Chronic low back pain: predicting response to nonsurgical treatment.

The effectiveness of several measures in predicting response to conservative treatment for chronic low back pain was studied. The measures included pretreatment demographic variables, the Minnesota Multiphasic Personality Inventory (MMPI), patient ratings of pain caused by daily activities as measured by the Activities Discomfort Scale (ADS), and physician ratings. The subjects consisted of 135 patients treated at a university hospital back clinic who completed questionnaires six months or more after the initiation of treatment. The only significant predictor of how successful treatment had been in relieving pain was the MMPI hypochondriasis scale. The best predictor of patient-rated average pain intensity at follow-up was the ADS. Physicians' predictions of pain relief and return to normal activities, the ADS and the hypochondriasis scale each was associated significantly with patient ratings of return to normal activities after treatment, and to about the same extent. In general, statistical combinations of pretreatment variables did not increase the power of prediction. The ADS appears to be a promising predictor of patient-rated pain intensity after treatment, and may be useful in identifying poor risk patients.