Preliminary development of a clinical prediction rule for determining which patients with low back pain will respond to a stabilization exercise program

Objective: To develop a clinical prediction rule to predict treatment response to a stabilization exercise program for patients with low back pain (LBP). Design: A prospective, cohort study of patients with nonradicular LBP referred to physical therapy (PT). Setting: Outpatient PT clinics. Participants: Fifty-four patients with nonradicular LBP. Intervention: A standardized stabilization exercise program. Main Outcome Measure: Treatment response (success or failure) was categorized based on changes in the Oswestry Disability Questionnaire scores after 8 weeks. Results: Eighteen subjects were categorized as treatment successes, 15 as treatment failures, and 21 as somewhat improved. After using regression analyses to determine the association between standardized examination variables and treatment response status, preliminary clinical prediction rules were developed for predicting success (positive likelihood ratio [LR], 4.0) and failure (negative LR, .18). The most important variables were age, straight-leg raise, prone instability test, aberrant motions, lumbar hypermobility, and fear-avoidance beliefs. Conclusions: It appears that the response to a stabilization exercise program in patients with LBP can be predicted from variables collected from the clinical examination. The prediction rules could be used to determine whether patients with LBP are likely to benefit from stabilization exercises.