Mean daily morphine equivalent dose in milligrams. Refer to eTable 5 in the Supplement for drug comparisons. Negative outcome values represent mean change from baseline. Egger P = .007 for short term and P = .42 for intermediate term.
A, Regression of log dose (morphine equivalents) on treatment effect for chronic low back pain; short-term pain relief; P = .046. B, Regression of study design on treatment effect; short-term pain relief; P = .69. Note circles represent one opioid analgesic vs placebo comparison. Negative values favor opioid analgesic. C, Effect size estimates from cutoff dose less than 100 mg morphine equivalents per day and greater than or equal to 100 mg morphine equivalents per day.
Loss to follow-up includes patients withdrawn from trial owing to noncompliance, patients electing to withdraw, patients without complete outcomes, and various other reasons.
eTable 1. Search Strategy ONE (OVID resources: MEDLINE, CENTRAL, PsycINFO, Cochrane database of Systematic Reviews).
eTable 2. Pedro ratings for eligible trials.
eTable 3. Data extraction Table.
eTable 4. Main effect sizes and grading of evidence.
eTable 5. Morphine equivalent dose conversion for opioid analgesic medicines evaluated in this review.
eTable 6. Proportion of participants who dropped out in run-in and trial phase.
eTable 7. Adverse event rates for opioid analgesic vs placebo trials.
eTable 8. Adverse event rates. Number (%) of participants with adverse events in placebo controlled studies of opioid analgesics.
eFigure 1. Funnel plot of standard error by treatment effect; chronic low back pain. Egger’s test (two-tailed p-value 0·007) for the primary time point and outcome: short term pain relief. Circles represent one opioid analgesic vs placebo comparison.
eFigure 2. Funnel plot of standard error by treatment effect; opioid analgesic studies; chronic low back pain: intermediate pain relief. Egger’s p=0·42.
eFigure 3. Intermediate and short term effects on pain from head-to-head opioid analgesic trials; chronic low back pain. Note each comparison is individually presented. BTDS = buprenorphine transdermal system; NTX = naltrexone; bd = twice daily dosing; qid = four times daily dosing; µg/h = microgram per hour.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Abdel Shaheed C, Maher CG, Williams KA, Day R, McLachlan AJ. Efficacy, Tolerability, and Dose-Dependent Effects of Opioid Analgesics for Low Back Pain: A Systematic Review and Meta-analysis. JAMA Intern Med. 2016;176(7):958–968. doi:10.1001/jamainternmed.2016.1251
What is the efficacy, tolerability, and dose-dependent effects of opioid analgesics for low back pain?
In this systematic review we found that recommended doses of opioid analgesics (range 40.0-240.0-mg morphine equivalents/day) did not provide clinically meaningful pain relief (> 20 points on a 0-100 point pain scale) in people with chronic low back pain.
For people with chronic low back pain who tolerate the medicine, opioid analgesics provide modest short-term pain relief, but the effect is not likely to be clinically important within guideline-recommended doses.
Opioid analgesics are commonly used for low back pain, however, to our knowledge there has been no systematic evaluation of the effect of opioid dose and use of enrichment study design on estimates of treatment effect.
To evaluate efficacy and tolerability of opioids in the management of back pain; and investigate the effect of opioid dose and use of an enrichment study design on treatment effect.
Medline, EMBASE, CENTRAL, CINAHL, and PsycINFO (inception to September 2015) with citation tracking from eligible randomized clinical trials (RCTs).
Placebo-controlled RCTs in any language.
Data Extraction and Synthesis
Two authors independently extracted data and assessed risk of bias. Data were pooled using a random effects model with strength of evidence assessed using the grading of recommendations assessment, development, and evaluation (GRADE).
Main Outcomes and Measures
The primary outcome measure was pain. Pain and disability outcomes were converted to a common 0 to 100 scale, with effects greater than 20 points considered clinically important.
