Reliability and reproducibility of clinical findings in low-back pain.

The data obtained from the history and examination of patients with low-back pain and sciatica have been subjected to an observer variation analysis. Redesign of the proformata resulted in a reduction of the observer error from 33% to 18% for the history, but no improvement in the finding of a 30% observer error in the examination. The clinician has a simple choice: if he wishes to obtain a large amount of detailed information he must realize that much of it will be unreliable; if he is prepared to limit the information then he will increase its reliability.