Reviews1 September 2015

Epidural Corticosteroid Injections for Radiculopathy and Spinal Stenosis

FREE
A Systematic Review and Meta-analysis
    Author, Article, and Disclosure Information

    Abstract

    Background:

    Use of epidural corticosteroid injections is increasing.

    Purpose:

    To review evidence on the benefits and harms of epidural corticosteroid injections in adults with radicular low back pain or spinal stenosis of any duration.

    Data Sources:

    Ovid MEDLINE (through May 2015), Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, prior systematic reviews, and reference lists.

    Study Selection:

    Randomized trials of epidural corticosteroid injections versus placebo interventions, or that compared epidural injection techniques, corticosteroids, or doses.

    Data Extraction:

    Dual extraction and quality assessment of individual studies, which were used to determine the overall strength of evidence (SOE).

    Data Synthesis:

    30 placebo-controlled trials evaluated epidural corticosteroid injections for radiculopathy, and 8 trials were done for spinal stenosis. For radiculopathy, epidural corticosteroids were associated with greater immediate-term reduction in pain (weighted mean difference on a scale of 0 to 100, −7.55 [95% CI, −11.4 to −3.74]; SOE, moderate), function (standardized mean difference after exclusion of an outlier trial, −0.33 [CI, −0.56 to −0.09]; SOE, low), and short-term surgery risk (relative risk, 0.62 [CI, 0.41 to 0.92]; SOE, low). Effects were below predefined minimum clinically important difference thresholds, and there were no longer-term benefits. Limited evidence showed no clear effects of technical factors, patient characteristics, or comparator interventions on estimates. There were no clear effects of epidural corticosteroid injections for spinal stenosis (SOE, low to moderate). Serious harms were rare, but harms reporting was suboptimal (SOE, low).

    Limitations:

    The review was restricted to English-language studies. Some meta-analyses were based on small numbers of trials (particularly for spinal stenosis), and most trials had methodological shortcomings.

    Conclusion:

    Epidural corticosteroid injections for radiculopathy were associated with immediate reductions in pain and function. However, benefits were small and not sustained, and there was no effect on long-term surgery risk. Limited evidence suggested no effectiveness for spinal stenosis.

    Low back pain is one of the most frequently encountered conditions in clinical practice (1–5). Although most low back pain is nonradicular, symptomatic spinal stenosis or herniated disc each occur in about 3% to 4% of patients (6). Epidural corticosteroid injections are most commonly performed for radiculopathy due to a herniated disc, but may also be given for spinal stenosis. Despite conflicting conclusions from systematic reviews (7–13) and discordant clinical practice guidelines (14–17), use of epidural injections has increased (18, 19).

    Challenges in interpreting the evidence on epidural corticosteroid injections include variability in the methods used to select patients for inclusion, the injection techniques used, choice of comparators, and when and how outcomes are assessed (10, 20). The purpose of this systematic review is to synthesize the current evidence on the effects of epidural corticosteroid injections for radiculopathy and spinal stenosis.

    Methods

    Detailed methods and data for this review, including the analytic framework, key questions, search strategies, inclusion criteria, study data extraction, and quality ratings, are available in the full report (21). The full report also addresses other types of injections, nonradicular and postsurgical back pain, and effects of epidural injections versus active comparators. The protocol was developed by using a standardized process (22) with input from experts and the public, and was posted on the Agency for Healthcare Research and Quality (AHRQ) Web site on 29 May 2014 (23). This article focuses on the effectiveness and harms of epidural corticosteroid injections for radiculopathy or spinal stenosis, and whether effectiveness estimates vary according to technical factors, patient characteristics, or type of placebo comparator.

    We defined “placebo interventions” as epidural saline or local anesthetic injections without corticosteroid, a soft-tissue injection, or no injection, on the basis of the assumption that therapeutic effects in the epidural space are primarily related to the corticosteroid. Technical factors included the corticosteroid or local anesthetic used, medication doses, volume of injectate, number of levels injected, frequency and number of injections, use of imaging guidance, and route of administration. Patient characteristics included demographic (for example, age, sex, race) and clinical factors (for example, imaging findings, duration of symptoms, and presence of psychosocial factors or neurologic findings).

    Data Sources and Searches

    A research librarian searched MEDLINE, the Cochrane Central Register of Controlled Trials, and the Cochrane Database of Systematic Reviews from 2008 through May 2015. Studies published before 2008 were identified from prior reviews that we conducted (7, 24). We also reviewed reference lists and searched ClinicalTrials.gov.

    Study Selection

    Two investigators independently reviewed abstracts and full-text articles against prespecified eligibility criteria. We included randomized trials of adults undergoing epidural corticosteroid injections versus placebo interventions for radicular low back pain or spinal stenosis of any duration. We considered “sciatica” to be synonymous with radiculopathy. We included epidural injections performed via any approach, as well as transforaminal injections that did not necessarily enter the epidural space (“periradicular” injections). We also included studies that directly compared injection techniques, corticosteroids, and corticosteroid doses. Outcomes were pain, function, composite outcomes, subsequent surgery measured at least 5 days after the injection, and local and systemic harms. For harms, we also included large treatment series (sample size >1000 patients).

    We excluded studies of back pain due to fracture, high-impact trauma, cancer, or infection.

    Data Extraction and Quality Assessment

    One investigator extracted details about the study design, patient sample, setting, interventions, and results. Another investigator verified extractions for accuracy. Two investigators independently assessed risk of bias (“quality”) for each randomized trial as good, fair, or poor by using predefined criteria (25). Discrepancies were resolved through a consensus process.

    Data Synthesis and Analysis

    We conducted meta-analyses by using the Dersimonian–Laird random-effects method in Stata/IC 13.0 (StataCorp LP). Statistical heterogeneity was measured with the Cochran chi-square test and the I2 statistic (26). When statistical heterogeneity was present, we repeated meta-analysis by using the profile likelihood method (27). All analyses were stratified by the approach used (transforaminal, interlaminar, or caudal). Outcomes were analyzed as immediate (5 days to ≤2 weeks), short-term (2 weeks to ≤3 months), intermediate-term (3 months to <1 year), and long-term (>1 year), using the longest-duration data available within each category. For continuous outcomes, pain scores were converted to a scale of 0 to 100 and pooled as weighted mean differences (WMDs); function was pooled as standardized mean differences (SMDs) unless all trials in an analysis reported the same functional outcome. We used pain scores for leg pain when available, and overall or back pain when leg pain was not reported. The mean difference was calculated from the change from baseline to follow-up; sensitivity analysis based on adjusted estimates (for example, analysis of covariance) or differences in follow-up scores gave similar results and are not reported further. We imputed missing SDs by using the mean value from other studies in that analysis.

    For dichotomous outcomes, we pooled relative risks (RRs) for successful (as defined in the trials) pain, function, and composite outcomes and rates of subsequent surgery. To investigate whether certain placebo interventions might have therapeutic effects, we also performed separate pooled analyses on the placebo group response rates for continuous and dichotomous outcomes, stratified by the specific type of placebo comparator.

    We performed sensitivity analyses excluding poor-quality and outlier studies, and subgroup analyses and meta-regression on the corticosteroid, corticosteroid dose (in prednisolone equivalents), the local anesthetic, the comparator, injectate volume, symptom duration, use of imaging correlation, use of fluoroscopic guidance, number of injections, exclusion of patients with prior surgery, year of publication, and blinding methods. For analyses with at least 10 studies, we created funnel plots and performed the Egger test for small sample effects (28).

    We defined a minimum clinically important difference as an improvement in 15 points on a pain scale of 0 to 100, 10 points on the Oswestry Low Back Pain Disability Index (ODI), and 5 points on the Roland–Morris Disability Questionnaire (RDQ) (29).

    We assessed the overall strength of each body of evidence as high, moderate, low, or insufficient on the basis of aggregate study quality, precision, consistency, and directness (22).

    Role of the Funding Source

    The AHRQ funded the review at the request of the Centers for Medicare & Medicaid Services, who assisted in developing the scope of the review and key questions. Neither organization had a role in study selection, quality assessment, or synthesis. The investigators are solely responsible for the content.

    Results

    The literature search and selection is summarized in the Appendix Figure. Database searches resulted in 202 potentially relevant articles. After full-text dual review, 59 trials and 4 observational studies met inclusion criteria for the interventions and comparisons addressed in this article.

    Appendix Figure. Summary of evidence search and selection.

    * Cochrane Central Register of Controlled Trials and the Cochrane Database of Systematic Reviews.

    † Reference lists of relevant articles and systematic reviews, among other sources.

    ‡ The full report (21) also addresses other types of injections, nonradicular and postsurgical back pain, and effects of epidural injections versus active comparators.

    § Some studies are included for more than 1 question.

    Thirty trials (26 to 239 participants) compared epidural corticosteroid injections via various approaches with placebo interventions for radiculopathy (30–58), and 8 trials (29 to 386 participants) compared epidural corticosteroid injections with placebo interventions for spinal stenosis (Appendix Table 1) (37, 40, 59–64). Duration of follow-up ranged from 1 week to 3 years. The trials primarily evaluated patients with chronic symptoms.

    Appendix Table 1. Trials of Epidural Corticosteroid Injections for Radicular Pain and Spinal Stenosis

    Appendix Table 1.

    Appendix Table 1. Continued

    Appendix Table 1.

    Appendix Table 1. Continued

    Appendix Table 1.

    Four trials of epidural injections for radiculopathy (60 to 106 participants) (65–68) and 1 trial of spinal stenosis (70 participants) (69) evaluated effects of one corticosteroid versus another, and 6 trials (33 to 60 participants) evaluated corticosteroid dose effects (70–75). Eleven trials (30 to 239 participants) directly compared alternative epidural injection techniques (46, 76–85). Two trials compared effects of different patient evaluation and selection methods involving imaging (86, 87).

    Five trials were rated as good-quality (36, 43, 44, 59, 84), 40 trials as fair-quality, and 14 trials as poor-quality. Methodological shortcomings included failure to report adequate randomization or allocation concealment methods; inadequate blinding of outcome assessors, injectionists, or patients; high or unclearly reported attrition; and failure to specify primary outcomes.

