Low back pain has a high prevalence, and the economic costs of this epidemic per year in the United States alone have an estimated range from $100 million to $50 billion with annual direct cost per person of around US $8000.1–3 Approximately 40% of cases of low back pain are associated with lumbosacral radicular pain, a condition caused usually by nerve root irritation and inflammation from herniated intervertebral disks.4,5 Lumbosacral radicular pain is characterized by pain arising in the back and radiating into the lower limbs in the distribution of one or more spinal nerves.6,7 In addition to the inflammatory reaction in the nerve roots, changes may occur in ion channel functioning of sensory neurons.8–10 More than half of the patients report a decrease in their daily activities and their ability to work.11 One in every 4 patients continues to have severe pain despite the use of oral pharmacologic treatments,5,12 whereas 14% of patients eventually require surgery for severe pain, especially if the pain is accompanied by neurologic deficits.13 Operative intervention, however, is expensive, is associated with morbidity, and its long-term benefits are unclear.14
Injection of steroids into the epidural space often is used in an attempt to ameliorate lumbosacral radicular pain and bypass the need for operative intervention, but the evidence for benefit from epidural steroids is unclear, despite the plethora of investigations in animals, case series, retrospective studies, prospective clinical trials, and systematic reviews. Previous systematic reviews and meta-analyses (SR-MAs) on the efficacy and effectiveness of epidural steroids for lumbosacral radicular pain have some important limitations. First, published reviews have included trials of a variety of epidural delivery routes besides transforaminal (e.g., interlaminar and caudal),15,16 but these routes often do not result in the steroids reaching the exiting spinal nerve, anterior epidural space, and the dorsal root ganglion17,18—the desired sites of action to counteract the inflammation secondary to compression from herniated intervertebral disks. It is now recognized that fluoroscopically guided injections through the transforaminal epidural (TFE) route (i.e., around the emerging spinal nerve roots) are the best option for ensuring delivery of steroids near the site of the pathology and are likely to be more effective.9
The second limitation of published reviews is the broad inclusion criteria in most published trials. Data from patients with lumbosacral radicular pain attributable to spinal stenosis (a chronic, often multivertebral level pathology) often are combined with data from subjects with herniated intervertebral disks.4,16,19 This makes it difficult for readers to appreciate the true therapeutic benefit of epidural steroids in patients with lumbosacral radicular pain due to herniated intervertebral disks. In the most recent SR-MA on TFE steroids for lumbosacral radicular pain published in 2012, data from only 3 trials conducted before 2010 could be used for quantitative analysis, and there was no attempt to stratify administered doses of steroids or to examine the quality of evidence.20 Lastly, previous reviews did not systematically grade the quality of the evidence of included trials.
The primary objective of this SR-MA was to determine the analgesic efficacy of TFE steroids compared with local anesthetics (LAs) and/or saline or conservative management for lumbosacral radicular pain secondary to herniated intervertebral disks 1 to 3 months after injection. We chose LA and saline as comparators because of a recently published multicenter large trial on epidural injections for spinal stenosis that showed equianalgesic benefit of LA against a combination of LA and steroids.21 Secondary objectives were to assess the impact of this intervention on pain-associated domains (anxiety, depression, physical activity, quality of life), incidence of operative intervention, and adverse effects. Clinical recommendations based on the guidelines from the Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group22 also were made based on the result of this SR-MA.
This systematic review was conducted according to the recommendations of the Cochrane Collaboration,23 and it is reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplemental Digital Content 1, https://links.lww.com/AA/B354).24
Search Strategy and Study Selection
We conducted comprehensive, serial searches of the literature through February 17, 2015. The following databases were searched: Embase, 1947 onward; Medline, 1946 onward; Medline In-Process and Other Non-Indexed Citations (all using the OvidSP Platform); and Cochrane Database of Systematic Reviews. PROSPERO and Cochrane Central Register of Controlled Trials were included to identify reviews or trials that may have been published but missed during the initial search on MEDLINE and EMBASE. We also searched Google Scholar (first 200 search results were reviewed) to compliment search results from the aforementioned databases with the objective of accessing all content relevant to the topic. Proceedings of the major annual meetings of anesthesiology and pain societies (American Society of Anesthesiologists, European Society of Anaesthesiology, International Association for the Study of Pain, American Society of Regional Anesthesia and Pain Medicine, European Society of Regional Anaesthesia and Pain Therapy, and World Institute of Pain) in the last 2 years also were searched. We also searched for randomized controlled trials (RCTs) in the meta-register of Controlled Trials (clinicaltrials.gov). We restricted our search to trials involving human subjects, but no limitations on language were imposed. Finally, experts with clinical and research experience on the role of epidural steroids for axial and radicular pain were also consulted.
For Embase and Medline, both controlled vocabulary terms (Embase—Emtree; Medline—MeSH) and text word searching were conducted for each of the following search segments: radicular pain/radiculopathy; herniated lumbar intervertebral disk/herniation; steroids (methylprednisolone, triamcinolone, betamethasone, dexamethasone); epidural/TFE/selective nerve root; injections/blocks. Each search was limited to English language and human subjects. We applied a highly sensitive search strategy to identify studies.25 Details of our search strategy are provided in Supplemental Digital Content 2 (https://links.lww.com/AA/B355). We completed the search by reviewing the bibliographies of every selected article to look for possible additional articles that had not been retrieved by the search. Two authors independently evaluated titles, abstracts, and full texts according to the inclusion criteria. All instances of discordance were discussed between the investigators to reach a consensus.
Criteria for Considering Studies for This Review
We considered only RCTs for this review that compared the effect of TFE steroid injections (with or without LA) versus TFE saline or LA injection or conservative management.
Trials with a follow-up period of at least 1 month after injection were included.
Diagnosis of Lumbosacral Radicular Pain Secondary to Herniated Intervertebral Disks:
Trials that included radiological diagnosis of herniated intervertebral disks with associated compression of exiting nerve roots were included. We prespecified eligibility criteria using the population, intervention, comparison, and outcomes (i.e., PICO) approach.
We included studies of community-dwelling participants of both sexes who were older than 18 year and who had lumbosacral radicular pain in the lower limb secondary to herniated intervertebral disks, with or without low back pain with intensity of pain equal to or greater than 4 on a 0 to 10 numerical rating scale (NRS; i.e., pain of moderate-to-severe intensity).
Interventions and Comparators
Intervention was defined as injection of any of the 4 different steroids (methylprednisolone, triamcinolone, betamethasone, and dexamethasone) administered through only the TFE route under fluoroscopic guidance. The comparators were TFE saline or LA injection or conservative management (pharmacotherapy [antidepressants, anticonvulsants, anti-inflammatories, opioids], physiotherapy, or a combination of these) for the treatment of lumbosacral radicular pain. We did not impose restrictions on the number of TFE injections. In studies with additional cohorts apart from those comparing combinations of steroids, LA, or saline, we attempted to extract the data for the comparison between TFE steroids and LA or saline. Studies involving TFE use of cytokine antagonists, clonidine, and hypertonic saline were excluded because of the experimental nature of these interventions. We also excluded trials on neuromodulation (central or peripheral), radiofrequency (pulsed or ablation), or other surgical modalities (microdiscectomy/laminectomy) for the treatment of lumbosacral radicular pain due to herniated intervertebral disks.
