|Home | About | Journals | Submit | Contact Us | Français|
Although Ernst and Canter's review (April 2006 JRSM1) attempts to reduce the confusion over manipulation it only adds to it. First, no justification is offered for aggregating a heterogeneous range of health complaints. Secondly, there is no evidence of systematic quality appraisal of these disparate data, except by comments on professional backgrounds or by self-assessing their own work as `rigorous and systematic'. Thirdly, the exclusion criteria neglect the very studies which test the effectiveness of manipulation as used in practice, i.e. as part of a package of care.
Manipulation is a biomedical intervention, used mainly for common musculoskeletal disorders by a wide range of healthcare practitioners, both within the banner of conventional medicine and outside it. Taken alone, it is like any other intervention for these conditions; it will work in some cases but not others. It is increasingly clear that no biomedical approach in isolation is adequate for common musculoskeletal conditions. The usefulness of manipulation is that it can be used within a package of care that provides advice about re-activation, reassurance about resuming activity, pain control, and the recognition and minimization of psychosocial risks for chronicity. The trials excluded by Ernst and Canters review (e.g., the UK BEAM trial)2,3 show that manipulation is effective and cost-effective within such a package of care. Current guidelines also recognize this.
Ernst has a record of publications that take a different approach4; and there is enough evidence about manipulation in the back pain area that further explanatory trials are probably no longer needed. Rather, as with many interventions including exercise, further research is needed to help clarify where it is best used in a package of care and for which patient subgroups; so that practitioners who have the training to use it can do so more selectively within a holistic approach.