Of 20 included RCTs of opioid analgesics (with a total of 7925 participants), 13 trials (3419 participants) evaluated short-term effects on chronic low back pain, and no placebo-controlled trials enrolled patients with acute low back pain. In half of these 13 trials, at least 50% of participants withdrew owing to adverse events or lack of efficacy. There was moderate-quality evidence that opioid analgesics reduce pain in the short term; mean difference (MD), −10.1 (95% CI, −12.8 to −7.4). Meta-regression revealed a 12.0 point greater pain relief for every 1 log unit increase in morphine equivalent dose (P = .046). Clinically important pain relief was not observed within the dose range evaluated (40.0-240.0-mg morphine equivalents per day). There was no significant effect of enrichment study design.
Conclusions and Relevance
For people with chronic low back pain who tolerate the medicine, opioid analgesics provide modest short-term pain relief but the effect is not likely to be clinically important within guideline recommended doses. Evidence on long-term efficacy is lacking. The efficacy of opioid analgesics in acute low back pain is unknown.
Low back pain is a common health problem and the leading cause of disability worldwide.1,2 While management guidelines encourage the prescription of simple analgesics for low back pain (eg, paracetamol or nonsteroidal antiinflammatory drugs [NSAIDs]), many people with low back pain are prescribed opioid analgesics.3,4Quiz Ref ID Findings from the United States show that more than half of the people regularly treated with prescription opioid analgesics have chronic low back pain.4 Similarly in Australia the 3 most commonly prescribed drugs for back pain are opioid analgesics or opioid analgesic combinations: oxycodone (11.7%), tramadol (8.2%), and paracetamol and codeine combination (12.1%) (percent of all prescribed medicines for back pain).5,6
Despite the widespread use of opioid analgesics for people with low back pain, there remains uncertainty about 3 central issues that would guide rational prescribing of these medicines for the treatment of people with low back pain. Two recent high-quality systematic reviews have shown that, because of a lack of trials, there is uncertainty regarding the efficacy of opioid analgesics for people with acute low back pain7 and also when these medicines are used long term for chronic low back pain.8 There is also uncertainty regarding the optimal dosing of opioid analgesics because, to our knowledge, there has been no systematic evaluation of how the dose of these medicines may influence the size of any treatment effect for people with low back pain.9-14 Finally, there has been little consideration of the extent to which an unselected group of opioid-naïve people with low back pain would tolerate or respond to an opioid medicine. For example, many opioid trials use an enrichment study design and exclude participants who do not tolerate or adequately respond to the opioid analgesic in the run-in phase. Some trials also exclude participants who do not respond to or tolerate the opioid analgesic in the randomized phase of a trial, and therefore the estimate of treatment efficacy is derived from only a proportion of participants who were enrolled in the study to receive opioid analgesics.
The aims of this systematic review were to (1) evaluate the efficacy of opioid analgesics in the management of low back pain, (2) investigate the effect of opioid dose and enrichment study design on treatment effect, and (3) quantify treatment discontinuation (owing to adverse events and lack of efficacy) and loss to follow-up in the run-in and randomized phases of trials.
MEDLINE, EMBASE, Cochrane Database of Systematic Reviews, CENTRAL, CINAHL, and PsycINFO (inception to end September 2015) were searched for randomized clinical trials (RCTs) evaluating opioid analgesic medicines for nonspecific low back pain (eTable 1 in the Supplement). In addition we screened reference lists of included RCTs and relevant systematic reviews to identify additional RCTs (Figure 1). We included only RCTs enrolling patients with nonspecific low back pain15 (ie, low back pain where a cause [eg, fracture, infection] had not been identified).
After screening titles and abstracts of retrieved studies, 2 reviewers drawn from a pool of 3 reviewers (C.A.S., A.J.M., and C.G.M.) independently inspected the full articles of potentially eligible RCTs to determine eligibility, with disagreements resolved by consensus.
We included studies in any language evaluating single-ingredient or combination medicines containing an opioid analgesic for nonspecific acute or chronic low back pain. Study selection was not restricted by pain duration, comorbid condition(s), or concurrent nonopioid or nonanalgesic medication use (eg, to treat hypertension) provided that participants were stabilized on these medications and the pattern of use was unchanged throughout the study. We included the anatomical therapeutic chemical (ATC)16 codes for drug classes relevant to this review in the search.