The primary outcome was pain at 1 to 12 months after the intervention, expressed on an NRS (0—no pain to 10—worst possible pain). Intensity scores reported on a visual analog scale (0—no pain to 100—worst possible pain) were transformed to a 0 to 10 NRS scale. We chose to include trials that assessed the primary outcome at a wide range of time intervals because duration of the analgesic effect of long-acting steroids (e.g., methylprednisolone) injected in the epidural space is often assessed at 1, 3, 6, and even 12 months in trials. It has also been suggested that larger doses may have a longer duration of action.4 The following were considered as secondary outcomes if measured on validated scales: physical disability for low back and leg pain (Oswestry Disability Index [ODI] or Roland-Morris Disability Questionnaire [RMDQ]), psychological disability (anxiety and depression), and quality of life. Incidence of surgery up to 1 year after the interventions and adverse effects of TFE injections of steroids (both procedure [dural puncture, hematoma, infections, injury to spinal nerves, or cord] and medication [hyperglycemia, hypertension, fluid retention, weight gain, myopathy, osteoporosis and fractures, infections, psychosis, and cataracts] related) also were included.
Risk of Bias Assessment for Individual Trials
Two review authors independently assessed the risk of bias for each included study using the Cochrane Collaboration’s tool for assessing the risk of bias.23 Any disagreement was resolved through discussion or, if necessary, arbitration by a third reviewer. This tool assesses bias in the following domains: generation of the allocation sequence, allocation concealment, blinding of investigators and participants, blinding of outcome assessors, incomplete outcome data, selective outcome reporting, and any other sources of bias. Each item is classified as low, unclear, or high risk of bias. A decision to classify “overall bias” as low, unclear, or high was made by the reviewers using the following method:
- High: any trial with a high risk of bias listed on 3 or more domains or significant methodologic concerns that may have affected results
- Unclear: any trial with a high risk of bias listed on 2 domains or moderate methodologic concerns that may have affected the study results
- Low: any trial with a high risk of bias on none or 1 domain and with no significant methodologic concerns that may have affected the study results.
We also intended to use a funnel plot (and assessment of its asymmetry to assess publication biases [citation bias, selective outcome reporting]) and the Begg rank correlation test26 if a sufficient number of studies (10 or more) were available to apply this test.
The reference data, populations, and outcomes were extracted from the articles into prespecified tables. The 2 authors used a standardized data extraction procedure. The data-collection form was pilot-tested before its use. We extracted information on studies’ general characteristics (including design, number of arms, and primary outcomes), participants (characteristics of the populations, sample size, duration, and intensity of pain), and experimental intervention (type of steroid, doses, and administration regimes). Dichotomous outcomes were extracted as the incidence of surgical intervention and the presence or absence of adverse effect. For continuous data (pain NRS scores, ODI, or RMDQ scores), we extracted means and SDs. If not reported, the SDs were obtained from confidence intervals (CIs) or P values that are related to the differences between means in the 2 groups. Median values and interquartile ranges were converted to mean and SD if data appeared to be normally distributed.27
Data Synthesis and Analysis
We expected heterogeneity (because of diverse populations and doses of steroids included), and therefore, we used DerSimonian and Laird random effects meta-analysis models.28 Heterogeneity was assessed with the Q test, and Higgins I2 statistic was used to quantify it (I2 > 50% indicates substantial heterogeneity). The estimated mean effect of each study of these outcomes was calculated with the respective 95% CIs, and the pooled effect was then assessed. The Mantel-Haenszel method was used for calculating the pooled relative risk (risk ratio) with corresponding 95% CI. Investigation of sources of heterogeneity was based on the analysis of prespecified subgroups for the primary outcome. The definition of the subgroups included steroid dose per injection (high versus low; low was defined as methylprednisolone dose of up to 40 mg or an equipotent dose of another steroid) and quality of trials (high versus unclear or low). Subgroup analysis for dose of steroids was performed to assess whether there was a “dose−response” relationship between amount of steroids used and analgesic benefit. We also performed a post hoc sensitivity analysis for the primary outcome to evaluate the effect of removing trials with a strongly positive result in favor of steroids on the primary outcome measure. This is based on a published literature on the analgesic effect of perineural steroids on peripheral neuropathic pain.29 All statistical analyses were performed with the Review Manager (RevMan version 5.2.5; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark).
Quality of Evidence
The quality of evidence for the outcomes was assessed with GRADE methodology.30,31 The quality of evidence was classified as high, moderate, low, or very low for each outcome based on the risk of bias, inconsistency, indirectness, imprecision, and other considerations (publication bias). A summary table was constructed with GRADEpro version 3.6 (http://www.guidelinedevelopment.org/; Evidence Prime Inc., Hamilton, Ontario, Canada). We mentioned the reason for downgrading quality of any evidence in a footnote in the GRADE table.
The search results and study selection flow chart are reported in Figure 1. From the initial 385 records identified through database searching, 348 records were screened but 296 of these were excluded because the publications were not an SR-MA/RCT or failed to report the primary or secondary outcomes. Fifty-two full-text articles were assessed for eligibility, and 44 were excluded because the study settings (pathology, intervention) did not meet our inclusion criteria (intraoperative setting, failed back surgery syndrome, piriformis syndrome, routes of epidural injection other than TFE, and comparison of variations of TFE approaches). Eight RCTs including 771 patients (366 in steroid and 405 in comparator groups) were included for SR-MA (Fig. 1).32–39 The 8 RCTs that were included in data synthesis were published during the last 16 years (2000–2015).
Trial, Participants, and Intervention Characteristics
The median (range) sample size of the 8 RCTs was 86 (50–160) patients. All the RCTs enrolled patients with lumbosacral radicular pain with/without low back pain secondary to herniated intervertebral disks, and the diagnosis of herniated intervertebral disks was confirmed by magnetic resonance imaging or computed tomography. Intensity of pain was moderate or severe (4 or greater on a 0–10 NRS score), and the minimum duration of pain for inclusion in the RCTs was 1 month in all trials except the study by Ghahreman et al.37 in which patients with lesser duration of pain (1 day and more) were included. The duration of preintervention pain was 1 to 6 months in most RCTs. All RCTs involved comparison of TFE steroids against epidural (interlaminar or transforaminal) LA or saline, except one trial that had injection of saline in the paraspinal muscles as a comparator.34 We chose to include this trial because the placebo effects from TFE or IM saline are probably similar. Some of these RCTs involved more than 1 injection of steroid and comparator,32,34,37–39 but data on the mean/median number of procedures were available in only 2 trials.37,38 The interval between procedures was 1 to 4 weeks in RCTs that used repeat procedures. All procedures were performed under fluoroscopic guidance (Table 1).
