|A systematic review on the effectiveness of pharmacological interventions for chronic non-specific low-back pain
|Kuijpers T, van Middelkoop M, Rubinstein SM, Ostelo R, Verhagen A, Koes BW, van Tulder MW
This generally well-conducted review concluded that there was low quality evidence that opioids and non-steroidal anti-inflammatory drugs reduced pain compared with placebo in patients with non-specific chronic low-back pain; opioids also improved function in the short-term (up to three months). The authors’ conclusions are suitably cautious and appropriately acknowledge the limitations in the evidence base.
To assess the effectiveness of pharmacological agents for non-specific chronic low-back pain.
MEDLINE, EMBASE, CINAHL, Cochrane Central Register of Controlled Trials (CENTRAL) and PEDro were searched from the latest data of relevant Cochrane reviews up to December 2008 for articles published in English, Dutch or German. Search terms were not reported, but were available on request. Reference lists of relevant articles were scanned. Experts were contacted. Existing Cochrane reviews were searched for relevant studies.
Randomised controlled trials (RCTs) that compared pharmacological agents with control in adults (18 years or older) with chronic (longer than 12 weeks) non-specific low-back pain were eligible for inclusion. Trials had to have at least one day follow-up. Eligible pharmacological agents were non-steroidal inflammatory agents (NSAIDs), opioids, muscle relaxants, and antidepressants. Co-interventions were allowed. Eligible trials had to report at least one outcome measure of pain intensity, back-specific functional status, perceived recovery, and/or return to work. Abstracts and unpublished studies were excluded. Trials of specific low-back pain, post-partum, prevention studies, or post-operative studies were excluded.
The primary outcomes were: pain measured by the visual analogue scale (VAS), McGill pain questionnaire, or numerical rating scales; and back specific functional status measured by Roland-Morris Disability Questionnaire, or the Oswestry Disability Index).
The included trials studied various antidepressants, opioids, and NSAIDs, generally in comparison with placebo in patients with non-specific low-back pain. In several of the trials, a flare design was used by which patients who were already responding to the intervention were only included when they showed a worsening in low-back pain complaints during washout. Some patients had received prior back surgery. Adverse events were also reported.
Three reviewers (working in pairs) independently performed study selection; disagreements were resolved by discussion and consensus.
Assessment of study quality
Trial quality was assessed using the Cochrane Back Review Group (CBRG) criteria, which assessed 11 items including randomisation, blinding, allocation concealment, and withdrawals and drop-outs, to give a maximum score out of 11 points. Trials scoring less than 6 were deemed poor quality. The strength of evidence was assessed using the GRADE (Grades of Recommendation Assessment, Development and Evaluation) criteria, which graded evidence for each outcome as very low, low, moderate, and high.
Two reviewers independently performed quality assessment; disagreements were resolved by consensus.
Data were extracted on pain, functional status, recovery, return to work and adverse events. This was used to calculate relative risks (RRs) and mean differences (MDs), together with 95% confidence intervals (CIs).
The authors did not state how many reviewers performed the data extraction.
Methods of synthesis
Meta-analysis was used to pool relative risks, standardised mean differences (SMDs) and weighted mean differences (WMDs), together with 95% confidence intervals. A fixed-effect meta-analysis was used if the trials were deemed clinically and statistically homogeneous. Statistical heterogeneity was assessed using I2 and X2. Random-effects meta-analysis was used if heterogeneity was detected (I2 value over 20% was used as an indicator of heterogeneity). Sensitivity analyses were conducted for prior surgery, influence of tramadol, and type of drug.
Results of the review
Seventeen RCTs were included in the review. The quality of the included trials was mixed: 14 trials were deemed high quality (scores ranged from 6 to 10 out of 11) and three trials were deemed low quality (all scored 5). Despite this, many trials had methodological flaws in allocation concealment, compliance, and drop-out rates.
Non-steroidal anti-inflammatory drugs (NSAIDs) (four trials; low quality evidence): Compared with placebo, NSAIDs had a statistically significantly greater reduction in 100mm VAS pain scores for low-back pain (WMD -12.40 mm, 95% CI -15.53 to -9.26; four trials; n=1,020 participants), but statistically significantly more adverse events (RR 1.24, 95% CI 1.07 to 1.43; four trials; n=1,034 participants).
Antidepressants (four trials; moderate quality evidence): Compared with placebo, there was no statistically significant difference in pain scores (four trials) or adverse events (two trials). When selective serotonin reuptake inhibitors (three trials) and tricyclic antidepressants (two trials) were analysed separately, there was no statistically significant difference in pain intensity compared with placebo.
Opioids (seven trials; low quality evidence): Compared with placebo, opioids had a statistically significantly greater reduction in pain scores (SMD -0.54, 95% CI -0.72 to -0.36; seven trials; n=2,350 participants), and disability (SMD -0.19, 95% CI -0.31 to -0.08; four trials; n=1,258 participants), but there were statistically significantly more adverse events (RR 1.28, 95% CI 1.14 to 1.44; four trials; n=1,176 participants).
There was no evidence of statistical heterogeneity (where reported). Sensitivity analyses did not significantly alter the results.
Results for the narrative synthesis were also presented.
There was low quality evidence that, compared with placebo, opioids and NSAIDs reduced pain intensity, and that opioids also improved function in the short-term (up to three months) in patients with non-specific chronic low-back pain.
Inclusion criteria for the review were broadly defined. Several relevant data sources were searched. There may have been the potential for both language and publication biases, which the authors acknowledged. Attempts were made to reduce reviewer error and bias during study selection and quality assessment, but it was unclear if the same methods were used for data extraction.
Quality assessment was undertaken using a standard checklist, which indicated the variable quality of the included trials, which the authors acknowledged. The authors also noted that the flare design used in some trials may have biased results in favour of the pharmacological intervention. Trials were combined using fixed-effect or random-effects meta-analysis. Both clinical and statistical heterogeneity were considered, which was appropriate. However, the variable quality of included trials and potential biases in the trials limited the reliability of the evidence base, which the authors acknowledged. The review was generally well conducted.
The authors’ conclusions are suitably cautious and appropriately acknowledge the limitations in the evidence base.
Implications of the review for practice and research
Practice: The authors stated that NSAIDs and opioids may be useful for short-term pain relief in patients with non-specific chronic low-back pain who responded with an exacerbation of symptoms after stopping their medication. However, the benefits should be weighed against possible side-effects associated with specific medications.
Research: The authors stated that further high quality RCTs in non-specific chronic low-back pain were needed. Prospective studies with large sample sizes are needed to investigate the incidence of adverse events.
The Dutch Health Insurance Council.
Kuijpers T, van Middelkoop M, Rubinstein SM, Ostelo R, Verhagen A, Koes BW, van Tulder MW. A systematic review on the effectiveness of pharmacological interventions for chronic non-specific low-back pain. European Spine Journal 2011; 20(1): 40-50
Subject indexing assigned by NLM
Analgesia /methods; Analgesics, Opioid /therapeutic use; Anti-Inflammatory Agents, Non-Steroidal /therapeutic use; Antidepressive Agents /therapeutic use; Humans; Low Back Pain /drug therapy; Randomized Controlled Trials as Topic; Treatment Outcome
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.