Ability of early response to predict discharge outcomes with physical therapy for chronic low back pain.

OBJECTIVE: To determine how well early responses to physical therapy for chronic low back pain predict outcomes at discharge.
METHODS: A prospective single cohort study of chronic low back pain patients seen at a university spine physical therapy clinic. Self-reported pain severity on a 100-mm visual analog scale was measured before each physical therapy visit. The predictive variables were the change in pain score at the second, third, and fourth visits. The first outcome variable was discharge change in pain severity for a correlation analysis. The second outcome variable was a dichotomous variable of whether the subjects had achieved at least a 30% reduction in pain severity at discharge for a discriminant analysis.
RESULTS: Spearman's rank order correlation coefficient showed that early responses at the second (r = 0.324, P = 0.02) third (r = 0.342, P = 0.01), and fourth visits (r = 0.615, P < 0.001) were all significantly correlated with discharge change in pain. The discriminant analysis showed that early responses from the second to fourth visits were able to correctly predict 80.4% of the discharge outcomes (P < 0.001).
CONCLUSIONS: Early responses with physical therapy help predict discharge outcomes for chronic low back pain.