Study of the diagnostic accuracy and interexaminer reliability of scoliosis diagnostic tests.
To estimate the sensitivity, specificity, and predictive value of the Scoliometer (National Scoliosis Foundation, Watertown, MA) and Adam's forward bend test in diagnosing scoliosis, and to determine the interexaminer reliability of the Scoliometer and Adam's forward bend test.
Exposure to diagnostic radiation in patients with adolescent idiopathic scoliosis may result in a small but significant increase in cancer rates. The full-spine radiographic examination remains the standard procedure for the assessment of scoliosis. There is a need for a valid and reliable noninvasive test to assess scoliosis.
Two examiners independently assessed 105 patients presenting to a scoliosis clinic for trunk asymmetry with Adam's forward bend test and axial trunk rotation with the Scoliometer. The Cobb method served as the gold standard.
The interexaminer agreement for the Scoliometer is excellent in the thoracic spine and substantial in the lumbar spine. The interexaminer measurement error shows poor precision for thoracic and lumbar Scoliometer measurements. The interexaminer agreement for Adam's forward bend test is substantial in the thoracic spine and poor in the lumbar spine. Adam's forward bend test is more sensitive than the Scoliometer in detecting thoracic curves measuring 20° or more by the Cobb method. Receiver operating characteristic curve analysis suggests that the use of the Scoliometer marginally improves the ability of diagnosing a scoliosis in the thoracic spine.
The Scoliometer and Adam's forward bend tests have adequate interexaminer reliability for the assessment of thoracic curves. The Scoliometer has better interexaminer agreement in the lumbar spine. However, the Scoliometer has a high level of interexaminer measurement error that limits its use as an outcome instrument. Because Adam's forward bend test is more sensitive than the Scoliometer, the authors believe that it remains the best noninvasive clinical test to evaluate scoliosis.
From the *Institute for Health and Outcomes Research, Department of Physical Medicine and Rehabilitation, and the †Department of Surgery, University of Saskatchewan, Saskatoon; and ‡Private Chiropractic Practice, Montreal, Quebec, Canada.
Supported by the Chiropractors' Association of Saskatchewan.
Acknowledgment date: August 13, 1996.
Acceptance date: July 29, 1997.
Device status category: 9.
Address reprint requests to: J. David Cassidy, DC, PhD, FCCSC; Director of Research; Institute for Health and Outcomes Research; PO Box 108, Royal University Hospital; 103 Hospital Drive; Saskatoon, Saskatchewan; Canada S7N 0W8.