Steroids used in the RCTs were betamethasone, methylprednisolone, and triamcinolone, and the dose per procedure (in terms of methylprednisolone dose equivalents)40,41 was 40 to 70 mg. Steroids were combined with LAs (lidocaine or bupivacaine) in all the RCTs, and the comparator injectates were equivalent volumes of LA32,35,36 or 0.9% saline (TFE or in paraspinal muscles)33,34 or both.37–39 Of the 2 RCTs that had both LA and 0.9% saline as comparator groups, 1 trial had 3 groups of TFE injectates (steroids and LA, saline and LA, and etanercept)38 and the other trial had 5 groups of injectates (TFE steroids and LA, TFE LA, TFE saline, IM steroids, and IM saline).37 Data for TFE steroid and LA, LA and saline, only LA, and only saline groups were extracted from these trials, and data from nonsteroid groups were combined as comparator for the purpose of meta-analysis. Data for subjects in TFE etanercept group from the RCT by Cohen et al.38 were not included in the meta-analysis.
Change in NRS for pain was the primary outcome in all RCTs except for the study by Riew et al.,32 in which the proportion of patients requiring surgical intervention was the primary outcome and pain scores were not reported. NRS for pain was recorded from 1 to 12 months after the intervention in 7 RCTs with 3 months being the median time interval. Secondary outcomes were recorded in 7 RCTs and included changes in physical disability as measured on ODI33,35,36,38,39 or RMDQ.34,37 The proportion of patients requiring surgery was recorded in all RCTs except 2,34,39 and the duration of follow-up for recording this outcome varied from 3 to 28 months, with 12 months as the median. Impact of the interventions on quality of life was reported in only 2 RCTs.33,37 None of the RCTs included evaluation of changes in psychological function (anxiety and depression) or specifically sought procedure and steroid-related adverse effects, but adverse effects were reported anecdotally in 3 trials.33,38,39
Risk of Bias Assessment of Included Studies
Overall risk of bias was assessed as low in 4 trials,33,35,36,38 high in 2 trials,32,34 and unclear in 2.37,39 The trials deemed to have a high risk of bias did not adequately describe the procedure for allocation concealment and had incomplete outcome data,32,34 and the trials with unclear bias had incomplete outcome data37 or lack of details regarding blinding of outcome assessment.39 There was no difference in opinion between the 2 reviewers for the assessment of risk of bias for all 8 trials. Quality of reporting (as per the U.S. Preventive Services Task Force criteria)42 was good in 3 trials,33,35,38 fair in 3 trials,36,37,39 and poor in 2 trials (Fig. 2).32,34
Primary Outcome: Pain Scores After Interventions
The interval between intervention and procedure was a median value of 3 months with the range being 1 to 3 months for 6 of the RCTs. Data in the study by Vad et al.,34 which “randomized” patients by choice to epidural steroid injections or trigger point injections, were provided only at 1 year after the intervention. We included this study because pain scores at the end of a year after the intervention can be expected to be greater and not lower than those at 3 months after the intervention, and this assumption strengthened the null hypothesis that steroids confer no analgesic benefit. It is important to note that no saline or LA was injected in the epidural space as a comparator in this trial. However, similar to studies that compared epidural steroids to epidural LA or saline, or IM injections performed in a blinded fashion, it shares the potential for the placebo effect to explain the prolonged analgesic benefit observed in the control group, which justifies its inclusion. Two of the RCTS showed clinically meaningful reduction in pain in the TFE steroid group (pain NRS scores 20% lower compared with control group),34,37 2 showed a modest reduction (pain NRS scores 5%–20% lower compared with control group),35,36 and the other 3 did not show any benefit of steroids over comparators (Table 2; Fig. 3).33,38,39
Meta-analysis of the 7 RCTs that recorded pain NRS scores (716 patients: 338 in steroid and 378 in comparator groups)33–39 showed that there was a significant but clinically modest reduction in mean pain scores (0–10 scale) in patients receiving steroids compared with LA/saline (−0.97 points; 95% CI: −1.42 to −0.51 points; P < 0.0001, I2 = 90%; Fig. 3) at a median duration of 3 months after the intervention. The difference in mean pain NRS scores between the 2 groups represented approximately 10% absolute reduction and 14% reduction from the mean baseline pain scores (average of 7 on a 0–10 NRS). Table 2 shows the “within-group” change in mean pain scores from baseline for both steroid and LA/saline groups: the range for steroid groups was 32% to 82% and that for LA/saline groups was 7% to 62%.
Impact of Interventions on Physical Disability
Validated measures of assessing self-rated disability because of low back were used in all 8 RCTs except 1.32 ODI was used in 5 RCTs,33,35,36,38,39 and RMDQ was used in 2 RCTs.34,37 ODI is scored from 0 to 100, whereas RMDQ is scored from 0 to 24, with 0 indicating no disability and greater scores indicating increasing disability on both measures.43 We conducted a meta-analysis of the 5 RCTs that used ODI. The results indicated no significant difference in mean ODI scores in patients receiving steroids (285 patients) compared with LA/saline (289 patients) (−0.89 points; 95% CI: −2.60 to 0.81 points; P = 0.30, I2 = 68%) at 1 to 3 months after the interventions (Fig. 4). Of the 2 RCTs that used RMDQ to assess disability, one showed significantly lower scores (indicating improvement in physical function) in the LA/saline group at 16 months after interventions34 and the other RCT did not report on the significance of lower reported scores in the steroid group (Table 3; Fig. 4).37
Impact of Interventions on Psychological Function and Quality of Life
None of the 8 RCTs used validated measures to assess change in anxiety and/or depression after the interventions, and change in quality of life was reported in only 2 RCTs.33,37 One of these trials used Nottingham Health Profile to assess the impact of interventions on quality of life.33 The Nottingham Health Profile scores of the 2 groups in this trial were similar at 1, 3, 6 months, and 1 year after the interventions. The other RCT used Short Form 36 Health Survey to track changes in health status, but no significant difference was noticed in the steroid and comparator groups after the interventions.37
Impact of Interventions on Incidence of Surgery
All but 234,39 of the 8 RCTs reported incidence of surgery after the interventions. The time period for recording operative intervention varied from 3 to 28 months with 12 months being the median value. Criteria for surgical intervention (e.g., persistent pain, motor deficits) were not clearly described in any of the RCTs. We conducted a meta-analysis of the 6 RCTs that had operative intervention as an outcome. The results indicated no difference in the probability of operative intervention in patients who had received TFE steroids (228 patients) compared with those who had received LA/saline (266 patients) (relative risk: 0.88; 95% CI: 0.55 to 1.41; P = 0.59, I2 = 53%; Fig. 5). Only 1 of the 6 RCTs showed a significant reduction in operative intervention in patients who received TFE steroids (8 of 28 patients had surgery) compared with those who received LA (18 of 28 patients required surgery) at a median interval of 23 months after the interventions (P < 0.004; Fig. 5).32
Impact of Interventions on Incidence of Adverse Effects
Only 2 RCTs mentioned procedure-related adverse effects. Karppinen et al.33 reported that one patient who was taking anticoagulants and who received TFE steroids developed a retroperitoneal hematoma. Manchikanti et al.39 reported an incidence of 28 (4.6%) for intravascular infiltrations and 9 (1.5%) for nerve root irritations in the 601 TFE injections performed in both groups in their study. None of the RCTs except one38 reported any other procedure-related (e.g., nerve or spinal cord injury) or steroid-related (e.g., hyperglycemia, hypertension, fluid retention, weight gain, myopathy, osteoporosis and fractures, infections, psychosis, cataracts) adverse effects. In the RCT by Cohen et al.,38 6 of 30 patients in TFE saline group and 1 of 28 patients in TFE steroid group had minor adverse events (e.g., worsening of pain after injection), 3 of 30 patients in TFE saline group developed nonlocal infection, and 1 of 28 patients in TFE steroid group reported a nonlocal rash.