    Effectiveness

    Radiculopathy

    Epidural corticosteroid injections were associated with greater immediate reduction in pain intensity compared with placebo interventions (6 trials; WMD on a scale of 0 to 100, −7.55 [95% CI, −11.4 to −3.74]; I2 = 30%; strength of evidence [SOE], moderate) (33, 38, 41, 44, 45, 57) (Figure 1 of the Supplement, but differences were smaller and not statistically significant at longer follow-up (SOE, low to moderate) (Appendix Table 2 and Figures 2 to 4 of the Supplement). For immediate functional improvement, effects favored epidural corticosteroids, but the difference was not statistically significant (4 trials; SMD, −0.75 [CI, −1.62 to 0.11]; I2 = 94%; SOE, low) (33, 44, 54, 57) (Figure 5 of the Supplement). Statistical heterogeneity was substantial owing to an outlier trial (54) that reported a much stronger effect than the other trials (SMD, −1.90 [CI −2.25 to −1.55] vs. −0.24 to −0.52, respectively). Effects were smaller but statistically significant when this trial was excluded (3 trials; SMD, −0.33 [CI, −0.56 to −0.09]; I2 = 0%). There were no statistically significant effects at other time points (SOE, low to moderate), with or without the outlier trial (Figures 6 to 8 of the Supplement).

    Appendix Table 2. Pooled Results: Epidural Corticosteroid Injections Versus Placebo Interventions for Radiculopathy

    Appendix Table 2.

    Epidural corticosteroid injections and placebo interventions did not differ in the likelihood of a successful outcome for pain (SOE, low to moderate), function (SOE, low), or a composite outcome (SOE, low to moderate) at any time point (Appendix Table 1 and Figures 9 to 17 of the Supplement). Epidural corticosteroid injections were associated with lower short-term risk for surgery than placebo interventions (8 trials; RR, 0.62 [CI, 0.41 to 0.92]; I2 = 0%; SOE, low) (38, 39, 45, 46, 54, 57, 88) (Figure 18 of the Supplement). The point estimate was similar but the difference no longer statistically significant when 3 poor-quality trials (38, 46) were excluded (5 trials; RR, 0.69 [CI, 0.42 to 1.13]). There was no difference in risk for long-term surgery (14 trials; RR, 0.97 [CI, 0.75 to 1.25]; I2 = 23%; SOE, moderate) (30, 34, 36, 37, 40, 41, 43, 44, 50, 53, 55, 56, 58, 89) (Figure 19 of the Supplement).

    For outcomes other than short-term surgery, exclusion of poor-quality trials had little effect on findings. Year of publication (before or after 2000) or blinding of patients or outcomes assessors also had no effect. Funnel plots did not suggest small sample effects (Figures 20 to 22 of the Supplement).

    Spinal Stenosis

    One good-quality trial (386 participants) found fluoroscopically guided interlaminar or transforaminal epidural corticosteroid injections to be associated with greater improvement in the RDQ at 3 weeks compared with an epidural local anesthetic (difference on a scale of 0 to 24, −1.8 [CI, −2.8 to −0.9]), although the difference was smaller and no longer statistically significant at 6 weeks (−1.0 [CI, −2.1 to 0.1]). There were no differences in the likelihood of having a greater than 30% or greater than 50% improvement in the RDQ or pain scores at 6 weeks, or on improvement in pain intensity at 3 or 6 weeks (59).

    Pooled analyses were consistent with the good-quality trial (Appendix Table 3), with small, non–statistically significant effects on pain intensity (WMD, 0.62 to 3.73 points) at short- and intermediate-term follow-up (SOE, low to moderate) (Figures 23 and 24 of the Supplement). Evidence on longer-term effects was sparse, and only 1 small trial (29 participants) evaluated effects on immediate pain intensity (WMD, −22.0 [CI, −36.0 to −8.00]; SOE, low) (61). There were no differences in functional improvement (Figure 25 of the Supplement) or likelihood of experiencing a successful pain, function, or composite outcome at any time point, although estimates were based on few trials (SOE low, except for short-term function [moderate]). Findings were similar when poor-quality trials were excluded.

    Appendix Table 3. Pooled Results: Epidural Corticosteroid Injections Versus Placebo Interventions for Spinal Stenosis

    Appendix Table 3.

    Technical Factors

    Five head-to-head trials of transforaminal versus interlaminar epidural corticosteroid injections for radiculopathy found no differences in reduction in pain intensity or function at any time point (76–78, 80, 90) (Appendix Table 4 and Figures 26 to 29 of the Supplement). Findings were similar when trials were stratified according to whether lower doses of corticosteroid were administered with the transforaminal approach (77, 90) or equivalent doses were administered with both approaches (76, 78, 80), or when a trial in which transforaminal injections did not clearly enter the epidural space (78) was excluded. There were also no clear differences on any outcome when placebo-controlled trials were stratified according to the approach used at any time point; however, estimates were imprecise (Appendix Table 5).

    Appendix Table 4. Pooled Results: Transforaminal Versus Interlaminar Epidural Corticosteroid Injections

    Appendix Table 4.

    Appendix Table 5. Epidural Corticosteroid Injections Versus Placebo Interventions, by Approach

    Appendix Table 5.

    A stratified analysis from a trial of epidural corticosteroid injections for spinal stenosis that permitted either the transforaminal or interlaminar approach found that only interlaminar corticosteroid injections were associated with greater improvement at 3 weeks on the RDQ (difference on a scale of 0 to 10, −2.5 [CI, −3.7 to −1.3]) and on leg pain (difference, −0.9 [CI, −1.5 to −0.3]) versus epidural local anesthetic (59). There was no effect of either approach on 6-week outcomes on the basis of the prespecified P value of 0.025 for subgroup analyses. Trials that compared alternative approaches (oblique interlaminar or lateral parasagittal) with standard interlaminar or transforaminal approaches for radiculopathy found no clear differences in pain, function, or other outcomes (46, 81–84). One fair-quality trial (239 participants) found the transforaminal ganglionic approach to be associated with a lower likelihood of overall “good” or “excellent” results compared with the preganglionic approach at 1 month (71% vs. 88%; RR, 0.80 [CI, 0.70 to 0.91]), although the difference was no longer present at longer (>6 month) follow-up (85).

    Head-to-head trials found no clear differences in outcomes among corticosteroids (4 trials [65–68]) or among corticosteroid doses (7 trials [69–75]). However, some estimates were imprecise, trials varied in the corticosteroids and doses that were compared, some trials that compared corticosteroids evaluated nontherapeutically equivalent doses (65, 66), and routes of administration and duration of follow-up varied (Appendix Table 1). All trials evaluated patients with radiculopathy, except for 1 trial of patients with spinal stenosis (69).

    No study directly compared epidural injections with versus without imaging guidance. One fair-quality trial (110 participants) found no differences between a caudal epidural injection with fluoroscopic plus Doppler guidance for chronic radicular pain versus fluoroscopic guidance alone in pain or ODI scores through 12 weeks (86). All placebo-controlled trials of the transforaminal approach used fluoroscopic guidance, and no trials of the interlaminar approach used fluoroscopic guidance. For spinal stenosis, there were no clear differences between trials that used or did not use fluoroscopic guidance, but analyses were limited by small numbers of trials. One good-quality trial of patients with radiculopathy (132 participants) found no difference between magnetic resonance imaging compared with history and physical examination alone to guide transforaminal or interlaminar epidural injections on any outcome through 3 months (87).

    No trial directly compared the effectiveness of epidural corticosteroid injections according to the local anesthetic used, number of injections, or number of levels injected. One trial found that if a first epidural corticosteroid injection was not successful, subsequent injections in the following 6 weeks were no more effective (30), and another study found no association between the number of injections and treatment response (68).

    Patient Characteristics

    Five trials found no association between duration of symptoms and epidural corticosteroid injection responsiveness after adjustment for other potentially contributing factors (30, 36, 85, 86, 91). A sixth trial found longer symptom duration to be associated with less favorable outcomes (56).

    Trials found no statistically significant interaction between age (36, 56, 85, 86), sex (36, 56, 85, 86), anxiety or depression (30, 56), opioid use (36), baseline function (30, 36), presence of neurologic abnormalities (30, 91), previous back episodes (30), or work status (30) on responsiveness to epidural corticosteroid injections. Three trials found no clear differences in estimates of effectiveness of injections for herniated disc versus spinal stenosis (37, 40, 86). Studies also found no clear effects of other specific imaging findings (56, 91, 92). No study evaluated effects of smoking status, body mass index, or use of opioid or other concomitant therapies.

    In meta-regression, exclusion of patients with prior surgery, requiring imaging or presence of herniated disc for enrollment, or duration of symptoms did not affect the estimates of effectiveness, although results were limited by small numbers of trials.

    Comparators

    Three trials of epidural corticosteroid injections for radiculopathy found no clear differences in estimates of effectiveness on the basis of different placebo comparators (41, 43, 45). There were no clear differences on any outcome when placebo-controlled trials were stratified by the type of comparator, although some estimates were imprecise (Appendix Table 6). There were also no clear differences between placebo comparators in the magnitude of improvement or proportion of responders (Table 1 of the Supplement).

    Appendix Table 6. Epidural Corticosteroid Injections Versus Placebo Interventions, by Type of Placebo Comparator

    Appendix Table 6.

    Harms

    In 30 placebo-controlled trials (2912 participants in total) of epidural corticosteroid injections for radiculopathy, 1 serious adverse event (a case of retroperitoneal hematoma in a patient receiving anticoagulation) (44) was reported. Methods for assessing harms were not well reported, and harms data were sparse. Thirteen trials did not report harms at all or reported no harms (32, 37, 39–42, 47, 48, 50, 51, 53, 58, 89).

    Three trials of the transforaminal versus interlaminar approaches did not report adverse events (77, 80, 90). One trial reported 1 case of transient hypertension with the transforaminal approach (78), and 1 trial reported no adverse events (76). Trials that compared alternative versus standard approaches reported few adverse events (81–84).

    Trials that compared corticosteroids did not report harms (66, 68) or reported no harms (65, 67). Harms were also poorly reported in 5 trials (452 participants) of corticosteroid dose comparisons, although no serious adverse events were reported (70, 71, 73–75).

    Eight placebo-controlled trials (821 participants in total) of epidural corticosteroid injections for spinal stenosis reported few harms, and no serious harms (37, 40, 59–64). One good-quality trial found transforaminal or interlaminar epidural corticosteroid injections to be associated with increased risk for at least 1 adverse event versus a local anesthetic injection (22% vs. 16%; RR, 1.39 [CI, 0.91 to 2.11]), but no difference in risk for serious adverse events (2.5% vs. 2.5%) (59). There was no clear difference in risk for adverse events between the interlaminar and transforaminal approaches. Among the other trials, 2 reported no harms (60) or no major harms (62), 2 did not report harms by treatment group (61, 62), and 3 did not report harms (37, 40, 64).

    Large observational studies of epidural and other spinal injections found serious adverse events to be rare, although minor adverse events, such as local hematoma, bleeding, return of blood, and dural puncture, were more common (93–96). In the largest study, there were no cases of nerve damage, infection, abscess, or epidural hematoma after 2760 lumbar epidural injections under fluoroscopic guidance (94). Rates of profuse bleeding ranged from 0.2% to 0.8%, depending on the approach used. There were no cases of transient nerve-root irritation after 3985 caudal injections, 4 cases (0.28%) among 1450 interlaminar injections, and 60 cases (4.6%) among 1310 transforaminal injections.