Placebo-controlled RCTs and RCTs comparing 2 drugs from the same class or different doses of the same drug were eligible for inclusion. Trials were included if they reported pain, disability, or adverse events outcomes. Trials involving patients with a range of pain conditions were included if results for participants with low back pain could be separately extracted.
Two reviewers (C.A.S., C.G.M.) independently extracted outcomes data from published studies. Missing data were obtained by contacting authors or estimated using the methods described in the Cochrane Handbook.17 Analysis of data from a crossover trial was performed according to recommendations in the Cochrane Handbook.17
Risk of bias was assessed using the 11-item PEDro scale (eTable 2 in the Supplement),18-20 a valid and reliable method of rating methodological quality of individual RCTs.18-20 Each item (excluding the item for external validity) was scored as either present (1) or absent (0) to give a total score out of 10. Rating of trials was carried out by 2 independent raters (C.A.S. plus A.J.M. or C.G.M.) with disagreements resolved by an independent third rater (eTable 3 in the Supplement). Trials scoring less than 7 out of 10 on the PEDro scale were considered to be at high risk of bias; those scoring 7 or more were considered to be at low risk of bias.18
Pain and disability outcomes were converted to a common 0 to 100 scale (0, no pain or disability, to 100, worst possible pain or disability). The pain intensity measures used in the retrieved trials were visual analog scale (VAS) scores (scale range, 0-100) and numerical rating scale (NRS) scores (range, 0-10). The NRS was converted to the same 0 to100 scale as in the VAS because these 2 pain measures have been shown to be highly correlated and, when transformed, can be used interchangeably.21 The disability measures used to calculate pooled effects were Oswestry Disability Index scores (range, 0-100) and Roland–Morris Disability Questionnaire (RMDQ) scores (range, 0-24). The RMDQ scores were converted to the same 0 to 100 scale as in the Oswestry Disability Index because these 2 questionnaires are highly correlated and share similar psychometric properties.22
We present results as mean differences (MDs) rather than standardized mean differences (SMDs) because the benchmarks for clinically important difference in pain and disability are expressed in points on a 0 to 100 scale, not proportions of a standard deviation.12,13 We considered a 10-point difference on this 0 to 100 point scale as a ”minimal” difference and a 20-point difference as “clinically important” difference, consistent with the proposed thresholds for clinically important changes in the chronic pain12 and low back pain literature.13 To reflect the way that patients typically rate their pain, we considered differences of less than 10 points (out of 100) on the pain or disability scale as unlikely to be noticed by most patients; some patients would notice a difference in the range of 10 to 20, points and most people are likely to notice a difference greater than 20 points on the pain or disability scale.
We considered short-term pain relief (follow-up < 3 months) as the primary outcome. Outcomes were grouped into 3 time categories (with respect to follow-up): short-term (<3 months) intermediate-term (≥3 months, <12 months), and long-term (≥12 months). Where multiple time points were available for a single category, the time closest to 6 weeks was chosen for short-term follow-up, 6 months for the intermediate-term, and 12 months for the long-term. We conducted a separate analysis to calculate treatment effect sizes for placebo-controlled trials evaluating combination opioid analgesics containing a simple analgesic (eg, paracetamol).
Where there were multiple comparisons from a single study, we divided the number of participants in the common arm by the number of comparisons, according to recommendations in the Cochrane Handbook.17 Meta-analysis was carried out using RevMan review management software (version 5.1) and Comprehensive Meta-Analysis software (version 3, Biostat).23 Pooled effects were calculated using a random effects model. Where possible, we explored possible causes of heterogeneity where I2 values were greater than 40%.