Heterogeneity and Publication Bias
For the primary outcome, the I2 statistic was 90% for comparison of mean postintervention pain scores, indicating high heterogeneity. Several characteristics of these studies can contribute to heterogeneity, and we explored 2 of these by subgroup analysis (Table 4). The characteristics that were evaluated included doses of TFE steroids (≤40 vs 40 mg) and risk of bias in included studies (low versus high).
Steroid Dose per Injection
We explored the analgesic impact of using methylprednisolone (or equivalent doses of other steroids) in low (up to 40 mg per injection) or high (greater than 40 mg per injection) doses. This “cutoff” was based on the dichotomy of TFE steroid doses for the treatment of lumbosacral radicular pain secondary to herniated intervertebral disks.4 Four trials used methylprednisolone or betamethasone in low doses.33,35,36,39 Meta-analysis of these RCTs (516 patients: 257 in steroid and 259 in comparator groups) showed that there was a significant, but clinically modest, reduction in mean pain scores (0–10 scale) in patients receiving steroids compared with LA/saline (−0.54 points; 95% CI: −0.67 to −0.42 points; P < 0.00001, I2 = 10%) at 1 to 3 months after the interventions. Meta-analysis of data from 3 trials that used greater doses of steroids34,37,38 (200 patients: 81 in steroid group and 119 in comparator groups) showed significant and clinically meaningful reduction in mean pain scores (0–10 scale) in patients receiving steroids compared with LA/saline (−2.04 points; 95% CI: −2.42 to −1.65 points; P < 0.00001, I2 = 0%) at 1 to 12 months after the interventions.
Risk of Bias
We performed a meta-analysis of data for the primary outcome from the 4 trials that had an overall low risk of bias.33,35,36,38 Meta-analysis of these RCTs (454 patients: 225 in steroid and 229 in comparator groups) showed that there was a significant, but clinically modest, reduction in mean pain scores (0–10 scale) in patients receiving steroids compared with LA/saline (−0.57 points; 95% CI: −0.68 to −0.46 points; P < 0.00001, I2 = 0%) at 1 to 3 months after the interventions. Meta-analysis of the 3 RCTs with an overall unclear or high risk of bias (262 patients: 113 in steroid and 149 in comparator groups) showed that there was a significant but greater reduction in mean pain scores (0–10 scale) in patients receiving steroids compared with LA/saline (−1.44 points; 95% CI: −2.72 to −0.15 points; P = 0.03, I2 = 93%) at 1 to 3 months after the interventions.34,37,39 CI results for meta-analysis of data from 5 or less trials should be interpreted with caution because they may be too narrow. Wider CIs for this analysis would result in the upper limit of the interval crossing zero, thereby indicating no significant analgesic effect.
Publication bias was not assessed with a funnel plot because low number of eligible studies (<10) decreases the power of the tests to distinguish chance from real asymmetry.43,44
A sensitivity analysis was not planned a priori, but it was performed by removing 2 trials with the results that were strongly in favor of TFE steroids.34,37 Meta-analysis of the 5 remaining RCTs (574 patients: 285 in steroid and 289 in comparator groups) showed that there was a significant but clinically modest (and lower) reduction in mean pain scores (0–10 scale) in patients receiving steroids compared with LA/saline (−0.55 points; 95% CI: −0.67 to −0.43 points; P < 0.00001, I2 = 6%) at 1 to 3 months after the interventions.
On the basis of the quality of evidence and the strength of effect, it can be recommended that, in outpatients with lumbosacral radicular pain secondary to herniated intervertebral disks, tansforaminal epidural steroids should be used to reduce pain at 3 months after the intervention (strong recommendation ↑; moderate-quality evidence). The modest analgesic benefit should be discussed with patients, and their preferences and values considered before proceeding with this intervention. This intervention should not be used to reduce physical disability at 1 to 3 months after the intervention (strong recommendation ↓; high-quality evidence) or incidence of surgery at 12 months after the intervention (strong recommendation ↓; moderate-quality evidence) (Fig. 6). Further research is justified because there is a lack of information about appropriate dosages and number of procedures, effect of the intervention on psychological disability and quality of life, and adverse effects.
This is the first SR-MA of RCTs published during the last 15 years on the analgesic efficacy of TFE steroids in relieving lumbosacral radicular pain secondary to herniated intervertebral disks. It reveals that 40 to 70 mg of methylprednisolone (or equivalent doses of other steroids) is associated with analgesic benefit compared with TFE LA or saline at 3 months after the intervention. The reduction in mean pain scores (0–10 scale) in patients receiving steroids compared with LA/saline (−0.97 points; 95% CI: −1.42 to −0.51 points; P < 0.0001, I2 = 90%) is significant but clinically modest. The quality of evidence is moderate and the strength of recommendation strong for the analgesic efficacy of TFE steroids in this clinical setting (Fig. 6). TFE steroids did not have any advantage over TFE LA or saline in reducing physical disability at 1 to 3 months or incidence of surgery at 1 year after the interventions. Moreover, the studies included in this meta-analysis have inadequately reported the effect of TFE steroids on psychological disability, quality of life, and adverse effects of the interventions.
Population Included in This Review
There was variation in the duration of pain reported by the participants (1 day to over 6 months), number of TFE injections performed (1 to 4), and time points for the assessment of primary and secondary outcomes (1 month to 1 year) in the 8 RCTs included in this SR-MA. These factors contributed to the clinical heterogeneity of our review.
To enhance the validity of our results, we attempted to restrict our inclusion to only studies that involved TFE injections under fluoroscopic guidance. Patients in all the studies included in this review had mean pain scores of over 6 of 10 on NRS, thus confirming that they had pain of moderate or severe intensity.