    Discussion

    Epidural corticosteroid injections for radiculopathy were associated with early improvements in some outcomes versus placebo interventions, but effects were small and unsustained, and epidural corticosteroid injections had no clear effects in patients with spinal stenosis. The strength of evidence ratings are summarized in Table 2 of the Supplement.

    Evidence was most robust on effects in patients with chronic radiculopathy. The only statistically significant effects were on immediate (5 days to ≤2 weeks) improvement in pain (WMD on a scale of 0 to 100, −7.55 [CI, −11.4 to −3.74]), and short-term (>2 weeks to ≤3 months) surgery risk (RR, 0.62 [CI, 0.41 to 0.92]). Immediate effects on function were statistically significant only when an outlier trial (54) was excluded (SMD, −0.33 [CI, −0.56 to −0.09]). Effects were below predefined minimum clinically important difference thresholds (15 points on a pain scale of 0 to 100, 10 points on the ODI, and 5 points on the RDQ [29]), with a WMD for pain of −7.55, and effects on function in the nonoutlier trials of 5.1 and 7.6 points on the ODI (33, 44) and 1.3 points on the RDQ (57). Effects were not present at longer-term follow-up, and there were no effects on the likelihood of experiencing a successful pain, function, or composite outcome. Results were generally robust in sensitivity and stratified analyses.

    Evidence for spinal stenosis was more limited, but showed no clear effects of epidural corticosteroid injections on pain or function. The only statistically significant effect was on immediate pain intensity, on the basis of a single small trial (29 participants) (61). Our analysis included a recent, well-conducted multicenter trial that was also the largest trial to date (59). This trial used a more pragmatic design in which injection approaches, corticosteroids, and doses varied, although there were no clear effects on the basis of such factors.

    Our findings are consistent with those of several recent systematic reviews, despite variability in the studies included and methods used for data synthesis and meta-analysis (8–10, 24, 97, 98). Our review strengthens and extends the findings of these prior reviews through the inclusion of additional trials; evaluation of continuous as well as dichotomous outcomes at predefined time points; and additional analyses on technical factors, patient characteristics, comparators, and methodological factors. Other systematic reviews reported more positive conclusions regarding the effectiveness of epidural corticosteroids (11–13, 99, 100). However, some of these reviews relied on qualitative synthesis, included observational studies, classified improvement from baseline after an epidural injection as demonstrating effectiveness even when there was no difference versus a placebo intervention, or focused on “positive” trials when there was inconsistency.

    Evidence on the effects of different approaches, corticosteroids, or doses on effectiveness of epidural corticosteroid injections was limited, but indicated no clear effects. There were also no clear effects from other patient or technical factors, such as use of imaging guidance, duration of symptoms, or injectate volume, on the basis of stratified and subgroup analyses. There were no clear differences in effectiveness and improvements were large across placebo comparators, suggesting that observed improvements represent the natural history or placebo effects, rather than a therapeutic effect of epidural local anesthetic, epidural saline, or soft-tissue injections. Although another systematic review found some evidence that epidural nonsteroid injections might be more effective than nonepidural injections, its conclusions were based on indirect comparisons that were highly discrepant with direct comparisons (20, 101).

    The assessment and reporting of harms data was suboptimal, but indicate a low risk for serious harms. Serious neurologic complications have been reported after lumbar epidural injections, and there was a recent outbreak of serious fungal infections due to contaminated methylprednisolone (7, 102, 103). Although there have been reports of increased risk for neurologic complications with use of particulate corticosteroids for cervical epidural injections, no cases were reported in lumbar injection trials.

    Our review had limitations. We used the Dersimonian–Laird random-effects model to pool studies, which may result in overly narrow CIs when heterogeneity is present (27). Therefore, we repeated analyses by using the profile likelihood method, which resulted in similar findings. Some meta-analyses were based on small numbers of trials, and we used indirect comparisons to supplement direct evidence; both should be interpreted with caution (104). We also excluded non–English-language articles. The evidence was limited by the small numbers of trials for some analyses and methodological limitations in the available trials. Of 58 trials, only 5 were rated good-quality. We did not include case series and other uncontrolled studies on harms, which could provide additional information (105).

    Additional research would clarify the benefits and harms of epidural corticosteroid injections. For radiculopathy, additional research is needed to determine whether such factors as the severity or duration of symptoms, presence of specific imaging findings, or presence of psychiatric comorbid conditions affect responsiveness to injections. Research is needed to determine whether injections are more effective when given in the context of a more comprehensive pain management approach. Additional trials that directly compare approaches, corticosteroids, doses, and use of imaging guidance are needed to augment limited data. For spinal stenosis, research is needed to determine whether there may be specific subgroups of patients who might benefit from epidural corticosteroids, who could be the target of future trials.

    In conclusion, epidural corticosteroid injections for radiculopathy are associated with immediate improvements in pain and might be associated with immediate improvements in function, but benefits are small and are not sustained, and there is no effect on long-term surgery risk. Evidence did not suggest that effectiveness varies on the basis of the approach used, corticosteroid, dose, or comparator. Limited evidence suggested that epidural corticosteroid injections are not effective for spinal stenosis.

