Normality and reliability in the clinical assessment of backache.

Backache is common yet its routine medical assessment is imprecise, unreliable, and poorly interpreted. Reproducibility studies on 475 patients improved the reliability of clinical interview and examination in backache, while studies of 335 normal subjects defined the limits of normality. Assessment of nerve function was found to be reliable but assessment of the back itself had to be considerably modified and examination improved by incorporating actual measurements. The validity and clinical utility of the information were analysed to determine the minimum amount of information which should be collected to permit clear diagnosis and management.