Efficacy of TFE Steroids in Relieving Lumbosacral Radicular Pain Secondary to Intervertebral Disk Herniation
Although the pooled results showed a modest reduction in pain NRS scores (0.9 points on a 0–10 scale) at 3 months after the interventions, meta-analysis of the 3 trials that used higher doses of steroids (equivalent to methylprednisolone 60–70 mg) showed an impressive reduction in pain NRS (2 points; Table 4).34,37,38 This may be an evidence of a dose−response effect. Moreover, baseline pain scores were relatively greater in one of these trials34 compared with the other trials, and the minimum interval between onset of pain and intervention was the shortest in the other trial (1 day).37 Finally, all 3 trials allowed repeat procedures (2 or 3) to be performed. One or more of these factors may have contributed to the enhanced analgesic response observed in these trials. Of the 3 trials that did not show an analgesic benefit of TFE steroid over LA or saline injections, relatively low doses of steroids were used in 2 trials (betamethasone 3 mg [equipotent dose of methylprednisolone 20 mg] and methylprednisolone 40 mg)33,39 but not in the third trial (methylprednisolone 40–60 mg),38 whereas a repeat procedure was allowed within 3 weeks in 2 of these trials.38,39
A review of published literature revealed that studies that compared 40 and 80 mg of methylprednisolone administered through the midline epidural route did not show better efficacy for the greater dose in patients with chronic low back pain45 or lumbosacral radicular pain secondary to herniated intervertebral disks.46 Another study involved 2 TFE steroid injections at an interval of 1 week to compare the analgesic effect of cumulative methylprednisolone equivalent doses of 40 mg or lower at 1 week after the second injection. There was no difference in analgesic benefit from doses of triamcinolone 20, 30, or 40 mg (same dose for methylprednisolone equivalents), but the main limitation was the short follow-up period.47 Finally, the analgesic efficacy of TFE steroids was explored for 4, 8, and 12 mg of dexamethasone (20, 40, and 60 mg doses in methylprednisolone equivalents), but there was no difference in mean pain scores at 12 weeks after the interventions.48 None of these studies had a comparator group of LA/saline/conservative management. It is important to recognize that none of the studies in our SR-MA compared one dose of TFE steroids with another dose or different steroids or a single procedure versus repeat procedures. These factors need to be explored in the setting of an RCT.
Efficacy of TFE Steroids in Reducing Physical Disability in Patients with Lumbosacral Radicular Pain Secondary to Intervertebral Disk Herniation
ODI and RMDQ were used to assess physical disability in the RCTs included in this SR-MA. We performed a meta-analysis only for ODI scores because it was reported in 5 of the RCTs. Our results did not indicate reduction of physical disability with TFE steroids compared with TFE LA or saline at 1 to 3 months after the interventions. There may be several reasons for this finding. First, ODI may be better at detecting change in the more seriously disabled patients, whereas the RMDQ may be more appropriate for patients with minor disability.49 Second, TFE steroids may have less impact on LBP than lumbosacral radicular pain and both ODI and RMDQ reflect predominantly LBP. Finally, none of the RCTs involved exposure of participants to a well-planned regime of physiotherapy focused on improving core muscle, low back, and lower limb strength. It is also possible that the beneficial effects of TFE steroids on reducing physical disability may require longer than 3 months to manifest.
Efficacy of TFE Steroids in Reducing Incidence of Surgery in Patients with Lumbosacral Radicular Pain Secondary to Intervertebral Disk Herniation
Meta-analysis of the RCTs in this review did not find a difference in the probability of operative intervention in patients who had received TFE steroids (228 patients) compared with those who had received LA/saline (266 patients) (relative risk: 0.88; 95% CI: 0.55 to 1.41; P = 0.59, I2 = 53%). The time period for recording operative intervention varied from 3 to 28 months, and only 1 of the 6 RCTs showed a significant reduction in operative intervention in patients who received TFE steroids compared with those who received LA at a median interval of 23 months after the interventions.32 The incidence of surgery, however, in both groups (especially in the LA group) in this RCT was high (25% in steroid group and 66% in LA group), which suggests that longer follow-ups may be required to ascertain the impact of TFE steroids on the incidence of surgery, and participant populations/subgroups with a greater likelihood of surgery may benefit from TFE steroids. A recent SR-MA on the effect of epidural steroid injections (midline interlaminar or transforaminal) versus control (LA or saline) injections for “spinal” pain found a trend of reduction in the incidence of surgery up to 1 year after the injection but not beyond.50
Considering that the trial by Riew et al.32 was conducted in 2000, another explanation for their results being significantly different from those of the other 5 RCTs may be that the indications for operative intervention in patients with lumbosacral radicular pain secondary to herniated intervertebral disks have evolved and become more restrictive over the years. Both these hypothesis are supported by a recent review of operative and nonoperative treatments for patients with herniated intervertebral disks and symptomatic radiculopathy at 8 years from the Spine Patient Outcomes Research Trial.51 Subgroup analyses in this review identified patients with sequestered fragments on magnetic resonance imaging, greater levels of baseline back pain accompanying radiculopathy, a longer duration of symptoms, and those who were neither working nor disabled at baseline with a greater relative advantage from surgery at 8 years.
Adverse Effects Associated with Administering TFE Steroids in Patients with Lumbosacral Radicular Pain Secondary to Intervertebral Disk Herniation
Although adverse effects were reported anecdotally by only 2 RCTs in our review, we came across several complications related to either administration of steroids into the epidural space during our literature search (Table 5). Hyperglycemia, myopathy, and neurologic injury (direct injury to nerve or spinal cord by the needle or injury to vascular supply of the spinal nerves) have been reported anecdotally after the use of both particulate and nonparticulate steroids for TFE injections.52–56 RCTs on TFE steroids or other agents for treating lumbosacral radicular pain secondary to herniated intervertebral disks in future should include monitoring of adverse effects defined a priori as an outcome.
Review of Literature on TFE Steroids for Patients with Lumbosacral Radicular Pain Secondary to Intervertebral Disk Herniation
We identified 6 recent (published in the last 3 years) systematic reviews that had explored the role of steroids administered through different epidural approaches in providing analgesia and reducing physical disability and incidence of surgical intervention in patients with LBP and lumbosacral radicular pain secondary to herniated intervertebral disks. Five of these reviews included trials involving the use of steroids administered through interlaminar, caudal, and transforaminal approaches for patients with lumbosacral radicular pain,4,16,19,57,58 whereas the sixth included trials of only TFE steroids.20 None of these systematic reviews restricted their inclusion criteria to only patients with lumbosacral radicular pain secondary to herniated intervertebral disks because patients with spinal stenosis as an etiology of lumbosacral radicular pain were also included. In all 6 reviews, the authors concluded that epidural steroids confer a modest analgesic benefit that persists over the immediate-to-short (5 days to 3 months) but not the long term (Table 5). Two of these reviews had conflicting findings regarding the efficacy of epidural steroids on reduction of physical disability, with one review showing no benefit20 whereas the other reported benefit over the short term.19 Another systematic review and network meta-analysis of 122 studies on management strategies for sciatica reported that compared with inactive control or conventional care, there was a statistically significant improvement after epidural steroid injections in pain and overall recovery.59
Limitations of Current Evidence
There remain several unanswered questions regarding efficacy and safety of TFE steroids in patients with lumbosacral radicular pain secondary to herniated intervertebral disks. This is because of the following limitations of this SR-MA:
- There was variation in doses of steroids, volumes of injectate, and use of additives. The comparators also varied in types, concentrations, and dose across RCTs.
- Analgesia was assessed over a wide range of time periods after the interventions.
- Validated, but 2 different, scales were used to assess physical disability at varying time periods after the interventions.
- The effect of other multimodal therapies (e.g., pharmacologic, physical, or psychological therapy) on the analgesic efficacy of transformational steroids was not assessed.
