Four RCTs (n=707) were included in the review.
Although validity was not formally assessed, aspects of methodological quality were discussed for each study. All studies demonstrated at least some of the following weaknesses: lack of blinding; inadequate description of the randomisation process; insufficient information on group characteristics; insufficient information to establish whether control and intervention groups were comparable at baseline; failure to control for possible confounding factors; loss to follow-up; failure to use intention-to-treat analysis; and under-powered studies.
The highest quality trial (n=40) reported a significant benefit of spinal manipulation compared with chemonucleolysis on pain at 2 and 6 weeks and on disability at 2 weeks. There was no difference between the groups at 12 months.
One study (n=233) reported that 80% of patients treated with spinal manipulation had recovered at 2 weeks compared with 67% of control patients treated with heat. One study (n=112) reported a significant benefit of manipulation compared with traction in an in-patient population at 5 weeks' follow-up. The final study (n=322) found no significant difference between a group receiving spinal manipulation and patients undergoing exercise, traction or corset interventions. All of these studies were deemed to be of relatively poor quality.