    References

    • 1. Deyo RAMirza SKMartin BIBack pain prevalence and visit rates: estimates from U.S. national surveys, 2002. Spine (Phila Pa 1976)2006;31:2724-7. [PMID: 17077742] CrossrefMedlineGoogle Scholar
    • 2. Walker BFThe prevalence of low back pain: a systematic review of the literature from 1966 to 1998. J Spinal Disord2000;13:205-17. [PMID: 10872758] CrossrefMedlineGoogle Scholar
    • 3. Luo XPietrobon RSun SXLiu GGHey LEstimates and patterns of direct health care expenditures among individuals with back pain in the United States. Spine (Phila Pa 1976)2004;29:79-86. [PMID: 14699281] CrossrefMedlineGoogle Scholar
    • 4. Martin BIDeyo RAMirza SKTurner JAComstock BAHollingworth Wet alExpenditures and health status among adults with back and neck problems. JAMA2008;299:656-64. [PMID: 18270354] doi:10.1001/jama.299.6.656 CrossrefMedlineGoogle Scholar
    • 5. Stewart WFRicci JAChee EMorganstein DLipton RLost productive time and cost due to common pain conditions in the US workforce. JAMA2003;290:2443-54. [PMID: 14612481] CrossrefMedlineGoogle Scholar
    • 6. Chou RQaseem AOwens DKShekelle PClinical Guidelines Committee of the American College of PhysiciansDiagnostic imaging for low back pain: advice for high-value health care from the American College of Physicians. Ann Intern Med2011;154:181-9. [PMID: 21282698] doi:10.7326/0003-4819-154-3-201102010-00008 LinkGoogle Scholar
    • 7. Chou RAtlas SJStanos SPRosenquist RWNonsurgical interventional therapies for low back pain: a review of the evidence for an American Pain Society clinical practice guideline. Spine (Phila Pa 1976)2009;34:1078-93. [PMID: 19363456] doi:10.1097/BRS.0b013e3181a103b1 CrossrefMedlineGoogle Scholar
    • 8. Benoist MBoulu PHayem GEpidural steroid injections in the management of low-back pain with radiculopathy: an update of their efficacy and safety. Eur Spine J2012;21:204-13. [PMID: 21922288] doi:10.1007/s00586-011-2007-z CrossrefMedlineGoogle Scholar
    • 9. Quraishi NATransforaminal injection of corticosteroids for lumbar radiculopathy: systematic review and meta-analysis. Eur Spine J2012;21:214-9. [PMID: 21892774] doi:10.1007/s00586-011-2008-y CrossrefMedlineGoogle Scholar
    • 10. Pinto RZMaher CGFerreira MLHancock MOliveira VCMcLachlan AJet alEpidural corticosteroid injections in the management of sciatica: a systematic review and meta-analysis. Ann Intern Med2012;157:865-77. [PMID: 23362516] LinkGoogle Scholar
    • 11. Benyamin RMManchikanti LParr ATDiwan SSingh VFalco FJet alThe effectiveness of lumbar interlaminar epidural injections in managing chronic low back and lower extremity pain. Pain Physician2012;15:E363-404. [PMID: 22828691] CrossrefMedlineGoogle Scholar
    • 12. Manchikanti LBuenaventura RMManchikanti KNRuan XGupta SSmith HSet alEffectiveness of therapeutic lumbar transforaminal epidural steroid injections in managing lumbar spinal pain. Pain Physician2012;15:E199-245. [PMID: 22622912] CrossrefMedlineGoogle Scholar
    • 13. Parr ATManchikanti LHameed HConn AManchikanti KNBenyamin RMet alCaudal epidural injections in the management of chronic low back pain: a systematic appraisal of the literature. Pain Physician2012;15:E159-98. [PMID: 22622911] MedlineGoogle Scholar
    • 14. Chou RLoeser JDOwens DKRosenquist RWAtlas SJBaisden Jet alAmerican Pain Society Low Back Pain Guideline PanelInterventional therapies, surgery, and interdisciplinary rehabilitation for low back pain: an evidence-based clinical practice guideline from the American Pain Society. Spine (Phila Pa 1976)2009;34:1066-77. [PMID: 19363457] doi:10.1097/BRS.0b013e3181a1390d CrossrefMedlineGoogle Scholar
    • 15. American Academy of NeurologyAAN Summary of Evidence-based Guideline for Clinicians: Use of Epidural Steroid Injections to Treat Radicular Lumbosacral Pain. St. Paul, MN: American Academy of Neurology; 2007. Google Scholar
    • 16. Manchikanti LAbdi SAtluri SBenyamin RMBoswell MVBuenaventura RMet alAn update of comprehensive evidence-based guidelines for interventional techniques in chronic spinal pain. Part II: guidance and recommendations. Pain Physician2013;16:S49-283. [PMID: 23615883] MedlineGoogle Scholar
    • 17. Kreiner DSHwang SWEasa JEResnick DKBaisden JLBess Set alNorth American Spine SocietyAn evidence-based clinical guideline for the diagnosis and treatment of lumbar disc herniation with radiculopathy. Spine J2014;14:180-91. [PMID: 24239490] doi:10.1016/j.spinee.2013.08.003 CrossrefMedlineGoogle Scholar
    • 18. Friedly JChan LDeyo RIncreases in lumbosacral injections in the Medicare population: 1994 to 2001. Spine (Phila Pa 1976)2007;32:1754-60. [PMID: 17632396] CrossrefMedlineGoogle Scholar
    • 19. Dean DDMaruszak MBSandoval LFSpinal injections: trends in the US outpatient population. South Med J2014;107:528-9. [PMID: 25084194] doi:10.14423/SMJ.0000000000000144 CrossrefMedlineGoogle Scholar
    • 20. Bicket MCGupta ABrown CHCohen SPEpidural injections for spinal pain: a systematic review and meta-analysis evaluating the “control” injections in randomized controlled trials. Anesthesiology2013;119:907-31. [PMID: 24195874] doi:10.1097/ALN.0b013e31829c2ddd CrossrefMedlineGoogle Scholar
    • 21. Chou RHashimoto RFriedly JFu RDana TSullivan SDet alPain management injection therapies for low back pain. Technology Assessment Report no. ESIB0813. (Prepared by the Pacific Northwest Evidence-based Practice Center under contract no. HHSA 290-2012-00014-I.). Rockville, MD: Agency for Healthcare Research and Quality; 2015. [PMID: 25879124] Google Scholar
    • 22. Agency for Healthcare Research and QualityMethods guide for effectiveness and comparative effectiveness reviews. AHRQ publication no. 10(13)-EHC063-EF. Rockville MD. 2013. [PMID: 21433403] Google Scholar
    • 23. Pacific Northwest Evidence-based Practice Center; Agency for Healthcare Research and Quality. Systematic review protocol—pain management injection therapies for low-back pain. 29 May 2014. Accessed at www.ahrq.gov/research/findings/ta/topicrefinement/injection-protocol.pdf on 5 June 2015. Google Scholar
    • 24. Hashimoto RRaich AEcker EHenrikson NBWallace LDettori JRet alHealth Technology Assessment: Spinal Injections. Olympia, WA: Washington State Health Care Authority; 2011. Google Scholar
    • 25. Furlan ADPennick VBombardier Cvan Tulder MEditorial Board, Cochrane Back Review Group2009 updated method guidelines for systematic reviews in the Cochrane Back Review Group. Spine (Phila Pa 1976)2009;34:1929-41. [PMID: 19680101] doi:10.1097/BRS.0b013e3181b1c99f CrossrefMedlineGoogle Scholar
    • 26. Higgins JPThompson SGDeeks JJAltman DGMeasuring inconsistency in meta-analyses. BMJ2003;327:557-60. [PMID: 12958120] CrossrefMedlineGoogle Scholar
    • 27. Cornell JEMulrow CDLocalio RStack CBMeibohm ARGuallar Eet alRandom-effects meta-analysis of inconsistent effects: a time for change. Ann Intern Med2014;160:267-70. [PMID: 24727843] LinkGoogle Scholar
    • 28. Sterne JASutton AJIoannidis JPTerrin NJones DRLau Jet alRecommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ2011;343:d4002. [PMID: 21784880] doi:10.1136/bmj.d4002 CrossrefMedlineGoogle Scholar
    • 29. Ostelo RWDeyo RAStratford PWaddell GCroft PVon Korff Met alInterpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change. Spine (Phila Pa 1976)2008;33:90-4. [PMID: 18165753] doi:10.1097/BRS.0b013e31815e3a10 CrossrefMedlineGoogle Scholar
    • 30. Arden NKPrice CReading IStubbing JHazelgrove JDunne Cet alWEST Study GroupA multicentre randomized controlled trial of epidural corticosteroid injections for sciatica: the WEST study. Rheumatology (Oxford)2005;44:1399-406. [PMID: 16030082] CrossrefMedlineGoogle Scholar
    • 31. Béliveau PA comparison between epidural anaesthesia with and without corticosteroid in the treatment of sciatica. Rheumatol Phys Med1971;11:40-3. [PMID: 5551095] CrossrefMedlineGoogle Scholar
    • 32. Breivik HHesla PMolnar ILind BTreatment of chronic low back pain and sciatica. Comparison of caudal epidural injections of bupivacaine and methylprednisolone with bupivacaine followed by saline. Adv Pain Res Ther1976;1:927-32. Google Scholar
    • 33. Buchner MZeifang FBrocai DRSchiltenwolf MEpidural corticosteroid injection in the conservative management of sciatica. Clin Orthop Relat Res2000:149-56. [PMID: 10853164] CrossrefMedlineGoogle Scholar
    • 34. Bush KHillier SA controlled study of caudal epidural injections of triamcinolone plus procaine for the management of intractable sciatica. Spine (Phila Pa 1976)1991;16:572-5. [PMID: 2053000] CrossrefMedlineGoogle Scholar
    • 35. Carette SLeclaire RMarcoux SMorin FBlaise GASt-Pierre Aet alEpidural corticosteroid injections for sciatica due to herniated nucleus pulposus. N Engl J Med1997;336:1634-40. [PMID: 9171065] CrossrefMedlineGoogle Scholar
    • 36. Cohen SPWhite RLKurihara CLarkin TMChang AGriffith SRet alEpidural steroids, etanercept, or saline in subacute sciatica: a multicenter, randomized trial. Ann Intern Med2012;156:551-9. [PMID: 22508732] doi:10.7326/0003-4819-156-8-201204170-00397 LinkGoogle Scholar
    • 37. Cuckler JMBernini PAWiesel SWBooth RERothman RHPickens GTThe use of epidural steroids in the treatment of lumbar radicular pain. A prospective, randomized, double-blind study. J Bone Joint Surg Am1985;67:63-6. [PMID: 3155742] CrossrefMedlineGoogle Scholar
    • 38. Datta RUpadhyay KKA randomized clinical trial of three different steroid agents for treatment of low backache through the caudal route. Med J Armed Forces India2011;67:25-33. CrossrefMedlineGoogle Scholar
    • 39. Dilke TFBurry HCGrahame RExtradural corticosteroid injection in management of lumbar nerve root compression. Br Med J1973;2:635-7. [PMID: 4577015] CrossrefMedlineGoogle Scholar
    • 40. el Zahaar MSThe value of caudal epidural steroids in the treatment of lumbar neural compression syndromes. J Neurol Orthop Med Surg1991;12:181-4. Google Scholar
    • 41. Ghahreman AFerch RBogduk NThe efficacy of transforaminal injection of steroids for the treatment of lumbar radicular pain. Pain Med2010;11:1149-68. [PMID: 20704666] doi:10.1111/j.1526-4637.2010.00908.x CrossrefMedlineGoogle Scholar
    • 42. Helliwell MRobertson JEllis ROutpatient treatment of low-back pain and sciatica by a single extradural corticosteroid injection. Br J Clin Pract1985;39:228-31. Google Scholar
    • 43. Iversen TSolberg TKRomner BWilsgaard TTwisk JAnke Aet alEffect of caudal epidural steroid or saline injection in chronic lumbar radiculopathy: multicentre, blinded, randomised controlled trial. BMJ2011;343:d5278. [PMID: 21914755] doi:10.1136/bmj.d5278 CrossrefMedlineGoogle Scholar
    • 44. Karppinen JMalmivaara AKurunlahti MKyllönen EPienimäki TNieminen Pet alPeriradicular infiltration for sciatica: a randomized controlled trial. Spine (Phila Pa 1976)2001;26:1059-67. [PMID: 11337625] CrossrefMedlineGoogle Scholar
    • 45. Klenerman LGreenwood RDavenport HTWhite DCPeskett SLumbar epidural injections in the treatment of sciatica. Br J Rheumatol1984;23:35-8. [PMID: 6697071] CrossrefMedlineGoogle Scholar
    • 46. Kraemer JLudwig JBickert UOwczarek VTraupe MLumbar epidural perineural injection: a new technique. Eur Spine J1997;6:357-61. [PMID: 9391811] CrossrefMedlineGoogle Scholar
    • 47. Manchikanti LSingh VCash KAPampati VFalco FJA randomized, double-blind, active-control trial of the effectiveness of lumbar interlaminar epidural injections in disc herniation. Pain Physician2014;17:E61-74. [PMID: 24452658] CrossrefMedlineGoogle Scholar
    • 48. Manchikanti LSingh VCash KAPampati VDamron KSBoswell MVEffect of fluoroscopically guided caudal epidural steroid or local anesthetic injections in the treatment of lumbar disc herniation and radiculitis: a randomized, controlled, double blind trial with a two-year follow-up. Pain Physician2012;15:273-86. [PMID: 22828681] CrossrefMedlineGoogle Scholar
    • 49. Manchikanti LCash KAPampati VFalco FJTransforaminal epidural injections in chronic lumbar disc herniation: a randomized, double-blind, active-control trial. Pain Physician2014;17:E489-501. [PMID: 25054399] CrossrefMedlineGoogle Scholar
    • 50. Mathews JAMills SBJenkins VMGrimes SMMorkel MJMathews Wet alBack pain and sciatica: controlled trials of manipulation, traction, sclerosant and epidural injections. Br J Rheumatol1987;26:416-23. [PMID: 2961394] CrossrefMedlineGoogle Scholar
    • 51. Ridley MGKingsley GHGibson TGrahame ROutpatient lumbar epidural corticosteroid injection in the management of sciatica. Br J Rheumatol1988;27:295-9. [PMID: 3408828] CrossrefMedlineGoogle Scholar
    • 52. Riew KDYin YGilula LBridwell KHLenke LGLauryssen Cet alThe effect of nerve-root injections on the need for operative treatment of lumbar radicular pain. A prospective, randomized, controlled, double-blind study. J Bone Joint Surg Am2000;82-A:1589-93. [PMID: 11097449] CrossrefMedlineGoogle Scholar
    • 53. Rogers PNash TSchiller DNorman JEpidural steroids for sciatica. Pain Clinic1992;5:67-72. Google Scholar
    • 54. Sayegh FEKenanidis EIPapavasiliou KAPotoupnis MEKirkos JMKapetanos GAEfficacy of steroid and nonsteroid caudal epidural injections for low back pain and sciatica: a prospective, randomized, double-blind clinical trial. Spine (Phila Pa 1976)2009;34:1441-7. [PMID: 19525834] doi:10.1097/BRS.0b013e3181a4804a CrossrefMedlineGoogle Scholar
    • 55. Snoek WWeber HJørgensen BDouble blind evaluation of extradural methyl prednisolone for herniated lumbar discs. Acta Orthop Scand1977;48:635-41. [PMID: 343479] CrossrefMedlineGoogle Scholar
    • 56. Tafazal SNg LChaudhary NSell PCorticosteroids in peri-radicular infiltration for radicular pain: a randomised double blind controlled trial. One year results and subgroup analysis. Eur Spine J2009;18:1220-5. [PMID: 19387704] doi:10.1007/s00586-009-1000-2 CrossrefMedlineGoogle Scholar
    • 57. Valat JPGiraudeau BRozenberg SGoupille PBourgeois PMicheau-Beaugendre Vet alEpidural corticosteroid injections for sciatica: a randomised, double blind, controlled clinical trial. Ann Rheum Dis2003;62:639-43. [PMID: 12810426] CrossrefMedlineGoogle Scholar
    • 58. Wilson-MacDonald JBurt GGriffin DGlynn CEpidural steroid injection for nerve root compression. A randomised, controlled trial. J Bone Joint Surg Br2005;87:352-5. [PMID: 15773645] CrossrefMedlineGoogle Scholar
    • 59. Friedly JLComstock BATurner JAHeagerty PJDeyo RASullivan SDet alA randomized trial of epidural glucocorticoid injections for spinal stenosis. N Engl J Med2014;371:11-21. [PMID: 24988555] doi:10.1056/NEJMoa1313265 CrossrefMedlineGoogle Scholar
    • 60. Fukusaki MKobayashi IHara TSumikawa KSymptoms of spinal stenosis do not improve after epidural steroid injection. Clin J Pain1998;14:148-51. [PMID: 9647457] CrossrefMedlineGoogle Scholar
