Abstract
OBJECTIVE: To identify potential prognostic factors that may predict clinical improvement of patients treated with different physical therapy interventions in the short-term. METHODS: This is a prospective cohort study. A total of 616 patients with chronic non-specific low back pain treated with interventions commonly used by physical therapists were included. These patients were selected from five randomized controlled trials. Multivariate linear regression models were used to verify if sociodemographic characteristics (age, gender, and marital status), anthropometric variables (height, body mass, and body mass index), or duration of low back pain, pain intensity at baseline, and disability at baseline could be associated with clinical outcomes of pain intensity and disability four weeks after baseline. RESULTS: The predictive variables for pain intensity were age (β=0.01 points, 95% CI=0.00 to 0.03, p=0.03) and pain intensity at baseline (β=0.23 points, 95% CI=0.13 to 0.33, p=0.00), with an explained variability of 4.6%. Similarly, the predictive variables for disability after four weeks were age (β=0.03 points, 95% CI=0.00 to 0.06, p=0.01) and disability at baseline (β=0.71 points, 95% CI=0.65 to 0.78, p=0.00), with an explained variability of 42.1%. CONCLUSION: Only age, pain at baseline and disability at baseline influenced the pain intensity and disability after four weeks of treatment. The beta coefficient for age was statistically significant, but the magnitude of this association was very small and not clinically important.