Five studies were included, with a total of 307 patients (range 35 to 100). Four studies were retrospective; one was prospective and non-randomised. None of the studies met the criteria for high quality. The prospective study did not adequately describe its study design, statistical analysis, and group baseline data. Validated outcome assessment scales were used in only one study. Follow-up lasted from two to six years.
Compared with posterolateral fusion, statistically significant, more favourable results were found with posterior lumbar interbody fusion, for non-union rates (RR 0.21, 95% CI 0.08 to 0.56; five studies, Ι²=0), disc height (WMD 3.2mm, 95% CI 1.9 to 4.4; four studies), and residual percentage slippage (WMD 6.3%, 95% CI 3.9 to 8.7; four studies). No significant differences were found for complications, functional outcomes, and segmental or total lumbar lordosis.