Is the methodological quality of trials on spinal manipulative therapy for low-back pain improving?

Background: Spinal manipulative therapy (SMT) is an intervention practiced worldwide by a variety of professionals. Numerous randomized controlled trials (RCTs) have examined the effectiveness of SMT; however, interpretation of those treatment effects can be hampered by a high risk of bias and poor reporting as well as small sample sizes.Objective: To provide a descriptive overview of RCTs on SMT for low-back pain as well as an analysis of the trends with time in relation to risk of bias and sample size.Methods: Descriptive data on 61 RCTs of SMT for low-back pain were extracted. RCTs published prior to March 31, 2011, which fulfilled the inclusion criteria, were included. Publication date of the individual RCTs was used in the analyses. Linear regression was conducted to test trends in sample size; however, trends in risk of bias was not formally tested due to insufficient data.Results: Of the included RCTs, SMT was delivered by either a chiropractor or manual therapist in most (68%), and approximately half (49%) examined a high-velocity thrust. Sufficient data are available for the outcomes, pain and functional status, but lacking for other outcomes. Overall, 28% of the RCTs met the criteria for a low risk of bias and the median sample size for SMT (interquartile range) was 60 (34, 90). There is a positive trend over time in studies with a low risk of bias, in addition to improvements in reporting of specific items related to selection bias and selective reporting bias as well as intention to-treat. Despite this trend, many items were fulfilled by less than half of the studies published in the last decade. In addition, there is a trend towards larger studies for SMT as the intervention, although this also demonstrates variation with time.Conclusions: The continuing uncertainty regarding the effect of SMT for low-back pain is hampered by too many studies with a high risk of bias, which in some cases, are too small to detect clinically-relevant differences. It is our wish that the lessons learned from this analysis be applied in the design of future trials of SMT as well as other non-pharmacological therapies. (c) 2012 Elsevier Ltd. All rights reserved.