    • 61. Koc ZOzcakir SSivrioglu KGurbet AKucukoglu SEffectiveness of physical therapy and epidural steroid injections in lumbar spinal stenosis. Spine (Phila Pa 1976)2009;34:985-9. [PMID: 19404172] doi:10.1097/BRS.0b013e31819c0a6b CrossrefMedlineGoogle Scholar
    • 62. Manchikanti LCash KAMcManus CDDamron KSPampati VFalco FJLumbar interlaminar epidural injections in central spinal stenosis: preliminary results of a randomized, double-blind, active control trial. Pain Physician2012;15:51-63. [PMID: 22270738] CrossrefMedlineGoogle Scholar
    • 63. Manchikanti LCash KAMcManus CDPampati VFellows BResults of 2-year follow-up of a randomized, double-blind, controlled trial of fluoroscopic caudal epidural injections in central spinal stenosis. Pain Physician2012;15:371-84. [PMID: 22996849] CrossrefMedlineGoogle Scholar
    • 64. Nam HSPark YBEffects of transforaminal injection for degenerative lumbar scoliosis combined with spinal stenosis. Ann Rehabil Med2011;35:514-23. [PMID: 22506167] doi:10.5535/arm.2011.35.4.514 CrossrefMedlineGoogle Scholar
    • 65. Cocelli LPKarakurum GCebesoy OKaradasli HOner UClinical comparison of effectiveness of epidural triamcinolone and betamethasone in discal radiculalgia: a prospective, randomized study. J Musculoskelet Pain2009;17:281-6. CrossrefGoogle Scholar
    • 66. Kennedy DJPlastaras CCasey EVisco CJRittenberg JDConrad Bet alComparative effectiveness of lumbar transforaminal epidural steroid injections with particulate versus nonparticulate corticosteroids for lumbar radicular pain due to intervertebral disc herniation: a prospective, randomized, double-blind trial. Pain Med2014;15:548-55. [PMID: 24393129] doi:10.1111/pme.12325 CrossrefMedlineGoogle Scholar
    • 67. Kim DBrown JEfficacy and safety of lumbar epidural dexamethasone versus methylprednisolone in the treatment of lumbar radiculopathy: a comparison of soluble versus particulate steroids. Clin J Pain2011;27:518-22. [PMID: 21562412] doi:10.1097/AJP.0b013e31820c53e0 CrossrefMedlineGoogle Scholar
    • 68. Park CHLee SHKim BIComparison of the effectiveness of lumbar transforaminal epidural injection with particulate and nonparticulate corticosteroids in lumbar radiating pain. Pain Med2010;11:1654-8. [PMID: 20807343] doi:10.1111/j.1526-4637.2010.00941.x CrossrefMedlineGoogle Scholar
    • 69. Huda NBansal PGupta SMRuhela ARehman MAfzal MThe efficacy of epidural depo-methylprednisolone and triamcinolone acetate in relieving the symptoms of lumbar anal stenosis: a comparative study. J Clin Diagn Res2010;4:2843-7. Google Scholar
    • 70. McCahon RARavenscroft AHodgkinson VEvley RHardman JA pilot study of the dose-response of caudal methylprednisolone with levobupivacaine in chronic lower back pain. Anaesthesia2011;66:595-603. [PMID: 21564047] doi:10.1111/j.1365-2044.2011.06764.x CrossrefMedlineGoogle Scholar
    • 71. Owlia MBSalimzadeh AAlishiri GHaghighi AComparison of two doses of corticosteroid in epidural steroid injection for lumbar radicular pain. Singapore Med J2007;48:241-5. [PMID: 17342295] MedlineGoogle Scholar
    • 72. Habib GJabbour ASalman JHakim GHaddad HThe effect of epidural methylprednisolone acetate injection on the hypothalamic-pituitary-adrenal axis. J Clin Anesth2013;25:629-33. [PMID: 23988802] doi:10.1016/j.jclinane.2013.07.002 CrossrefMedlineGoogle Scholar
    • 73. Becker CHeidersdorf SDrewlo Sde Rodriguez SZKrämer JWillburger REEfficacy of epidural perineural injections with autologous conditioned serum for lumbar radicular compression: an investigator-initiated, prospective, double-blind, reference-controlled study. Spine (Phila Pa 1976)2007;32:1803-8. [PMID: 17762286] CrossrefMedlineGoogle Scholar
    • 74. Kang SSHwang BMSon HJCheong IYLee SJLee SHet alThe dosages of corticosteroid in transforaminal epidural steroid injections for lumbar radicular pain due to a herniated disc. Pain Physician2011;14:361-70. [PMID: 21785479] MedlineGoogle Scholar
    • 75. Ahadian FMMcGreevy KSchulteis GLumbar transforaminal epidural dexamethasone: a prospective, randomized, double-blind, dose-response trial. Reg Anesth Pain Med2011;36:572-8. [PMID: 22005659] doi:10.1097/AAP.0b013e318232e843 CrossrefMedlineGoogle Scholar
    • 76. Ackerman WEAhmad MThe efficacy of lumbar epidural steroid injections in patients with lumbar disc herniations. Anesth Analg2007;104:1217-22. [PMID: 17456677] CrossrefMedlineGoogle Scholar
    • 77. Gharibo CGVarlotta GPRhame EELiu ECBendo JAPerloff MDInterlaminar versus transforaminal epidural steroids for the treatment of subacute lumbar radicular pain: a randomized, blinded, prospective outcome study. Pain Physician2011;14:499-511. [PMID: 22086091] CrossrefMedlineGoogle Scholar
    • 78. Kolsi IDelecrin JBerthelot JMThomas LProst AMaugars YEfficacy of nerve root versus interspinous injections of glucocorticoids in the treatment of disk-related sciatica. A pilot, prospective, randomized, double-blind study. Joint Bone Spine2000;67:113-8. [PMID: 10769103] MedlineGoogle Scholar
    • 79. Rados ISakic Zdravcevic KHrgovic ZpainDETECT questionnaire and lumbar epidural steroid injection for chronic radiculopathy. Eur Neurol2013;69:27-32. [PMID: 23128915] doi:10.1159/000338265 CrossrefMedlineGoogle Scholar
    • 80. Thomas ECyteval CAbiad LPicot MCTaourel PBlotman FEfficacy of transforaminal versus interspinous corticosteroid injectionin discal radiculalgia—a prospective, randomised, double-blind study. Clin Rheumatol2003;22:299-304. [PMID: 14579160] CrossrefMedlineGoogle Scholar
    • 81. Candido KDRana MVSauer RChupatanakul LTharian AVasic Vet alConcordant pressure paresthesia during interlaminar lumbar epidural steroid injections correlates with pain relief in patients with unilateral radicular pain. Pain Physician2013;16:497-511. [PMID: 24077196] CrossrefMedlineGoogle Scholar
    • 82. Ghai BVadaje KSWig JDhillon MSLateral parasagittal versus midline interlaminar lumbar epidural steroid injection for management of low back pain with lumbosacral radicular pain: a double-blind, randomized study. Anesth Analg2013;117:219-27. [PMID: 23632053] doi:10.1213/ANE.0b013e3182910a15 CrossrefMedlineGoogle Scholar
    • 83. Candido KDRaghavendra MSChinthagada MBadiee STrepashko DWA prospective evaluation of iodinated contrast flow patterns with fluoroscopically guided lumbar epidural steroid injections: the lateral parasagittal interlaminar epidural approach versus the transforaminal epidural approach. Anesth Analg2008;106:638-44. [PMID: 18227326] doi:10.1213/ane.0b013e3181605e9b CrossrefMedlineGoogle Scholar
    • 84. Ghai BBansal DKay JPVadaje KSWig JTransforaminal versus parasagittal interlaminar epidural steroid injection in low back pain with radicular pain: a randomized, double-blind, active-control trial. Pain Physician2014;17:277-90. [PMID: 25054387] CrossrefMedlineGoogle Scholar
    • 85. Jeong HSLee JWKim SHMyung JSKim JHKang HSEffectiveness of transforaminal epidural steroid injection by using a preganglionic approach: a prospective randomized controlled study. Radiology2007;245:584-90. [PMID: 17940309] CrossrefMedlineGoogle Scholar
    • 86. Park YLee JHPark KDAhn JKPark JJee HUltrasound-guided vs. fluoroscopy-guided caudal epidural steroid injection for the treatment of unilateral lower lumbar radicular pain: a prospective, randomized, single-blind clinical study. Am J Phys Med Rehabil2013;92:575-86. [PMID: 23636087] doi:10.1097/PHM.0b013e318292356b CrossrefMedlineGoogle Scholar
    • 87. Cohen SPGupta AStrassels SAChristo PJErdek MAGriffith SRet alEffect of MRI on treatment results or decision making in patients with lumbosacral radiculopathy referred for epidural steroid injections: a multicenter, randomized controlled trial. Arch Intern Med2012;172:134-42. [PMID: 22157067] doi:10.1001/archinternmed.2011.593 CrossrefMedlineGoogle Scholar
    • 88. Ng LChaudhary NSell PThe efficacy of corticosteroids in periradicular infiltration for chronic radicular pain: a randomized, double-blind, controlled trial. Spine (Phila Pa 1976)2005;30:857-62. [PMID: 15834326] CrossrefMedlineGoogle Scholar
    • 89. Riew KDPark JBCho YSGilula LPatel ALenke LGet alNerve root blocks in the treatment of lumbar radicular pain. A minimum five-year follow-up. J Bone Joint Surg Am2006;88:1722-5. [PMID: 16882893] CrossrefMedlineGoogle Scholar
    • 90. Rados ISakic KFingler MKapural LEfficacy of interlaminar vs transforaminal epidural steroid injection for the treatment of chronic unilateral radicular pain: prospective, randomized study. Pain Med2011;12:1316-21. [PMID: 21914118] doi:10.1111/j.1526-4637.2011.01213.x CrossrefMedlineGoogle Scholar
    • 91. Ghahreman ABogduk NPredictors of a favorable response to transforaminal injection of steroids in patients with lumbar radicular pain due to disc herniation. Pain Med2011;12:871-9. [PMID: 21539702] doi:10.1111/j.1526-4637.2011.01116.x CrossrefMedlineGoogle Scholar
    • 92. Karppinen JOhinmaa AMalmivaara AKurunlahti MKyllönen EPienimäki Tet alCost effectiveness of periradicular infiltration for sciatica: subgroup analysis of a randomized controlled trial. Spine (Phila Pa 1976)2001;26:2587-95. [PMID: 11725240] CrossrefMedlineGoogle Scholar
    • 93. McGrath JMSchaefer MPMalkamaki DMIncidence and characteristics of complications from epidural steroid injections. Pain Med2011;12:726-31. [PMID: 21392252] doi:10.1111/j.1526-4637.2011.01077.x CrossrefMedlineGoogle Scholar
    • 94. Manchikanti LMalla YWargo BWCash KAPampati VFellows BA prospective evaluation of complications of 10,000 fluoroscopically directed epidural injections. Pain Physician2012;15:131-40. [PMID: 22430650] CrossrefMedlineGoogle Scholar
    • 95. Manchikanti LMalla YWargo BWCash KAPampati VFellows BComplications of fluoroscopically directed facet joint nerve blocks: a prospective evaluation of 7,500 episodes with 43,000 nerve blocks. Pain Physician2012;15:E143-50. [PMID: 22430660] CrossrefMedlineGoogle Scholar
    • 96. Manchikanti LMalla YWargo BWCash KAMcManus CDDamron KSet alA prospective evaluation of bleeding risk of interventional techniques in chronic pain. Pain Physician2011;14:317-29. [PMID: 21785475] CrossrefMedlineGoogle Scholar
    • 97. Shamliyan TAStaal JBGoldmann DSands-Lincoln MEpidural steroid injections for radicular lumbosacral pain: a systematic review. Phys Med Rehabil Clin N Am2014;25:471-89. [PMID: 24787344] doi:10.1016/j.pmr.2014.02.001 CrossrefMedlineGoogle Scholar
    • 98. Choi HJHahn SKim CHJang BHPark SLee SMet alEpidural steroid injection therapy for low back pain: a meta-analysis. Int J Technol Assess Health Care2013;29:244-53. [PMID: 23769210] doi:10.1017/S0266462313000342 CrossrefMedlineGoogle Scholar
    • 99. MacVicar JKing WLanders MHBogduk NThe effectiveness of lumbar transforaminal injection of steroids: a comprehensive review with systematic analysis of the published data. Pain Med2013;14:14-28. [PMID: 23110347] doi:10.1111/j.1526-4637.2012.01508.x CrossrefMedlineGoogle Scholar
    • 100. Roberts STWillick SERho MERittenberg JDEfficacy of lumbosacral transforaminal epidural steroid injections: a systematic review. PM R2009;1:657-68. [PMID: 19627959] doi:10.1016/j.pmrj.2009.04.008 CrossrefMedlineGoogle Scholar
    • 101. Edwards SJClarke MJWordsworth SBorrill JIndirect comparisons of treatments based on systematic reviews of randomised controlled trials. Int J Clin Pract2009;63:841-54. [PMID: 19490195] doi:10.1111/j.1742-1241.2009.02072.x CrossrefMedlineGoogle Scholar
    • 102. Epstein NEThe risks of epidural and transforaminal steroid injections in the spine: commentary and a comprehensive review of the literature. Surg Neurol Int2013;4:S74-93. [PMID: 23646278] doi:10.4103/2152-7806.109446 CrossrefMedlineGoogle Scholar
    • 103. Smith RMSchaefer MKKainer MAWise MFinks JDuwve Jet alMultistate Fungal Infection Outbreak Response TeamFungal infections associated with contaminated methylprednisolone injections. N Engl J Med2013;369:1598-609. [PMID: 23252499] doi:10.1056/NEJMoa1213978 CrossrefMedlineGoogle Scholar
    • 104. Song FLoke YKWalsh TGlenny AMEastwood AJAltman DGMethodological problems in the use of indirect comparisons for evaluating healthcare interventions: survey of published systematic reviews. BMJ2009;338:b1147. [PMID: 19346285] doi:10.1136/bmj.b1147 CrossrefMedlineGoogle Scholar
    • 105. U.S. Food and Drug Administration. Slides for the November 24-25, 2014 Meeting of the Anesthetic and Analgesic Drug Products Advisory Committee (AADPAC). 2014. Accessed at www.fda.gov/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/AnestheticAndAnalgesicDrugProductsAdvisoryCommittee/ucm425962.htm on 13 February 2015. Google Scholar
    • 106. Price CArden NCoglan LRogers PCost-effectiveness and safety of epidural steroids in the management of sciatica. Health Technol Assess2005;9:1-58. [PMID: 16095548] CrossrefMedlineGoogle Scholar
    • 107. Manchikanti LSingh VFalco FJCash KAPampati VEvaluation of the effectiveness of lumbar interlaminar epidural injections in managing chronic pain of lumbar disc herniation or radiculitis: a randomized, double-blind, controlled trial. Pain Physician2010;13:343-55. [PMID: 20648203] CrossrefMedlineGoogle Scholar
    • 108. Manchikanti LSingh VCash KAPampati VFalco FJThe role of fluoroscopic interlaminar epidural injections in managing chronic pain of lumbar disc herniation or radiculitis: a randomized, double-blind trial. Pain Pract2013;13:547-58. [PMID: 23279452] doi:10.1111/papr.12023 CrossrefMedlineGoogle Scholar
    • 109. Manchikanti LSingh VCash KAPampati VDamron KSBoswell MVA randomized, controlled, double-blind trial of fluoroscopic caudal epidural injections in the treatment of lumbar disc herniation and radiculitis. Spine (Phila Pa 1976)2011;36:1897-905. [PMID: 21897343] doi:10.1097/BRS.0b013e31823294f2 CrossrefMedlineGoogle Scholar
    • 110. Manchikanti LSingh VCash KAPampati VDamron KSBoswell MVPreliminary results of a randomized, equivalence trial of fluoroscopic caudal epidural injections in managing chronic low back pain: part 2—disc herniation and radiculitis. Pain Physician2008;11:801-15. [PMID: 19057627] CrossrefMedlineGoogle Scholar
    • 111. Manchikanti LCash KAMcManus CDPampati VFellows BFluoroscopic caudal epidural injections with or without steroids in managing pain of lumbar spinal stenosis: one-year results of randomized, double-blind, active-controlled trial. J Spinal Disord Tech2012;25:226-34. [PMID: 22652990] doi:10.1097/BSD.0b013e3182160068 CrossrefMedlineGoogle Scholar
    • 112. Manchikanti LCash KAMcManus CDPampati VAbdi SPreliminary results of a randomized, equivalence trial of fluoroscopic caudal epidural injections in managing chronic low back pain: part 4—spinal stenosis. Pain Physician2008;11:833-48. [PMID: 19057629] CrossrefMedlineGoogle Scholar

