Physician and patient factors influencing the treatment of low back pain.

Previous retrospective studies have suggested that patient demographics may influence analgesic administration. These studies have not taken physicians' impression of patient pain into account. This prospective study investigates the influence of (i) physician impression of the degree of pain and (ii) patient demographics on the use of analgesic. A convenience sample of adults with non-traumatic lower back pain was studied. Possible predictors of analgesic administration included physician pain scores (assessed by visual analogue scale), patient ethnicity, gender, age, and insurance. These variables were tested individually and then using logistic regression. For the total of 91 patients enrolled, only physician pain scale was found to be associated with analgesic use. Median scores were 68 mm (interquartile range = 62-80 mm) for those receiving treatment versus 48 mm (interquartile range = 30-58 mm) for those who did not (P < 0.001). This study therefore suggests that physician impression of patient pain rather than patient demographics influences analgesic use.