Review: non-steroidal anti-inflammatory drugs and muscle relaxants are moderately effective for low back pain
Dr R Chou, Oregon Evidence-based Practice Center, Portland, OR, USA;
Medline (to November 2006), Cochrane Library (Issue 4, 2006), reference lists, and experts.
Study selection and assessment:
English-language systematic reviews of randomised controlled trials (RCTs) that were published since 2000 and evaluated medications for acute or chronic low back pain, with or without sciatica, in non-pregnant adults. If no high-quality systematic review was found for a given drug category, primary RCTs were sought. 7 systematic reviews and 24 additional RCTs met the selection criteria.
pain, back-specific function, and work disability.
The table shows a summary of the findings.
Good evidence exists that nonsteroidal anti-inflammatory drugs (NSAIDs) and skeletal muscle relaxants are moderately effective for acute low back pain. NSAIDs have moderate benefit, and tricyclic antidepressants have small-to-moderate benefit for chronic low back pain. Systemic corticosteroids are not effective for low back pain.
Chou R, Huffman LH. Medications for acute and chronic low back pain: a review of the evidence for an American Pain Society/American College of Physicians clinical practice guideline. Ann Intern Med 2007;147:505–14.
Clinical impact ratings: Emergency 6/7; Family/General practice 6/7; General/Internal medicine 6/7; Pain management 6/7
The review by Chou and Huffman summarised current evidence about the effectiveness of drug treatments for acute and chronic back pain, based on a clear set of principles for assessing the quality of systematic reviews and RCTs. It showed that the quality of much of the available evidence is not good. The paucity of studies of opioid analgesia, which is widely used in practice, is particularly surprising.
While some moderate reductions in pain (10-point to 20-point improvements on a 100-point visual analogue scale) were confirmed, these effects may not be clinically important, and there is a lack of evidence as to whether most of the drugs studied lead to valuable or sustained functional improvements. For example, return to work or usual activity is generally considered an important indicator for success of pain treatment, particularly for acute pain, yet this outcome is notable for its absence in the studies reviewed.
Most of the trials of chronic pain were of short duration and provided virtually no information on long-term outcomes. Data on activity levels, depression, and quality of life were sparse or difficult to interpret. The review highlighted concerns that long-term adverse effects and drug tolerance have not been sufficiently scrutinised. It is astonishing that benzodiazepines are considered as potential drug treatment for chronic pain in view of the known high risk of addiction and the recommendation that these drugs should not be used for >2–4 weeks.1
The management of all back pain needs to be viewed from a biopsychosocial perspective that addresses patients’ concerns, beliefs, and activity patterns in the context of their daily lives. A drug-induced “moderate” reduction in pain may be useful if it facilitates an increase in self-management and improves function and quality of life. The review by Chou et al highlights the limitations of drug trials that fail to account for a broad range of important outcomes assessed over time. In the absence of such patient-centred outcomes, the term “effective” is potentially misleading.