    Comments

    0 Comments
    Sign In to Submit A Comment
    Harry W. Daniell, MD, FACP23 October 2015
    Comment

    In their comprehensive review, Chou et al (1) were unable to identify factors which could be used to predict any benefit resulting from epidural corticosteroid injections (ECI), a benefit presumed to result from corticosteroid anti-inflammatory activity. None of the reviewed investigations discussed potential decreases in this activity by the pro-inflammatory activity induced by suppression of post-ECI systemic cortisone levels, which includes any associated increases in pro-inflammatory cytokines, any therapeutic benefit from ECI likely reflecting the sum of these anti-inflammatory and pro-inflammatory factors.
    Cortisol levels are independently suppressed by opioids (2), benzodiazepines (3), and multiple other medications including propofol. Post-ECI cortisol decreases typically last for 2-3 weeks but Kay et al (4) demonstrated them to be enhanced and prolonged following midazolam premedication in 14 patients randomized to receive midazolam or placebo premedication before each of 3 weekly ECIs. Therapeutic benefits in their subjects were not reported.
    Kirpalani et al (5) suggested less effective cervical ECI in chronic opioid consumers, with it documented in only 1 of 5 but present in 7 of 10 non-consumers (P=0.06).
    After treating 3 chronic sustained-action opioid consumers hospitalized with symptomatic adrenal insufficiency soon after lumbar ECI, and four similar outpatients in a solo practice of general internal medicine, serial cortisol levels were measured after lumbar ECI in 6 additional sustained-action opioid consumers. Each of these 13 patients had received 40-80 mg of triamcinolone by ECI, 4 male and 9 female, ages 20-89 with daily morphine sulfate equivalent (MSE) consumption of 50-250 (median 90). None had known adrenal disease or had received other corticosteroids within the preceding 6 months. Each had received premedication before ECI with midazolam, fentanyl, and propofol. Post-ECI cortisol levels are presented in Figure 1.Seven subjects had reported post-ECI weakness, hypotension, or syncope which had been partially explained by chronic benzodiazepine use in 4 and anti-hypertensive medications in 4 others.
    Cortisol levels remained subnormal for more than 4 weeks post-ECI in 9 of 11 patients, with none of the 13 developing convincing benefits from ECI.
    Our observations support strong inhibition of cortisol formation by their combined medications, without associated clinical benefit from ECI, suggesting potential value in studies designed to examine the possibility of more predictable post-ECI benefit during limitation of post-ECI cortisol levels by minimizing the chronic and premedication use of drugs known to suppress cortisol formation, while also suggesting possible added therapeutic benefit by post-ECI cortisol replacement.

    References:

    1.     Chou R, Hashimoto R, Friedly J, et al. Epidural Corticosteroid Injections for Radiculopathy and Spinal Stenosis. Ann Intern Med. 2015;163:373-381.

    2.     Brennan MJ. The effect of opioid therapy on endocrine function. Am J Med. 2013;126(3 suppl 1):S12-8.

    3.     Arvat E, Giordano R, Grottoli S, Ghigo E. Benzodiazepines and anterior pituitary function. J Endocrinol Invest. 2002;25:735-747.

    4.     Kay J, Findling JW, Raff H. Epidural triamcinolone suppresses the pituitary-adrenal axis in human subjects. Anesth Analg. 1994;79:501-505.

    5.     Kirpalani D, Mitra R. Is chronic opioid use a negative predictive factor for response to cervical epidural steroid injections? J Back Musculoskelet Rehabil. 2011;24:123-127.