- The effect of TFE steroids on psychological disability and on quality of life was not assessed.
- Incidence of surgical intervention was assessed over varying periods of time, and criteria for surgery were not standardized in the RCTs.
- Lack of measurement of biomarkers (e.g., perineural cytokine levels, gene expression, functional/perfusion of brain/spinal cord on magnetic resonance imaging) before and after the intervention in the existing RCTs makes it difficult to understand the mechanistic basis of action of TFE steroids.
- Assessment of procedure and steroid-related adverse effects was not a secondary outcome in any of the RCTs.
- The RCTs included in this review had heterogeneity and variable quality.
- The number of trials for subgroup analysis (low versus high dose of methylprednisolone) to explore heterogeneity in our meta-analysis was low (4 and 3 trials, respectively). The DerSimonian and Laird approach with Z-statistic used for random effects meta-analysis can result in overly narrow CIs, and alternative approaches for analysis have been suggested.60 Consequently, our CI results for meta-analysis of data from 5 or less trials should be interpreted with caution because they may be too narrow.
SUGGESTIONS FOR FUTURE RESEARCH
Future RCTs should use equipotent doses of steroids with standardized volumes and additives (LA or saline), and the comparators also should be standardized. Impact of other multimodal therapies used for patients (e.g., pharmacologic, physical, or psychological therapy) on the analgesic efficacy of transformational steroids also should be assessed. For example, a recent randomized, comparative-effectiveness study comparing a series of epidural steroid injections, medications and physical therapy, and combination therapy for cervical radiculopathy found that a multimodal treatment approach was more effective than stand-alone therapy.61 Important “core” outcomes of anxiety and depression in addition to satisfaction with treatment also should be included as end points. The primary outcome should be measured at a range of intervals after the intervention.
It is important to perform large, multicenter RCTs that evaluate the efficacy of this intervention. Given our finding that combining steroids with LA may increase efficacy, that results from other studies that suggest LA on their own may confer long-term analgesic benefits,50,62 and the fact that the placebo analgesic effect is strong,63 there is a need to perform an RCT with 4 groups of perineural injectates: no/sham injection, LA, steroid, and a combination of steroid with LA, respectively. We propose these 4 groups because it is important to understand the mechanistic basis of potential analgesic effects from TFE steroid injections—whether it is an LA (membrane-stabilizing effect) or a mechanical effect (washout of inflammatory mediators and breakdown of scar tissue) or a combination of both. An RCT with 4 arms, as suggested here, may yield insights into these mechanisms. The sample size for this trial would be at least 71 patients per group (i.e., 284 patients in all) for an expected difference between pain NRS (0–10 scale) means of 1.8 and the within-group SD of 3. This sample size has been determined to give a 90% chance of rejecting the null hypothesis of no difference between means at an α of 0.0125 and using a Bonferroni adjustment to the size of the test to compensate for multiple comparisons.64 Standardized supportive management (pharmacologic, physical, and psychological supportive therapy) for lumbosacral radicular pain should be defined and provided to all groups. Pain scores should be measured at multiple time intervals after intervention: acutely (at 1 week), intermediate (1 week to 3 months), and long-term (3–12 months). Important secondary outcomes that need to be measured in this trial include psychological, functional, and patient satisfaction with treatment along with meticulous attempts to identify and record adverse effects.65 Duration of monitoring (after TFE injection) for surgical intervention (and the criteria thereof) should be clearly specified. Another potential area of research is the comparison of efficacy and safety of particulate (methylprednisolone, triamcinolone) and nonparticulate (dexamethasone, betamethasone) steroids because nonparticulate steroids may have a better safety profile.52 Finally, repeat procedures should be evaluated in RCTs with comparisons of different regimens in terms of number of procedures and intervals between procedures. Measurement of perineural (epidural) cytokines should be considered in these RCTs because repeat procedures provide an opportunity to study trends in levels of perineural cytokines over time (before and after intervention) without exposing patients to additional procedures. However, measurement of epidural cytokines is likely to be expensive and may lack correlation with presentation and/or resolution of symptoms of lumbosacral radicular pain.
Epidural injections and steroids can be associated with multiple local and systemic effects (epidural abscess, nerve or spinal cord injury causing neurologic deficits, myopathy, epidural lipomatosis, and hyperglycemia).66–72 Only 2 trials in this review reported procedural adverse effects of perineural steroids.33,38 There is a need to monitor patients for predefined adverse effects in RCTs evaluating injections of steroids. Hyperglycemia and osteoporosis at multiple time points after interventions are examples of 2 effects that could be relatively easy to identify. Fluid retention, hypertension, infections, myopathy, hypokalemia, cataracts, and psychosis are other adverse effects of steroids that need to be assessed. Procedural complications to be assessed should include number of attempts, short-term (paresthesias, dural puncture) and long-term (neurologic deficits, infections) associated with the injections.
Name: Anuj Bhatia, MBBS, MD, FRCA, FRCPC, FIPP, FFPMRCA, EDRA, CIPS.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Anuj Bhatia approved the final manuscript.
Name: David Flamer, MD, FRCA.
Contribution: This author helped conduct the study, analyze the data, and write the manuscript.
Attestation: David Flamer approved the final manuscript.
Name: Prakesh S. Shah, MSc, MBBS, MD, DCH, MRCP, FRCPC.
Contribution: This author helped design the study, analyze the data, and write the manuscript.
Attestation: Prakesh S. Shah approved the final manuscript.
Name: Steven P. Cohen, MD.
Contribution: This author helped design the study, analyze the data, and write the manuscript.
Attestation: Steven P. Cohen approved the final manuscript.
This manuscript was handled by: Spencer S. Liu, MD.
We thank Ms. Marina Engelasakis, Librarian, University Health Network, for her assistance in designing the literature search strategy.
