Diagnostic accuracy of patient history and clinical examination to identify pathology in the lumbar spine of elite gymnasts with MRI as the reference standard.

OBJECTIVE: To evaluate the diagnostic accuracy of patient history and physical examination for diagnosing common causes of low back pain in elite gymnasts, using MRI as the reference standard. DESIGN: Prospective cross-sectional study reported according to STARD guidelines. Index tests were acquired by an experienced musculoskeletal physiotherapist blinded to imaging. MRI scans were independently analysed by two specialist musculoskeletal radiologists. SETTING: MRI centre, Birmingham, UK. PARTICIPANTS: Non-probability consecutive total-population sample of elite gymnasts, regardless of low back pain status. MAIN OUTCOME MEASURES: Sensitivity and specificity of individual index tests and exploratory test clusters. RESULTS: Forty participants consented; 75% had abnormal MRI findings. Most specific findings were 'eased by flexion' and 'erector spinae hypertrophy' for degenerative disc disease, 'erector spinae hypertrophy', 'pain on flexion in standing', and 'pain on extension/left side flexion' for disc herniation, 'pain on flexion in standing' plus 'pain on extension/left side flexion' for pars oedema/spondylolysis; and 'eased by flexion' with the same cluster for spondylolisthesis. CONCLUSIONS: Highly specific index tests and exploratory clusters showed diagnostic accuracy for identifying gymnasts without degenerative disc disease, disc herniation, and pars oedema/spondylolysis. All gymnasts were recruited regardless of low back pain; these findings need to be considered cautiously when applied clinically.