Study design: Numerous authors have attempted to sub-classify low back pain in order that valid homogenous subsets of low back pain presentations might be recognised. This review systematically appraises these papers. Methods: Medline, Embase, Cinahl, AMED and PEDro electronic databases were searched with subsequent hand searching of bibliographies. Papers were included between June 1983 and June 2003. Two reviewers independently reviewed 32 papers using a standard scoring criteria for assessment. A third reviewer mediated disagreements. Results: Thirty-two papers were reviewed, with classification systems being grouped by method of classification. Classification has been attempted by implication of patho-anatomical source, by clinical features, by psychological features, by health and work status and in one case by a biopsychosocial weighting system. Scores were generally higher for systems using a statistical cluster analysis approach to classification than a judgemental approach. Both approaches have specific advantages and disadvantages with a synthesis of both methodologies being most likely to generate an optimal classification system. Conclusions: The classification of NSLBP has traditionally involved the use of one paradigm. In the present era of biopsychosocial management of NSLBP, there is a need for an integrated classification system that will allow rational assessment of NSLBP from biomedical, psychological and social constructs.

No abstract

No abstract