Fifteen RCTs were included (sample sizes not reported). Studies were considered to be of moderate quality with PEDro scores that ranged from 5 to 10 out of 11 (mean 7.7). Flaws included lack of blinding, lack of reporting of intention-to-treat analysis and high drop-out rates and losses to follow-up. Follow-up was reported at 12 months in 12 studies and at two or more years in three studies.
Pain: Compared against control interventions, physiotherapy exercise programmes were associated with a significant reduction in pain scores at six months follow-up (SMD -0.57, 95% CI -0.75 to -0.39; six studies, 555 patients) and a non-significant reduction in pain scores at 12 months (SMD -0.25, 95% CI -0.44 to 0.06; 434 patients). There was no significant difference between intervention groups after 12-month follow-up (three studies).
Recurrence of low back pain: Compared against control interventions, physiotherapy exercise programmes were associated with a significant reduction in the proportion of patients with recurrence of low back pain at six months (RR 1.74, 95% CI 1.33 to 2.27, NNT=3; six studies, 296 patients). There was no significant difference between intervention groups at 12-month follow-up (seven studies, 472 patients).