The Basis for Recommending Repeating Epidural Steroid Injections for Radicular Low Back Pain: A Literature Review

Article Outline

• Abstract
• Methods
• Results
• Discussion
  • Study Limitations
• Conclusions
• Appendix 1. Classification of Evidence
• References
• Copyright

Abstract 

Novak S, Nemeth WC. The basis for recommending repeating epidural steroid injections for radicular low back pain: a literature review.

Objectives

To determine the current evidence to support guidelines for frequency and timing of epidural steroid injections (ESIs), to help determine what sort of response should occur to repeat an injection, and to outline specific research needs in these areas.

Data Sources

A PubMed, Medline (EBSCO), and Cochrane library search (January 1971–December 2005), as well as additional references found from the initial search.

Study Selection

There were no studies that specifically addressed the objectives outlined. Eleven randomized controlled trials, 1 prospective controlled trial, and 2 prospective cohort studies were identified that included a protocol involving repeat epidural injections for radicular pain secondary to herniated nucleus pulposus or spinal stenosis. One qualitative survey was also identified. Five review articles were also included that discussed this topic.

Data Extraction

Data were extracted from clinical trials if they included the following: (1) protocols in clinical trials on ESIs that included repeat injections and the response required to trigger these injections, (2) any evidence given for establishing these protocols, and (3) similar studies that included only 1 injection. Specific mention of repeat ESIs and partial response that was mentioned in review articles was also included.

Data Synthesis

There is limited evidence to suggest guidelines for frequency and timing of ESIs or to help to define what constitutes the appropriate partial response to trigger a repeat injection. No study has specifically evaluated these objectives. Methodologically limited research suggests that repeat injections may improve outcomes, but the evidence is insufficient to make any conclusions.

Conclusions

There does not appear to be any evidence to support the current common practice of a series of injections. Recommendations for further research are made, including a possible study design.

Key Words: Injections, epidural, Rehabilitation, Review article [publication type]

EPIDURAL STEROID INJECTIONS (ESIs) have been suggested for treatment of radiculopathy in patients with lumbar disk disease to address symptoms secondary to mechanical irritation and inflammation. In 2005, McLain et al¹ described what they considered the most commonly used time protocol for ESIs. This included repeating this intervention at 2-week intervals if partial improvement in radicular signs and symptoms were observed. Repeat injections were not generally indicated if there was either no relief or full relief. The suggested maximum number of injections is usually 3,² although large ranges in frequency of ESIs have been reported. Cluff et al³ found in academic institutions that the mean maximal number of ESIs performed in 1 patient each year was 4.74±2.6 and in private practice was 6.9±6.98.

It has been stated that the decisions regarding selection and timing of ESIs are largely based on the following: (1) personal experiences and preferences of the treating physician, (2) anecdotal accounts, and/or (3) selected reviews of the literature.², ³, ⁴ There is currently no consensus as to the minimum or maximum number of ESIs or ideal timing for repeat injections.⁵ There is also little guidance as to how to define the type of response that should occur before a repeat injection. This is despite the fact that several researchers⁶, ⁷, ⁸ have suggested that studies addressing specifics of ESIs require research, including number of injections and temporal spacing.

The objectives of this review were to determine the following for patients treated with epidural steroids for radiculopathy from either a herniated nucleus pulposus (HNP) or spinal stenosis: (1) the evidence to suggest guidelines for frequency and timing of ESIs and (2) the evidence to help determine what sort of partial response should occur to repeat an injection. These objectives specifically involved the effect of repeat injections and/or the timing of these injections on the success rate of ESI therapy. A final objective was to outline the specific research needs in these areas.

Back to Article Outline

Methods 

A search of current literature found that there was no published research that specifically addressed the above objectives. We then decided to perform a PubMed, Medline (EBSCO), and Cochrane Library search (January 1971–December 2005) to identify clinical trials and review articles involving the role of epidural steroids for the treatment of sciatica and/or radiculopathy in the lumbar region secondary to HNP and/or spinal stenosis. Articles were not included if they specifically addressed the use of these injections for treatment of axial low-back pain (LBP), treatment for failed back surgery, or treatment in the cervical region. Search words included the following free-text terms: epidural, epidural corticosteroid, caudal, caudal corticosteroid, lumbosacral radiculopathy, and sciatica. Additional articles were identified for references identified in the original studies.

Inclusion criteria for the clinical trials were identified that investigated the following: (1) protocols in articles researching ESIs for treatment of radiculopathy that involved more than 1 injection, (2) any evidence given for establishing these protocols, and (3) studies that involved only 1 injection that used similar protocols to the earlier-identified multiple injection studies. These later articles were identified to both compare outcomes and to add to the rationale given for protocols. The criteria for selection gave preference to prospective, controlled randomized studies. Nonrandomized or retrospective study designs were included if they had a focus on repeat injections and, in particular, if they were frequently cited in this regard. Although it was impossible to grade the articles as they applied to the outlined objectives, each article was given a class of evidence score based on the criteria outlined in appendix 1. This was performed to give readers a general understanding of the methodologic quality of the study design of the described articles.

