Osteopathic intervention in chronic non-specific low back pain: a systematic review

Stratification of non-specific low back pain patients for therapeutic trials in general and illustrative suggestions for trials of manipulation in particular.

Brian Sweetman   (2013-05-08 17:31)  NHS

Orrock and Myers (1) have reviewed the present status of osteopathic intervention for so-called ��non-specific�� low back pain. They conclude that further clinical trials into this subject are required that have consistent and rigorous methods. However their introduction states that Non Specific Low Back Pain (CNSLBP) is a common, complex and disabling condition, but go on in the discussion to note that many studies had mixed back pain populations. There in lies one of the greatest problems of trial design. Should initial data collection allow for subgroup analysis? Which symptoms and signs are best for such stratification?
Subgrouping test suggestions for such secondary (or primary) analysis are summarized in a monograph (2) based on a classification analysis based on an earlier randomized controlled therapeutic trial (3). The reatments included Short Wave Diathermy, Traction, and Extension Exercises. A recent review of the original database (4) analyzed patient comments on previous treatments including hard bed, hot bath, SWD, exercises, traction, manipulation and injection. The results are given as amalgamated standardised deviates. ��Good�� scores are positive over 25 and include more than average better and/or not worse. ��Bad�� scores are negative worse than -25 and included more than average saying worse and/or not better. ��No effects�� score between +25 and -25.
Manipulation was best for thoracolumbar junction pain radiating into the low back (+52), then the facet joint syndrome (+27).
The patients suggested there was ��No effect�� with disc prolapse and those with switching side of lumbago between episodes.
Worst was for those with what is probably disc degeneration (-262) and then rotation back strain (-63). See the monograph (2) for the clinical tests that allow for these sub diagnoses.
These results suggest how much difference there is between subgroups. The no effect cases will dilute out overall evidence of effects, and good and bad results will cancel out if looking at all cases together.
This shows how important stratification could be if studies are to make any progress. But these results should under no circumstances be taken as meaningful of them selves, and there might be dangers for example in accepting that ��manipulation�� would not be harmful for cases of disc prolapse. Moreover there was patient selection bias as they were by definition needing more help despite previous treatments, they were giving subjective and retrospective opinions on outcome, and the treatment titles were very very broad.
Brian Sweetman FRCP, MD, PhD.
Swansea, UK.
Email: bjsweetman@hotmail.com
References
1. Orrock PJ, Myers SP. Osteopathic intervention in chronic non-specific low back pain: a systematic review. BMC Musculoskeletal Disorders 2013, 14:129. ISSN Publication date 9 April 2013.
2. Sweetman BJ. Low back pain, some real answers. 2005. tfm Publishing, Harley SY5 6LX; UK.
3. Sweetman BJ. Numerical classification of common low back pain. MD Thesis, London University, 1985.
4. Sweetman BJ, Sweetman SJ. Back pain database analysis. 2012; Unpublished. Details available on enquiry via email to bjsweetman@hotmail.com

Competing interests

None

top