    Roger Chou, Robin Hashimoto, Janna Friedly21 October 2015
    In response
    Manchikanti et al suggest that local anesthetic epidural injections have long-term therapeutic effects and should not be considered a placebo. Our definition of placebo was based on the assumption that epidural injection therapeutic effects are primarily related to the corticosteroid (1). This is consistent with clinical practice and the rational for epidural injections, which are predicated on corticosteroid anti-inflammatory properties. Further, our findings were similar when analyses were stratified according to the type of placebo comparator used (local anesthetic epidural, saline epidural, soft tissue injection, or no injection), arguing against a specific local anesthetic therapeutic effect. In fact, one trial found that pain relief rates were higher with intramuscular saline (13%) or epidural saline (19%) than with epidural local anesthetic (7%) (1).

    Maus et al suggest that diagnoses of radiculopathy were too heterogeneous to reach reliable conclusions. To clarify, all of the radiculopathy trials enrolled patients with radicular symptoms, with or without low back pain. Based on meta-regression analyses, requiring imaging correlation or a herniated disc on imaging did not impact estimates.

    Maus et al critique our article for including outdated injection techniques, such as trials without imaging guidance. Yet we stratified analyses by the approach used; results for transforaminal injections (all of which were performed with imaging guidance) were similar to the overall findings. It is also incorrect to suggest that analyses were solely based on continuous outcome measures. We also found no differences in categorical measures of pain, function, or composite measures of success at any time point.

    Riaz et al note that it is helpful to discuss the findings of systematic reviews in the context of prior reviews, which we did. The review cited by Riaz et al (3) did not include the largest trial on epidural corticosteroids for spinal stenosis, relied on qualitative synthesis, and classified trial results as positive based on improvement from baseline, even in the absence of differences versus a control intervention. The review (4) cited by Maus et al did not rate study quality and appeared to weight observational studies equally with or higher than randomized trials.

    Manchikanti et al suggest that the report was not conducted in accordance with current standards for conducting systematic reviews. However, our review adheres to current methodological standards (5), including risk of bias assessment, evidence synthesis methods, and disclosure of financial and non-financial conflicts of interest prior to conducting the review.

    References

    (1) Chou R, Hashimoto R, Friedly J, Fu R, et al. Epidural corticosteroid injections for radiculopathy and spinal stenosis: A systematic review and meta-analysis. Ann Intern Med 2015;163:373-81.
    (2) Ghahreman A, Ferch R, Bogduk N. The efficacy of transforaminal injection of steroids for the treatment of lumbar radicular pain. Pain Medicine 2010;11:1149-68.
    (3) Manchikanti L, Kaye AD, Manchikanti K, Boswell M, Pampati V, Hirsch J. Efficacy of epidural injections in the treatment of lumbar central spinal stenosis: A systematic review. Anesth Pain Med 2015;5:e23139.
    (4) MacVicar J, King W, Landers MH, Bogduk N. The effectiveness of lumbar transforaminal injection of steroids. A comprehensive review with systematic analysis of the published data. Pain Med 2013;14:14-28.
    (5) Methods Guide for Effectiveness and Comparative Effectiveness Reviews. AHRQ Publication No. 10(14)-EHC063-EF. Rockville, MD: Agency for Healthcare Research and Quality. January 2014. Chapters available at: www.effectivehealthcare.ahrq.gov.

    Disclosures: Grants received by Dr. Chou, Dr. Friedly, and Dr. Hashimoto from the Agency for Healthcare Research and Quality for conducting the review that is the subject of this response.

    Timothy Maus, Belinda Duszynski2 October 2015
    Comment
    The publication by Chou and colleagues (1) raises significant concerns among physicians treating patients suffering from radicular pain and associated functional impairment. Fourteen medical societies formulated a consensus response to the Agency for Healthcare Quality and Research’s technology assessment, the basis of the current publication, addressing the flawed methodology and resulting aberrant conclusions. (2)

    The authors assert the nihilistic position, without evidence, that back and leg pain are un-attributable to a specific cause and, therefore, include studies with patient selection by symptoms, not diagnosis. Current literature demonstrates that radicular and somatic back pain can be specifically diagnosed with systematic application of diagnostic blocks or provocative procedures, synthesized with clinical examination, advanced imaging and electrophysiology. (3) Their position has led to inclusion of heterogeneous study populations; in 29 studies of “epidural steroid injection” versus placebo, radicular pain alone was specified in 22, a mixture of radicular and back pain in six, and back pain alone in one. Correlative imaging findings were required in only 11 studies, leaving the nature of the compressive lesions and degree of compression unknown. With literature demonstrating these factors influence the natural history and efficacy of epidural steroid injections, it is inappropriate to draw conclusions from these heterogeneous studies.

    The review is a corruption of evidence-based medicine -- omitting the best available evidence: high quality outcome studies of homogenous patients with contrast confirmation of injectate delivery to the target. Rather, it includes decades-old trials of unguided epidural injections by several routes. Only 7 of the 29 placebo-controlled trials utilized image guidance. Reliance on flawed RCTs leads the authors to conclude there is no evidence supporting the use of image guidance, placing it in conflict with the FDA Safe Use Initiative, which mandates image guidance.

    The authors’ conclusions are based on invalid statistical analyses, primarily changes in mean pain scores, which are insufficient for drawing conclusions about effectiveness. A National Institutes of Health task force recommended the utilization of categorical outcomes for studying low back pain. (4)

    When inappropriate statistics are applied to review heterogeneous populations given heterogeneous treatments, with equal weight given to outdated procedural techniques, results should be viewed with skepticism. A comprehensive examination of the literature, including high quality contemporary outcomes studies of homogenous patient populations, reveals that in carefully selected patients, epidural steroid injections performed to exacting procedural standards provide pain relief and functional improvement in patients suffering from radicular pain. (5)

    Sincerely,


    American Association of Neurological Surgeons
    American Academy of Pain Medicine
    American Academy of Physical Medicine and Rehabilitation
    American College of Radiology
    American Pain Society
    American Society of Anesthesiologists
    American Society of Neuroradiology
    American Society of Regional Anesthesia and Pain Medicine
    American Society of Spine Radiology
    Congress of Neurological Surgeons
    Spine Intervention Society
    North American Neuromodulation Society
    North American Spine Society
    Society of Interventional Radiology



    1. Chou R, Hashimoto R, Friedly J, Fu R, Bougatsos C, Dana T, Sullivan SD, Jarvik J. Epidural corticosteroid injections for radiculopathy and spinal stenosis: a systematic review and meta-analysis. Ann Intern Med 2015;163(5):373-381.
    2. Multisociety Pain Work Group Letter to Dr. Elise Berliner, AHRQ; July 29, 2015. http://1515docs.org/pdfs/MPW_Letter_to_AHRQ_7-29-2015.pdf
    3. DePalma MJ. Diagnostic nihilism toward low back pain: what once was accepted, should no longer be. Pain Med 2015;16(8):1453-4.
    4. Deyo RA, Dworkin SF, Amtmann D, Andersson G, Borenstein D, Carragee E, Carrino J, Chou R, Cook K, DeLitto A, Goertz C, Khalsa P, Loeser J, Mackey S, Panagis J, Rainville J, Tosteson T, Turk D, Korff MV, Weiner DK. Report of the NIH Task Force on research standards for chronic low back pain. Pain Med 2014;15(8):1249-67.
    5. MacVicar J, King W, Landers MH, Bogduk N. The effectiveness of lumbar transforaminal injection of steroids: a comprehensive review with systematic analysis of the published data. Pain Med 2013;14(1):14-28.

    Laxmaiah Manchikanti, MD, Alan David Kaye MD PhD, Joshua A. Hirsch, MD 3 September 2015
    Comment
    We are concerned that the systematic review by Chou et al (1) will have far reaching consequences on patients who might potentially benefit from epidural corticosteroid injections for radiculopathy and spinal stenosis. This systematic review is similar to an earlier publication by Pinto et al (2). The fundamental flaw of this systematic review and previous one (2) is that the authors have converted all placebo to active-control trials, with unproven hypothesis that all therapeutic effects in the epidural space are secondary to steroids (1). In our opinion, this is a scientifically and clinically incorrect methodology. Noteworthy, the therapeutic effectiveness of local anesthetics on a long-term basis has been illustrated in systematic reviews with an efficacy that was equivalent to steroids except in some circumstances (3). Further, methodologically sound systematic reviews showed efficacy of epidural injections in managing radiculopathy, spinal stenosis, and other ailments (3-5). Chou et al’s conversion of active-control trials to placebo control is analogous to studies comparing whole milk to water or skim milk, with water being placebo and skim milk being an active-control. With extensive literature available in reference to placebos and nocebos and their influence on various trials, this approach is unscientific and unjustifiable.
    Further, epidural injections have an excellent risk-benefit ratio compared to opioids and NSAIDs, which are themselves responsible for almost 17,000 deaths a year and numerous hospitalizations. Lumbar surgery alone is responsible for over almost 1,300 deaths a year, while deaths over the past two decades related to epidural injections were 131 -- significantly less than any other modality (4,5).
    Other deficiencies include inconsistency with standards developed by the Institute of Medicine (IOM) for systematic reviews, perceived intellectual bias and inappropriate methodological quality assessment of the manuscripts. Further, they (1,2) have misinterpreted outcomes assessment data as absolute difference between 2 active control groups, which is not feasible, because active control trials only demonstrate superiority, non-inferiority, or equivalency rather than efficacy. We posit that the absolute effect size can only be measured by a true placebo control – not an impure placebo or one converted from an active agent to placebo on paper.
    The policy implications of Chou et al’s systematic review are such that patients will lose access to epidural injections for radiculopathy and spinal stenosis, and seek alternative treatments including narcotic medications and surgery.

