Eleven RCTs (n=1,245) were included.
The quality scores ranged from 3 out of 10 to 8 out of 10. Eight studies were considered high quality.
McKenzie versus passive therapy.
The meta-analysis indicated a significant decrease in pain (WMD -4.16, 95% CI: -7.12, -1.20; based on 2 studies) and disability (WMD -5.22, 95% CI: -8.28, -2.16; based on 2 studies) favouring McKenzie at the 1-week follow-up, but there was no difference in disability at 4 weeks.
McKenzie versus advice to stay active.
There was no significant difference in pain at 12 weeks' follow-up, but disability scores favoured the advice group at 12 weeks (WMD 3.85, 95% CI: 0.30, 7.39; based on 2 studies).
McKenzie versus flexion exercises.
Individual studies reported that McKenzie was as effective as flexion for chronic pain at 2 weeks (1 study) and marginally better for acute pain at 8 weeks (1 study). Another study reported a large effect on acute disability (WMD -22, 95% CI: -26, -18) favouring McKenzie over flexion exercises at 5 days' follow-up.
Other comparisons.
Three trials comparing McKenzie with spinal manipulation therapy gave contradictory results. One study compared McKenzie with back school (lesson on back care and avoidance of inactivity) and found a higher return to work rate in the McKenzie group (RR 2.05, 95% CI: 1.43, 2.95). One trial compared McKenzie with strengthening exercises and found no significant differences in pain or disability at 8, 10 and 32 weeks.
Sensitivity analysis.
A pooled analysis could not be performed. Single studies indicated that classification-based McKenzie was as effective as an educational booklet, advice to stay active and strengthening exercises, and more effective than flexion exercises and spinal manipulative therapy.