Thirty-four studies (n=1,137 reported in tables) were included: 4 randomised controlled trials (RCTs; n=178) and 30 retrospective studies.
Twenty-six studies (n=890) evaluated isolated posterior fusion procedures, 5 studies (n=77) evaluated isolated anterior lumbar interbody fusion procedures and 9 studies (n=170) evaluated combined anterior and posterior stabilisation procedures. Some studies evaluated more than one type of procedure.
Ten studies reported blinding of the assessment of radiological and functional outcomes. Nineteen studies reported follow-up of at least 24 months after surgery.
Fusion rates were 98.2% (167 out of 170) for combined procedures, 74% (57 out of 77) for anterior procedures and 83.3% (741 out of 890) for posterior approaches. The combined procedure was associated with a statistically significantly higher fusion rate than either the solely posterior or the anterior approach (p<0.0001 for both). Fusion rates were higher with the posterior approach than with the anterior approach, but the difference was not statistically significant (p=0.059).
Clinical success rates were 86.4% (108 out of 125) for combined procedures, 89.6% (60 out of 67) for anterior procedures and 74.8% (608 out of 814) for posterior approaches. The combined procedure was associated with a statistically significant higher clinical success rate than the posterior approach (p=0.0045) but not when compared with the anterior approach (p=0.65). Success rates were statistically significantly higher with the anterior approach than with the posterior approach (p=0.0047).
Influences on outcomes of posterior surgery.
Posterolateral fusion plus internal fixation was associated with statistically significant higher fusion rates (90.2% versus 77.4%, p<0.0001) and clinical success rates (84.9% versus 64.4%, p<0.0001) than posterior fusion without spinal fixation.
Smokers had statistically significant lower fusion rates (74.4% versus 91.4%, p=0.0023) and lower clinical success rates (59.6% versus 80.6%, p=0.0017) than nonsmokers.
Patients involved in compensation or litigation claims had similar fusion rates compared with others, but clinical success rates were significantly lower (35.6% versus 81.8%, p<0.0001).