We used the grading of recommendations assessment, development, and evaluation (GRADE) criteria24 to evaluate the overall quality of the evidence for an intervention. This method is described elsewhere,25,26 but briefly the quality of evidence was downgraded a level for each of 4 factors: poor study design (≥25% of trials, weighted by sample size, had a low PEDro score [<7 of 10]), inconsistency of results (≥25% of the trials, weighted by sample size, had results which were not in the same direction), imprecision (sample size <300), and publication bias (assessed using funnel plot analysis/ Egger regression test). Where Egger regression 2-tailed P value was less than .10, the overall quality of evidence was downgraded by 1 level.27 We did not assess for indirectness (when the trial context is not the same as the review question) because this review encompassed a specific population. The quality of evidence was defined as “high quality,” “moderate quality,” “low quality,” and “very low quality.”24
As our primary analysis, the pooled effect of single-ingredient opioid analgesics was evaluated at short-term follow-up. To evaluate the effect of opioid dose we converted all doses of opioid analgesics to morphine equivalent doses28 and conducted metaregression to determine the effects of the log-transformed morphine equivalent dose on treatment effect size (for single ingredient opioid analgesic trials). We also conducted a simple stratified analysis exploring pooled effects at 100 mg or more and less than 100 mg morphine equivalents per day.
For each trial we collated the number of participants receiving an opioid analgesic who were withdrawn from the trial because of adverse events, lack of efficacy, or were lost to follow-up, in both the run-in and trial phase of a study. The proportion remaining were those who contributed data to the trial’s estimate of treatment efficacy. In addition, we explored the effects of study design (enrichment design [in which patients who tolerate and respond to the trial medicine in the run-in phase are allowed to continue to the randomization phase of the study] vs nonenrichment) on treatment effect using metaregression. At the same time we investigated the effect of opioid dose on treatment effect. We entered both items in the 1 model and report results from this multivariate model.
A total of 20 trials of opioid analgesics (a total of 7295 participants) were included in this review (see Table). Nineteen opioid analgesic trials evaluated participants with chronic low back pain, and 1 head-to-head trial evaluated participants with subacute low back pain. Seventeen RCTs compared an opioid analgesic with placebo and 3 trials compared 2 opioid analgesics.29-31Quiz Ref ID None of the trials evaluated long-term use or outcomes, the maximum treatment period was 12 weeks, and the maximum follow-up was also 12 weeks. Seventeen of the 20 trials reported industry funding. The medicines used in these trials were oral hydromorphone,32 oxymorphone,29,33,34 morphine,30,35 tramadol monotherapy31,36-38 or in combination with paracetamol,31,39-41 tapentadol,42 oxycodone monotherapy,42-44 oxycodone in combination with naloxone,45 or naltrexone,44 transdermal buprenorphine,43,46,47 transdermal fentanyl,30 and hydrocodone.48 The trials were typically of high quality with a mean (SD) PEDro score of 7.8 (1.5). The PEDro ratings are summarized in eTable 2 in the Supplement.
There is moderate-quality evidence from 13 studies of chronic low back pain (3419 participants) of an effect of single-ingredient opioid analgesics on pain in the short term; mean difference (MD), −10.1 (95% CI, −12.8 to −7.4) (Figure 2). There is high-quality evidence from 6 studies (2605 participants) that single-ingredient opioid analgesics relieve pain in the intermediate term; MD, −8.1 (95% CI, −10.2 to −6.0). Combination opioid analgesics containing a simple analgesic showed moderate evidence of pain relief in the intermediate term; MD, −11.9 (95% CI, −19.3 to −4.4). Results from a single trial of the combination of tramadol/paracetamol provided very low quality evidence of pain relief for the short term; MD, −8 (95% CI, −16.2 to 0.2).41 The effects of single-ingredient and combination opioids were minimal and approximately half the 20-point threshold for clinical importance. A funnel plot of standard error by treatment effect for the short and intermediate term is shown in eFigures 1 and 2 in the Supplement, respectively. The evidence for short-term efficacy was downgraded 1 level for publication bias (Egger P = .007). There were no long-term outcomes data.
