Radiography - methods; Radiography - statistics & numerical data; Reproducibility of Results; Research Design; Scoliosis - diagnostic imaging; Sleep. Vigilance; Thoracic Vertebrae - diagnostic imaging; Vertebrates: nervous system and sense organs
• ISSN: 0362-2436; 1528-1159
• DOI: 10.1097/01.brs.0000153401.78638.cb

A comparison between measurement of radiographs using a traditional protractor method and the Oxford Cobbometer, which has the potential to reduce error. To assess measurement variability of Cobb angles using the Oxford Cobbometer and to compare it to that of measurements made using the traditional protractor method. Studies of the Cobb method have multiple sources of error and subsequent intraobserver variability. Estimates of intraobserver variability are from 2.8 degrees to 10 degrees. Fifty-three scoliosis curves were measured by 3 examiners. Two measurement sets were performed using the traditional protractor method and two measurement sets performed using the Oxford Cobbometer. For the protractor method, intraobserver variability was 9.01 degrees (95% confidence interval 7.32-10.88). For the Cobbometer method, the value was 5.77 degrees (95% confidence interval 3.25-7.63). The difference between error for construction and Cobbometer methods was significant (P < 0.001). This study demonstrates a lower intraobserver variability for the Oxford Cobbometer compared to the traditional construction method. The Oxford Cobbometer, besides being quick and easy to use, does not require the drawing of lines on films or the use of wide diameter radiographic markers and hence removes some sources of intrinsic error incurred during the traditional method of measuring Cobb angles.