1. Crow WT, Willis DR. Estimating cost of care for patients with acute low back pain: a retrospective review of patient records. J Am Osteopath Assoc. 2009;109:229–33
2. Frymoyer JW. Back pain and sciatica. N Engl J Med. 1988;318:291–300
3. Juniper M, Le TK, Mladsi D. The epidemiology, economic burden, and pharmacological treatment of chronic low back pain in France, Germany, Italy, Spain and the UK: a literature-based review. Expert Opin Pharmacother. 2009;10:2581–92
4. Cohen SP, Bicket MC, Jamison D, Wilkinson I, Rathmell JP. Epidural steroids: a comprehensive, evidence-based review. Reg Anesth Pain Med. 2013;38:175–200
5. Ropper AH, Zafonte RD. Sciatica. N Engl J Med. 2015;372:1240–8
6. Tarulli AW, Raynor EM. Lumbosacral radiculopathy. Neurol Clin. 2007;25:387–405
7. Olmarker K, Størkson R, Berge OG. Pathogenesis of sciatic pain: a study of spontaneous behavior in rats exposed to experimental disc herniation. Spine (Phila Pa 1976). 2002;27:1312–7
8. Bogduk N Clinical Anatomy of the Lumbar Spine and Sacrum. 20054th ed Amsterdam, Netherlands Elsevier:183–6
9. Bogduk N. On the definitions and physiology of back pain, referred pain, and radicular pain. Pain. 2009;147:17–9
10. Brisby H. Pathology and possible mechanisms of nervous system response to disc degeneration. J Bone Joint Surg Am. 2006;88(suppl 2):68–71
11. Heliövaara M, Knekt P, Aromaa A. Incidence and risk factors of herniated lumbar intervertebral disc or sciatica leading to hospitalization. J Chronic Dis. 1987;40:251–8
12. Weber H. The natural history of disc herniation and the influence of intervention. Spine (Phila Pa 1976). 1994;19:2234–8
13. Weinstein JN, Lurie JD, Tosteson TD, Tosteson AN, Blood EA, Abdu WA, Herkowitz H, Hilibrand A, Albert T, Fischgrund J. Surgical versus nonoperative treatment for lumbar disc herniation: four-year results for the Spine Patient Outcomes Research Trial (SPORT). Spine (Phila Pa 1976). 2008;33:2789–800
14. Peul WC, van Houwelingen HC, van den Hout WB, Brand R, Eekhof JA, Tans JT, Thomeer RT, Koes BWLeiden-The Hague Spine Intervention Prognostic Study Group. . Surgery versus prolonged conservative treatment for sciatica. N Engl J Med. 2007;356:2245–56
15. Koes BW, Scholten RJ, Mens JM, Bouter LM. Efficacy of epidural steroid injections for low-back pain and sciatica: a systematic review of randomized clinical trials. Pain. 1995;63:279–88
16. Shamliyan TA, Staal JB, Goldmann D, Sands-Lincoln M. Epidural steroid injections for radicular lumbosacral pain: a systematic review. Phys Med Rehabil Clin N Am. 2014;25:471–89.e1
17. Weinstein SM, Herring SA, Derby R. Contemporary concepts in spine care. Epidural steroid injections. Spine (Phila Pa 1976). 1995;20:1842–6
18. Bogduk N. Epidural steroids. Spine (Phila Pa 1976). 1995;20:845–8
19. 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;157:865–77
20. Quraishi NA. Transforaminal injection of corticosteroids for lumbar radiculopathy: systematic review and meta-analysis. Eur Spine J. 2012;21:214–9
21. Friedly JL, Comstock BA, Turner JA, Heagerty PJ, Deyo RA, Sullivan SD, Bauer Z, Bresnahan BW, Avins AL, Nedeljkovic SS, Nerenz DR, Standaert C, Kessler L, Akuthota V, Annaswamy T, Chen A, Diehn F, Firtch W, Gerges FJ, Gilligan C, Goldberg H, Kennedy DJ, Mandel S, Tyburski M, Sanders W, Sibell D, Smuck M, Wasan A, Won L, Jarvik JG. A randomized trial of epidural glucocorticoid injections for spinal stenosis. N Engl J Med. 2014;371:11–21
22. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, Schünemann HJGRADE Working Group. . GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336:924–6
23. Higgins JPT, Green S Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. 2011 The Cochrane Collaboration Available at: www.cochrane-handbook.org
. Accessed March 1, 2015
24. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ. 2009;339:b2700
25. Lefebvre C, Manheimer E, Glanville JHiggins JPT, Green S. Chapter 6.4: searching for studies. In: Cochrane Handbook for Systematic Reviews of Interventions, Version 5.0.1.. 2008 The Cochrane Collaboration Available at: www.cochrane-handbook.org
. Accessed March 1, 2015
26. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50:1088–101
27. Higgins JP, White IR, Wood AM. Imputation methods for missing outcome data in meta-analysis of clinical trials. Clin Trials. 2008;5:225–39
28. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–88
29. Bhatia A, Flamer D, Shah PS. Perineural steroids for trauma and compression-related peripheral neuropathic pain: a systematic review and meta-analysis. Can J Anaesth. 2015;62:650–62
30. Andrews JC, Schünemann HJ, Oxman AD, Pottie K, Meerpohl JJ, Coello PA, Rind D, Montori VM, Brito JP, Norris S, Elbarbary M, Post P, Nasser M, Shukla V, Jaeschke R, Brozek J, Djulbegovic B, Guyatt G. GRADE guidelines: 15. Going from evidence to recommendation-determinants of a recommendation’s direction and strength. J Clin Epidemiol. 2013;66:726–35
31. Guyatt GH, Oxman AD, Kunz R, Brozek J, Alonso-Coello P, Rind D, Devereaux PJ, Montori VM, Freyschuss B, Vist G, Jaeschke R, Williams JW Jr, Murad MH, Sinclair D, Falck-Ytter Y, Meerpohl J, Whittington C, Thorlund K, Andrews J, Schünemann HJ. GRADE guidelines 6. Rating the quality of evidence–imprecision. J Clin Epidemiol. 2011;64:1283–93
32. Riew KD, Yin Y, Gilula L, Bridwell KH, Lenke LG, Lauryssen C, Goette K. The 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 Am. 2000;82-A:1589–93
33. Karppinen J, Malmivaara A, Kurunlahti M, Kyllönen E, Pienimäki T, Nieminen P, Ohinmaa A, Tervonen O, Vanharanta H. Periradicular infiltration for sciatica: a randomized controlled trial. Spine (Phila Pa 1976). 2001;26:1059–67
34. Vad VB, Bhat AL, Lutz GE, Cammisa F. Transforaminal epidural steroid injections in lumbosacral radiculopathy: a prospective randomized study. Spine (Phila Pa 1976). 2002;27:11–6
35. Ng L, Chaudhary N, Sell P. The efficacy of corticosteroids in periradicular infiltration for chronic radicular pain: a randomized, double-blind, controlled trial. Spine (Phila Pa 1976). 2005;30:857–62
36. Tafazal S, Ng L, Chaudhary N, Sell P. Corticosteroids in peri-radicular infiltration for radicular pain: a randomised double blind controlled trial. One year results and subgroup analysis. Eur Spine J. 2009;18:1220–5
37. Ghahreman A, Ferch R, Bogduk N. The efficacy of transforaminal injection of steroids for the treatment of lumbar radicular pain. Pain Med. 2010;11:1149–68
38. Cohen SP, White RL, Kurihara C, Larkin TM, Chang A, Griffith SR, Gilligan C, Larkin R, Morlando B, Pasquina PF, Yaksh TL, Nguyen C. Epidural steroids, etanercept, or saline in subacute sciatica: a multicenter, randomized trial. Ann Intern Med. 2012;156:551–9
