The interobserver reliability of thoracic spinal examination.
The interobserver reliability of thoracic spinal examination was tested in a general practice referral clinic on 97 subjects with any combination of pain in the back in the back, chest, or abdomen or either with no pain. Subjects were examined by a general practitioner and a physiotherapist without knowledge of their history. Kappa coefficients indicated agreement that was generally fair for kyphosis, scoliosis, movement restriction, pain with gross active movements and pain with overpressure at end range. Agreement was moderate for point tenderness, moderate to substantial for regional tenderness and slight for maximal tenderness within a region. Spearman's coefficients for pressure threshold readings for these sites of maximal tenderness showed substantial agreement. Based on examination findings alone, agreement about intervertebral dysfunction was moderate for its presence, fair for its side and exact region and slight for its exact level. The results question the utility of some thoracic spinal signs and demonstrate that the cost of anatomical precision in assessment is decreased reliability.
Australasian Musculoskeletal Medicine
PRE2009-Medical and Health Sciences