    References
    1. Chou R, Hashimoto R, Friedly J, Fu R, Bougatsos C, Dana T, Sullivan SD, Jarvik J. Epidural corticosteroid injections for radiculopathy and spinal stenosis: A systematic review and meta-analysis. Ann Intern Med 2015; [Epub ahead of print].
    2. Pinto RZ, Maher CG, Ferreira ML, Hancock M, Oliveira VC, McLachlan AJ, Koes B, Ferreira PH. Epidural corticosteroid injections in the management of sciatica: A systematic review and meta-analysis. Ann Intern Med 2012; 157:865-877.
    3. Manchikanti L, Nampiaparampil DE, Manchikanti KN, Falco FJE, Singh V, Benyamin RM, Kaye AD, Sehgal N, Soin A, Simopoulos TT, Bakshi S, Gharibo CG, Gilligan CJ, Hirsch JA. Comparison of the efficacy of saline, local anesthetics, and steroids in epidural and facet joint injections for the management of spinal pain: A systematic review of randomized controlled trials. Surg Neurol Int 2015; 6:S194-S235.
    4. Manchikanti L, Benyamin RM, Falco FJ, Kaye AD, Hirsch JA. Do epidural injections provide short- and long-term relief for lumbar disc herniation? A systematic review. Clin Orthop Relat Res 2015; 473:1940-1956.
    5. Manchikanti L, Kaye AD, Manchikanti KN, Boswell MV, Pampati V, Hirsch JA. Efficacy of epidural injections in the treatment of lumbar central spinal stenosis: A systematic review. Anesth Pain Med 2015; 5:e23139.
    Siefferman J, Kiritsy M25 November 2015
    Challenges of studying epidural injections
    Studying epidural injections for “low back pain” (LBP) presents many technical challenges. LBP may involve multiple anatomical and physiological etiologies, has acute and chronic phases, and the epidural itself may be performed in a variety of ways, with different approaches and using a variety of medications. Additionally, no true placebo has been identified. The available literature evaluating the efficacy of lumbar epidural injections are highly varied in methodology, with heterogeneous diagnostic inclusion criteria, placebo controls, procedures tested, medications tested, and follow up times.
    Similar to Chou and colleagues, and in advance of their publication, we also performed a systematic review of lumbar epidural injections with the specific goal of teasing out which factors effect efficacy: anatomic etiology of pain, injection approach, and type and volume of injectate. Our conclusions differed about the following:
    • Epidural steroid is effective for reducing radicular pain associated with a herniated disc. (Multiple level I evidence studies)
    • For undifferentiated causes of LBP, epidural injection of any variety is superior to non-epidural interventions. (Multiple level I evidence studies)
    • Epidural injection (steroid or saline) is effective for reducing LBP associated with spinal stenosis. (Multiple level I and II evidence studies)
    When comparing epidurals with or without steroid, 9 out of 14 studies demonstrated no difference in the degree of pain relief. However, the 4 of the 5 studies that did favor steroid were examining patients with radicular pain from herniated discs, and the fifth was a meta-analysis examining all causes of low back pain [1-5]. Therefore, there is mixed level I evidence favoring the use of epidural steroid for low back pain of any cause, but strong level I evidence supporting its use for radicular pain due to herniated disc.

    References
    1. Bicket, M.C., et al., Epidural injections for spinal pain: a systematic review and meta-analysis evaluating the "control" injections in randomized controlled trials. Anesthesiology, 2013. 119(4): p. 907-31.
    2. Datta, R. and K.K. Upadhyay, A randomized clinical trial of three different steroid agents for treatment of low backache through the caudal route. Medical Journal Armed Forces India. 67(1): p. 25-33.
    3. Ghahreman, A., R. Ferch, and N. Bogduk, The efficacy of transforaminal injection of steroids for the treatment of lumbar radicular pain. Pain Med, 2010. 11(8): p. 1149-68.
    4. Manchikanti, L., et al., A randomized, controlled, double-blind trial of fluoroscopic caudal epidural injections in the treatment of lumbar disc herniation and radiculitis. Spine (Phila Pa 1976), 2011. 36(23): p. 1897-905.
    5. Manchikanti, L., et al., A randomized, double-blind, active-control trial of the effectiveness of lumbar interlaminar epidural injections in disc herniation. Pain Physician, 2014. 17(1): p. E61-74.

    Irbaz Bin Riaz, Robert G Badgett2 October 2015
    Updating PRISMA to avoid conflicting evidence from Systematic Reviews and Meta Analyses
    The recent meta-analysis by Chou et al (1)showed that epidural corticosteroids injections were not effective for the Spinal Stenosis. Although the results of this systematic review sponsored by Agency for Healthcare Research and Quality (AHRQ) are similar to a previous Cochrane meta-analysis (2), juxtaposition of these reviews highlights two difficulties for readers attempting to synthesize a conclusion from contradictory reviews. Firstly, regarding the inclusion of trials, two rigorous sponsors of meta-analyses, AHRQ and Cochrane, disagreed on the included studies. Of the 5 trials published during a period accessible to both analyses, only two (Cuckler and Koc) were included by both reviews. Secondly, regarding conclusions, the recent AHRQ review might have better served readers if it had reconciled its conclusion with the review by Manchikanti (3). For example, the AHRQ review could have acknowledged the review of Manchikanti and explained that the review did not pool studies and so did not have a basis for the positive conclusion that it offered. Admittedly, the difference in included studies and lack of reconciliation of conclusions did not significantly influence the outcome in this example, it certainly could happen in future reviews of more controversial topics. We propose that the PRISMA statement be amended so that when prior, relevant systematic reviews exist, authors of new reviews should provide a reconciliation of conclusions of the current and prior review, along with a table reconciling the studies included. We believe this will assist readers when they are confronted by conflicting meta-analyses.

    References:

    1. Chou R, Hashimoto R, Friedly J, Fu R, Bougatsos C, Dana T, et al. Epidural Corticosteroid Injections for Radiculopathy and Spinal StenosisA Systematic Review and Meta-analysisEpidural Corticosteroid Injections for Radiculopathy and Spinal Stenosis. Annals of Internal Medicine. 2015;163(5):373-81.
    2. Ammendolia C, Stuber KJ, Rok E, Rampersaud R, Kennedy CA, Pennick V, et al. Nonoperative treatment for lumbar spinal stenosis with neurogenic claudication. The Cochrane database of systematic reviews. 2013;8:Cd010712.
    3. Manchikanti L, Kaye AD, Manchikanti K, Boswell M, Pampati V, Hirsch J. Efficacy of epidural injections in the treatment of lumbar central spinal stenosis: a systematic review. Anesthesiology and pain medicine. 2015;5(1):e23139.
    G Allison Glass DSc, Neil Maneck BS, Howard T Katz MD9 September 2015
    Heterogeniety and use of guidance
    We read with great interest Chou et al’s (1) review of the benefits of epidural corticosteroid injections (ECI) in adults with radiculopathy and concluded that the benefits were small and short lived. We are treated a 53-year-old female with left S1 radiculopathy who has recently received her second ECI and reports significant and lasting relief in pain and improved function as measured both in our office and that of another provider. The injections were performed under guidance by a third provider. As this is counter to the finding of Chou et al (1), we did our own review of the eight sources cited by the authors in their analysis of long-term follow-up of function for individuals with radiculopathy (Appendix Table 2 – Function). Of these 8 studies, 5 found significant improvement in function and pain following injection up to at least one year. Three of the studies found no long term benefit. It was notable that of the two studies showing little or no long term benefit, Arden et al (2) based their conclusion on the WEST study which did not use imaging guidance for the placement of the injection, Bush and Hilliar (3) who did not use imaging guidance and Iversen et al (4) who used anatomic markers and a 10 MHz ultrasound transducer. Four studies were performed using real time fluoroscopic imaging and found positive results with both steroid and local anesthetic injections. One study employed the careful use of anatomical markers, and large volumes (up to 20 ml) and saw long term positive results. There was not consistency with respect to corticosteroid or local anesthetic used from study to study nor with the injection volume employed.

    A careful examination of the sources cited by Chou et al (1) leads us to suggest that epidural injections, employing fluoroscopic guidance, do in fact provide statistically significant long term (greater than one year) relief from radicular pain and improvement in function. Whether this effect is due to the injection itself, the local anesthetic or the corticosteroid remains an open question. The effect of the choice of corticosteroid and the injection volume employed also remains a question. Perhaps greatest conclusion that can be deduced from Chou et al (1) is the need for a standardized protocol for ECIs.

    References:
    1. Chou R, Hashimoto R, Friedly J, Fu R, Bougatses C, Dana T, Sullivan SD, Jarvik J. Epidural Corticosteroid Injections for radiculopathy and spinal stenosis. Ann Intern Med. 2015,;163:171-81.
    2. Arden NK, Price C, Reading I., Stubing J, Hazelgrove J., Dunne C, et al; WEST Study Group. A multicentre randomized controlled trail of epidural corticosteroid injections for sciatica: the WEST study. Rheumatology (Oxford). 2005;44:1399-406.
    3. Bush K, Hillier S. A controlled study of caudal epidural injections of triamcinologne plus procaine for the management of intractable sciatica. Spine. 1991;16:572-5.
    4. Iversen T, Solberg TK, Romner B, Wilsgaard T, Twisk J, Anke Am, et al. Effect of caudal epidural steroid injection in chronic lumbar radiculopathy:multicentre, linded, randomize control trail. Spine. 2001;343:d5278.

    Denis F. Darko, MD, FACP, DLFAPA8 September 2015
    Epidural Corticosteroid Injections for Radiculopathy and Spinal Stenosis
    TO THE EDITOR: Chou and colleagues have tried to address an important area of patient care (1) with all the literature that is available, which unfortunately is not much. This small body of data illustrates more what we do not know than what we do know. The paucity of quality studies in this high profile area of medicine is embarrassing. As the authors describe in the Discussion, much, much more remains to be done.

    The authors only briefly mention near the end of the Discussion that, "Research is needed to determine whether injections are more effective when given in the context of a more comprehensive pain management approach." While limited evidence exists for the usefulness of specifically designed physical therapy (PT) programs for both radiculopathy (2, 3) and degenerative lumbar spinal stenosis (2, 4), clinical experience suggests that PT might be effective in improving patient outcomes when put in place as part of an appropriate overall treatment plan.

    In practical terms, the immediate-term reduction in pain with epidural corticosteroid injections as was found by the authors might provide a window to begin a productive program of active PT. The combination of epidural corticosteroid use followed by active PT should be studied with longer-term follow-up using the outcome measures of improvements in pain and function and change in risk for requiring surgical intervention.

    Denis F. Darko, MD, FACP, DLFAPA
    NeuroSci R&D Consultancy, LLC
    Plymouth, MN 55446-3745

    1. Chou R, Hashimoto R, Friedly J, Fu R, Bougatsis C, Dana T, et al. Epidural Corticosteroid Injections for Radiculopathy and Spinal Stenosis: A Systematic Review and Meta-analysis. Ann Intern Med. 2015; 163:373-381. [PMID: 26302454] doi:10.7326/M15-0934

    2. Chou R, Qaseem A, Snow V, Casey D, Cross JT Jr., Shekelle P, et al. Diagnosis and Treatment of Low Back Pain: A Joint Clinical Practice Guideline from the American College of Physicians and the American Pain Society. Ann Intern Med. 2007;147:478-491. [PMID: 17909209]

    3. North American Spine Society, Clinical Guidelines for Multidisciplinary Spine Care, Diagnosis and Treatment of Lumbar Disc Herniation with Radiculopathy, 2012: 30.

    4. North American Spine Society, Clinical Guidelines for Multidisciplinary Spine Care, Diagnosis and Treatment of Degenerative Lumbar Spinal Stenosis, 2011: 43.