39. Manchikanti L, Cash KA, Pampati V, Falco FJ. Transforaminal epidural injections in chronic lumbar disc herniation: a randomized, double-blind, active-control trial. Pain Physician. 2014;17:E489–501
40. Mager DE, Lin SX, Blum RA, Lates CD, Jusko WJ. Dose equivalency evaluation of major corticosteroids: pharmacokinetics and cell trafficking and cortisol dynamics. J Clin Pharmacol. 2003;43:1216–27
41. Lennard TA Pain Procedures in Clinical Practice. 2011 Philadelphia, PA Elsevier Saunders:7
42. Harris RP, Helfand M, Woolf SH, Lohr KN, Mulrow CD, Teutsch SM, Atkins DMethods Work Group, Third US Preventive Services Task Force. . Current methods of the US Preventive Services Task Force: a review of the process. Am J Prev Med. 2001;20:21–35
43. Roland M, Fairbank J. The Roland-Morris Disability Questionnaire and the Oswestry Disability Questionnaire. Spine (Phila Pa 1976). 2000;25:3115–24
44. Sterne JAC, Egger M, Moher DHiggins JPT, Green S. Chapter 10: addressing reporting biases. In: Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0.. 2011 The Cochrane Collaboration Available at: http://www.cochrane.org/sites/default/files/Handbook510pdf_Ch10_ReportingBias.pdf
. Accessed March 28, 2015
45. Whynes DK, McCahon RA, Ravenscroft A, Hardman J. Cost effectiveness of epidural steroid injections to manage chronic lower back pain. BMC Anesthesiol. 2012;12:26
46. Owlia MB, Salimzadeh A, Alishiri G, Haghighi A. Comparison of two doses of corticosteroid in epidural steroid injection for lumbar radicular pain. Singapore Med J. 2007;48:241–5
47. Kang SS, Hwang BM, Son HJ, Cheong IY, Lee SJ, Lee SH, Chung TY. The dosages of corticosteroid in transforaminal epidural steroid injections for lumbar radicular pain due to a herniated disc. Pain Physician. 2011;14:361–70
48. Ahadian FM, McGreevy K, Schulteis G. Lumbar transforaminal epidural dexamethasone: a prospective, randomized, double-blind, dose-response trial. Reg Anesth Pain Med. 2011;36:572–8
49. Fairbank JC, Pynsent PB. The Oswestry Disability Index. Spine (Phila Pa 1976). 2000;25:2940–52
50. Bicket MC, Gupta A, Brown CH IV, Cohen SP. Epidural injections for spinal pain: a systematic review and meta-analysis evaluating the “control” injections in randomized controlled trials. Anesthesiology. 2013;119:907–31
51. Kerr D, Zhao W, Lurie JD. What are long-term predictors of outcomes for lumbar disc herniation? A randomized and observational study. Clin Orthop Relat Res. 2015;473:1920–30
52. Kennedy DJ, Plastaras C, Casey E, Visco CJ, Rittenberg JD, Conrad B, Sigler J, Dreyfuss P. Comparative 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 Med. 2014;15:548–55
53. Bicket MC, Chakravarthy K, Chang D, Cohen SP. Epidural steroid injections: an updated review on recent trends in safety and complications. Pain Manag. 2015;5:129–46
54. Rathmell JP, Benzon HT, Dreyfuss P, Huntoon M, Wallace M, Baker R, Riew KD, Rosenquist RW, Aprill C, Rost NS, Buvanendran A, Kreiner DS, Bogduk N, Fourney DR, Fraifeld E, Horn S, Stone J, Vorenkamp K, Lawler G, Summers J, Kloth D, O’Brien D Jr, Tutton S. Safeguards to prevent neurologic complications after epidural steroid injections: consensus opinions from a multidisciplinary working group and national organizations. Anesthesiology. 2015;122:974–84
55. Ward A, Watson J, Wood P, Dunne C, Kerr D. Glucocorticoid epidural for sciatica: metabolic and endocrine sequelae. Rheumatology (Oxford). 2002;41:68–71
56. Boonen S, Van Distel G, Westhovens R, Dequeker J. Steroid myopathy induced by epidural triamcinolone injection. Br J Rheumatol. 1995;34:385–6
57. Manchikanti L, Buenaventura RM, Manchikanti KN, Ruan X, Gupta S, Smith HS, Christo PJ, Ward SP. Effectiveness of therapeutic lumbar transforaminal epidural steroid injections in managing lumbar spinal pain. Pain Physician. 2012;15:E199–245
58. 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;163:373–81
59. Lewis RA, Williams NH, Sutton AJ, Burton K, Din NU, Matar HE, Hendry M, Phillips CJ, Nafees S, Fitzsimmons D, Rickard I, Wilkinson C. Comparative clinical effectiveness of management strategies for sciatica: systematic review and network meta-analyses. Spine J. 2015;15:1461–77
60. IntHout J, Ioannidis JP, Borm GF. The Hartung-Knapp-Sidik-Jonkman method for random effects meta-analysis is straightforward and considerably outperforms the standard DerSimonian-Laird method. BMC Med Res Methodol. 2014;14:25
61. Cohen SP, Hayek S, Semenov Y, Pasquina PF, White RL, Veizi E, Huang JH, Kurihara C, Zhao Z, Guthmiller KB, Griffith SR, Verdun AV, Giampetro DM, Vorobeychik Y. Epidural steroid injections, conservative treatment, or combination treatment for cervical radicular pain: a multicenter, randomized, comparative-effectiveness study. Anesthesiology. 2014;121:1045–55
62. Andreae MH, Andreae DA. Local anaesthetics and regional anaesthesia for preventing chronic pain after surgery. Cochrane Database Syst Rev. 2012;10:CD007105
63. Farrar JT, Young JP Jr, LaMoreaux L, Werth JL, Poole RM. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain. 2001;94:149–58
65. Dworkin RH, Turk DC, Farrar JT, Haythornthwaite JA, Jensen MP, Katz NP, Kerns RD, Stucki G, Allen RR, Bellamy N, Carr DB, Chandler J, Cowan P, Dionne R, Galer BS, Hertz S, Jadad AR, Kramer LD, Manning DC, Martin S, McCormick CG, McDermott MP, McGrath P, Quessy S, Rappaport BA, Robbins W, Robinson JP, Rothman M, Royal MA, Simon L, Stauffer JW, Stein W, Tollett J, Wernicke J, Witter JIMMPACT. . Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain. 2005;113:9–19
66. Habib GS. Systemic effects of intra-articular corticosteroids. Clin Rheumatol. 2009;28:749–56
67. Kabbara A, Rosenberg SK, Untal C. Methicillin-resistant Staphylococcus aureus epidural abscess after transforaminal epidural steroid injection. Pain Physician. 2004;7:269–72
68. Houten JK, Errico TJ. Paraplegia after lumbosacral nerve root block: report of three cases. Spine J. 2002;2:70–5
69. Martin DP, Huntoon MA. Spinal cord infarction following therapeutic computed tomography-guided left L2 nerve root injection. Spine (Phila Pa 1976). 2005;30:1558
70. Huntoon MA, Martin DP. Paralysis after transforaminal epidural injection and previous spinal surgery. Reg Anesth Pain Med. 2004;29:494–5
71. Kennedy DJ, Dreyfuss P, Aprill CN, Bogduk N. Paraplegia following image-guided transforaminal lumbar spine epidural steroid injection: two case reports. Pain Med. 2009;10:1389–94
72. Sandberg DI, Lavyne MH. Symptomatic spinal epidural lipomatosis after local epidural corticosteroid injections: case report. Neurosurgery. 1999;45:162–5