Abstract
Objectives: To assess the evidence for a similar pattern of response to treatment among non-specific low back pain (NSLBP) patients in clinical trials.
Design: A systematic review of published trials on NSLBP and meta-analysis of within-group treatment effect calculated as the Standardised Mean Difference (SMD).
Data source: The Cochrane Register of Controlled Trials’ database (CENTRAL), April 2007.
Review methods: We included randomised controlled trials that investigated the effectiveness of primary care treatments in patients with NSLBP aged 18 years or over. We excluded trials conducted in patients with LBP of identifiable cause (e.g. disc herniation or arthritis), post-operative or post-traumatic back pain, or back pain during pregnancy or labour. We chose outcome measures commonly used in the majority of NSLBP trials, namely the Visual Analogue scale (VAS) for pain severity, Roland Morris Disability questionnaire (RMDQ) and Oswestry Disability questionnaire (ODQ) for physical functioning.
Results: 118 trials investigating a wide range of primary care treatment for NSLBP were included. In spite of heterogeneity, we found evidence for a similar pattern of symptom improvement represented by large SMDs at six weeks follow up ((0.86 for pain, 95% CI = 0.65,1.07, 0.97 for RMDQ, 95% CI = 0.66,1.28 and 0.98 for ODQ, 95% CI = 0.62,1.33) followed by much smaller further change at 13 week (pain 1.07 95% CI = 0.87,1.27, RMDQ 0.93 95% CI = 0.67,1.20, ODQ 0.92 95% CI = 0.70,1.14), 27 week (pain 1.03 95% CI = 0.82,1.25, RMDQ 0.91 95% CI = 0.59,1.24, ODQ 1.08 95% CI = 0.80,1.36 and 52 week (pain 0.88 95% CI = 0.60,1.1, RMDQ 1.01 95% CI = 0.68,1.34, ODQ 1.14 95% CI = 0.88,1.39). There was no statistically significant difference between responses in various trials arms (index treatment, active comparator treatment, placebo or sham treatment, usual care or waiting list controls). There was also no statistically significant difference between responses to pharmacological and non-pharmacological treatments.
Conclusions: NSLBP symptoms seem to improve very well and in a similar pattern in clinical trials following a wide variety of active as well as inactive treatments. It is important to explore factors other than the treatments themselves that might influence symptom improvement. Exploring possible sources of underlying heterogeneity in responses might lead to some of these factors.
Conflict of Interest: None
Source of Funding: This work is part of a PhD fellowship funded by the arc (Arthritis research campaign).
Correspondence should be addressed to: SBPR at the Royal College of Surgeons, 35–43 Lincoln’s Inn Fields, London WC2A 3PE, England.