There were limited data on disability outcomes. Results from a single trial showed no significant effect of the combination of tramadol-paracetamol41 on disability for the short term; MD, −6.2 (95% CI, −8.5 to 20.9), and results from another single trial40 of the tramadol-paracetamol combination showed no significant effect for the intermediate term; MD, −3.7 (−11.8 to 4.4). The evidence from these trials is of very low quality. A single study of morphine35 showed no clinically significant reduction in disability for the short term; MD, −6.3 (95% CI, 0.5-12.1), with the evidence being of very low quality.
See eTable 4 in the Supplement for overall grading of evidence and eTable 5 in the Supplement for morphine equivalent conversions.
In the 13 RCTs evaluating short-term efficacy, the morphine-equivalent dose ranged from 40.0 mg to 242.7 mg per day. Seven of the 13 RCTs used an enrichment study design whereby only the participants who responded favorably to the study medication, and tolerated the medicine in the trial run-in phase (prerandomization) were eligible to continue in the trial proper and be randomized to the study treatment. Trial results grouped by log opioid dose and enrichment study design are shown in Figure 3A and B, respectively. Results from the stratified analysis are shown in Figure 3C. The meta-regression model, including log opioid dose and enrichment study design, showed there was a significant effect of opioid dose on treatment effect, with a 12.0 point greater pain relief for every 1 log unit increase in dose (P = .046) (ie, 10 mg of morphine equivalents). The model predicts that clinically important effects are not seen within the dose range evaluated (≤240-mg morphine equivalents per day). The effect of enrichment study design was not statistically significant (P = .69). Together, log morphine-equivalent dose and enrichment study design accounted for 11% of variance in treatment effect.
Results from the multivariate analysis exploring dichotomous dose and enrichment study design showed no statistically significant effect of dose (P = .52) or enrichment study design (P = .40).
The proportion of participants given an opioid analgesic who were withdrawn from a trial owing to adverse events or lack of efficacy and the proportion lost to follow-up are shown in Figure 4 with more detailed information in eTable 6 in the Supplement. In the 8 trials (10 treatment contrasts) using an enrichment study design, only 20.3% to 48.4% remained in the trial and contributed data to the efficacy estimate, and in the nonenrichment study design this was 39.8% to 75.0%. Irrespective of study design, the predominant causes for dropout were adverse events or lack of efficacy, with half of trials having 50% of participants drop out owing to these 2 reasons. Even in the enrichment trials, where participants entered the trial only if they tolerated and responded to the medicine in the run-in phase, from 31.4% to 61.9% withdrew owing to adverse events and 3.3% to 29.6% withdrew owing to lack of efficacy in the randomized phase of the trial.
Eight trials32-34,37,38,40,42,47 provided details on the proportion of participants experiencing at least 1 adverse event in the run-in and/or RCT phase. Overall, the median rates (interquartile range [IQR]) of adverse events in the RCT phase were 49.1% (44.0%-55.0%) for placebo and 68.9.% (55.0%-85.0%) for treatment groups (risk ratio [RR], 1.3; P < .01). Common adverse events reported by participants in opioid analgesic trials included central nervous system adverse events (headache, somnolence, dizziness), gastrointestinal tract adverse events (constipation, nausea, vomiting), and autonomic adverse events, such as dry mouth. In some studies,33,34,38 over half of participants who experienced an adverse event completed the study. Studies rarely reported the severity or duration of adverse events, therefore it was not possible to categorize adverse events based on severity (see eTables 7 and 8 in the Supplement).
Treatment effects for head-to-head opioid analgesic trials are shown in eFigure 3 in the Supplement. There was a statistically significant difference in treatment outcome between different strengths of transdermal buprenorphine for both the short and intermediate term, with the 20-μg/h patch providing greater pain relief than the 5-μg/h patch (MD, 4.8 [95% CI, 0.6-9.0] and MD, 6.7 [95% CI, 2.3-11.2], respectively). Oxycodone, 40 mg per day, provided greater pain relief than transdermal buprenorphine, 5-μg/h, for both the short and intermediate term (MD, 5.9 [95% CI, 1.6-10.2] and MD, 7.6 [95% CI, 3.0-12.2], respectively). An evaluation of morphine equivalent conversions (eTable 5 in the Supplement) showed similar mean daily morphine equivalents for these trials, providing a plausible explanation for the small difference.
Quiz Ref IDThis review has found that there is evidence that opioid analgesics relieve pain in the short and intermediate term for people with chronic (but not acute) low back pain, but it is uncertain if they improve disability.Quiz Ref ID Treatment effects are small, being half the threshold for clinical importance. The medicines are also commonly associated with adverse events. We found some evidence of a greater effect of opioid analgesics with larger doses; however, the effects are not likely to be clinically important even at high doses. A detailed analysis of drop-outs from trials revealed that under half of participants entering these trials contributed to treatment effect size estimates. There is no evidence on long-term use and limited evidence for acute low back pain.
The strengths of this review include a consideration of opioid dose and study design as well as a comprehensive search strategy covering single-ingredient and combination opioid analgesics used to treat low back pain. The PEDro scale was used to assess risk of bias because it has acceptably high clinometric properties,18-20 whereas limitations have been reported for the Cochrane risk of bias scale.49,50 Limitations of this review include possible publication bias, as only studies published in peer-reviewed journals were included. A limitation of the meta-regression is that it does not account for variability in dose response as a result of duration of treatment or intrinsic factors (eg, genetic variability).
Our review challenges the prevailing view that opioid medicines are powerful analgesics for low back pain. Opioid analgesics had minimal effects on pain, and even at high doses the magnitude of the effect is less than the accepted thresholds for a clinically important treatment effect on pain. Importantly, the magnitude of the treatment effect we observed is similar to that reported for NSAIDs vs placebo in the Cochrane review of NSAIDs.51 This review focuses on the magnitude and clinical significance of treatment effects12,13 and presents a more detailed evaluation of run-in failure, treatment discontinuation, and loss to follow-up compared with existing reviews.7,9,10 Typically, the trials’ estimate of treatment effect was derived from approximately 50% of participants who initially entered the trial, with participants withdrawn mainly because they did not tolerate or respond to the medicine. Taken together, these issues suggest there would be much lower treatment effects among unselected groups of people new to these medicines.52
An unexpected finding of our study was that the use of an enrichment design was not associated with an exaggerated treatment effect. This result is in alignment with that of a study comparing enriched and nonenriched trials of opioid analgesics for chronic noncancer pain, in which it was similarly noted that enriched study designs do not significantly influence the estimates of treatment efficacy.53
An important feature of this review is that we also conducted a metaregression to explore association between the opioid analgesic dose and treatment effect and found there is a 12.0-point increase in pain relief for every 1 log unit increase in morphine-equivalent dose. Adjusting for use of an enrichment design, the predicted treatment effect was a 6.7-point increase in pain relief at 40-mg morphine equivalent units and a 16.1-point increase in pain relief at 240-mg morphine equivalent units. The metaregression findings suggest that while increasing dose was associated with greater effect, even at the higher doses the effect of opioid was less than the 20-point criterion we adopted for a clinically important effect. However, it is important to note that these are group mean differences, and some individuals may experience meaningful pain relief from clinically used doses whereas others may get very minimal relief or no relief at all. Concerns over adverse events and misuse continue to rise in line with the increased use and misuse of these medicines.11,54 However, despite the widespread use of opioid analgesics there has been uncertainty about their clinical benefits for low back pain until now. This review dispels the common public misconception that opioid analgesics are powerful analgesics, showing this is not the case for chronic low back pain and that clinically significant pain relief is not likely to be achieved at higher doses of opioid analgesics, which can be associated with potentially harmful effects (240-mg morphine equivalents per day).55 The latest review of opioid prescribing guidelines55 cautions against exceeding 200 mg of morphine equivalents per day as a way of alleviating the risk of opioid-related complications, such as life-threatening respiratory depression. Higher doses of opioid analgesics have also been associated with misuse, physical dependence, hyperalgesia, and clinically significant hormonal changes.55-63 In 2010, there were 16 65164 opioid-related deaths reported in the United States. Of concern are the paradoxical effects of opioid analgesics including opioid-induced hyperalgesia,65 a vexing issue thought to propagate the cycle of misuse and dependency on these medications. There is currently no evidence to support the long-term use of opioid analgesics in low back pain at any dose. Our review highlights the need to revise existing guidelines to acknowledge that (1) the effect of opioid analgesics in chronic low back pain is modest and not likely to be clinically significant at recommended doses and (2) there is considerable uncertainty about long-term use of these medicines for back pain.
Previous reviews excluded studies that evaluated combination medicine products containing opioid analgesics and simple analgesics or opioid antagonists.9,10 The addition of simple analgesics to opioid analgesics may have synergistic effects and potentially reduce the need for higher doses of opioid analgesics.61-63,66 This review has found moderate-quality evidence that paracetamol combined with the weak opioid tramadol provides modest, but not clinically significant reduction in pain (but not disability). This result is similar to findings from single-ingredient opioid analgesics. Another trial of the paracetamol-codeine combination66 not included in the pooled analysis of this review showed that there were significantly more people with chronic low back pain who experienced more than 30% pain relief in the paracetamol/tramadol group than the placebo group; however, it was not possible to interpret the clinical significance of these effects.
Tramadol and tapentadol exhibit their analgesic effects by 2 complementary pharmacological actions: binding to μ-opioid receptors, and inhibition of reuptake of noradrenaline and serotonin.67,68 Based on these known pharmacological properties, tramadol and tapentadol are widely regarded as synthetic analgesic with opioid-like effects (World Health Organization Anatomical Therapeutic Chemical code N02AX02 and N02AX06, respectively, within the N02A class) and, as such, were included in the analysis as opioid analgesics.
Quiz Ref IDVery few opioid analgesic trials reported on global recovery outcomes, and the effects on disability, where reported in this review, were small. Even with short-term use, there is no substantive reduction in disability. It is important to consider long-term functional outcomes with these medicines because a reduction in pain alone may not be correlated with improved disability outcomes. It is possible that the combination with nonpharmacological treatment strategies, such as physical activity, may have a synergistic effect and future research should focus on evaluating the benefits of such an approach. While this review suggests that opioid analgesics provide pain relief which is comparable to the effects reported in previous reviews of NSAIDs, head-to-head comparisons would yield valuable insight into effective dose regimens and choice of medicine.
In people with chronic low back pain, opioid analgesics provide short and/or intermediate pain relief, though the effect is small and not clinically important even at higher doses. Many trial patients stopped taking the medicine because they did not tolerate or respond to the medicine. There is no evidence on opioid analgesics for acute low back pain or to guide prolonged use of these medicines in the treatment of people with chronic low back pain.
Corresponding Author: Andrew J. McLachlan, PhD, Pharmacy and Bank Building A15, Faculty of Pharmacy, University of Sydney, Sydney, NSW 2006, Australia (firstname.lastname@example.org).
Published Online: May 23, 2016. doi:10.1001/jamainternmed.2016.1251.
Author Contributions: All authors approved the final version of the article. Dr Abdel Shaheed had access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Abdel Shaheed, Maher, Williams, McLachlan.
Acquisition, analysis, or interpretation of data: AbdelShaheed, Day, McLachlan.
Drafting of the manuscript: Abdel Shaheed, Maher, McLachlan.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: AbdelShaheed, Maher.
Administrative, technical, or material support: Maher, McLachlan.
Study supervision: Maher, Williams, Day, McLachlan.
Other: Abdel Shaheed.
Conflict of Interest Disclosures: Drs McLachlan, Maher, and Day are investigators on 3 trials evaluating medicines for low back pain.
Funding/Support: Dr Maher is funded by a research fellowship from the National Health and Medical Research Council of Australia. Dr McLachlan is the Program Director on the National Health and Medical Research Council of Australia Centre for Research Excellence on Medicines and Ageing.
Additional Contributions: We thank Nathaniel Katz, MD, for assisting with our data request. We also thank Patricia Parreira, BA, for providing translation. They were not compensated for their contributions.