Table 1 was constructed to include the following: (1) documentation of the study design including a grading designation; (2) number of subjects, treatment, and inclusion and exclusion criteria, including duration of symptoms and fluoroscopic documentation of diagnosis when given; (3) outcome measures; (4) results; (5) number of ESIs and time to repeat; (6) basis for repeat injections; and (7) comments by researchers. Table 2 discusses current recommendations for repeating ESIs. This table does not include a grading system.

Table 1. Literature Regarding Repeating ESIs (1972–2005)

Abbreviations: avg, average; CT, computed tomography; IM, intramuscular; MRI, magnetic resonance imaging; NA, not applicable; ODQ, Oswestry Disability Questionnaire; RMDQ, Roland-Morris Disability Questionnaire; SF-36, Medical Outcomes Study 36-Item Short-Form Health Survey; VAS, visual analog scale.

Table 2. Review Articles on Repeating ESIs

Abbreviations: AHCPR, Agency for Health Care and Policy Research; SNRBs, selective nerve root blocks.

Back to Article Outline

Results 

A total of 91 articles were originally identified, and 21 additional articles were found using the references from the original articles. Based on the inclusion criteria identified earlier, 11 randomized controlled trials (RCTs), 1 prospective controlled trial, and 2 prospective cohort studies were identified that included a protocol involving repeat epidural injections for radicular pain secondary to HNP or spinal stenosis. One qualitative survey was also identified. These articles are included in table 1; 5 review articles are included in table 2.

In 1973, Dilke et al⁹ (see table 1) used a blind interlaminar approach to evaluate ESI use on patients with unilateral sciatica due to lumbar disk disease. A second injection was suggested at a few days to a week after the first if there was a negative or partial response. A third injection was not thought to be helpful. Blind injections were again investigated by Cuckler et al¹⁰ in 1985 (see table 1). All patients had radiographic confirmation of nerve-root compression and were diagnosed with either HNP or spinal stenosis. Cuckler found that there was no significant difference either short-term or long-term when comparing patients who received blind lumbar interlaminar methylprednisolone with procaine versus patients who received normal saline and procaine (again through ESI). Repeat injections were performed at 24 hours, and Cuckler stated that based on the long-term follow-up there was no reason to repeat a second ESI in a patient who did not respond within 24 hours to either of the injections offered.

The general protocols through the 1980s were summarized by Benzon⁷ in 1986 in a review of ESIs for LBP and lumbosacral radiculopathy (see table 2). He recommended 1 injection followed by observation for response. No further injections were advised if there was complete relief of symptoms. Although he did state that it was not recommended to repeat an injection if there was no relief of symptoms, it was noted that other earlier researchers⁹, ¹¹ had suggested a repeat injection if there was no response to the first injection. What was not noted was that these were studies performed before a general introduction of fluoroscopic guidance for injections. Benzon⁷ also stated that there was probably no justification for more than 3 injections. Through the early 1990s prospective trial protocols included 2 to 3 injections at 1- to 2-week intervals (see table 1).¹², ¹³, ¹⁴

In 1997, Carette et al¹⁵ (see table 1) performed a prospective RCT on patients with sciatica due to HNP that compared blind interlaminar ESIs with either methylprednisolone and saline or saline alone using multiple outcomes including visual analog scale (VAS) scores and the Oswestry Disability Questionnaire (ODQ) for LBP. The mean number of injections required was 2.1 in both the experimental and control groups. Injections given were blind and were given at 3- and 6-week intervals. Significant differences were found at 3 weeks for the steroid group on finger-to-floor distance and sensory deficits. At 6 weeks, the only significant difference was a greater improvement in leg pain in the steroid group. There was no significant difference at 3 months between the 2 groups. The intervention did not reduce the need for surgery.

In 1998, Lutz et al¹⁶ (see table 1) in a prospective case series of patients with magnetic resonance imaging (MRI) documentation of HNP and a chief complaint of leg pain used a definition of a VAS score of 4 out of 10 or less to determine the need for repeat transforaminal ESIs that were fluoroscopically guided. Eighty percent of patients received 2 or fewer injections. These injections were given at 2-week intervals. Overall 75.4% of patients studied had a successful long-term outcome (>50% reduction of VAS score with an average follow-up period of 88wk).

In 1999, a treatment recommendation for repeating ESIs was made in what is one of the most commonly cited review articles (see table 2).² This review specifically addressed a blind intralaminar ESI approach using steroid and local anesthetic. Abram suggested that 1 injection be given and reevaluation should occur after “1 to several” weeks. Reinjection was not suggested if symptoms improved dramatically. If there was no improvement, the suggestion was made to reinject, because this might indicate a failure of the drug to reach the affected root. It was also suggested that a different approach might be taken. A third injection was suggested if there was further but incomplete relief. Abram did state that although these treatment recommendations represented a single opinion, they were based on an extensive review of the literature and were “shared by a substantial number of anesthesiologists.”^2(p1937)

In a study published around this same time, Cluff et al³ (see table 1), using survey research, found that the mean number of epidurals given in an academic setting per year was 4.74±2.6 (upper range, 20) and in private practice was 6.9±6.9 (upper range, 40). Cluff noted that there appeared to be no consensus on what constituted a “proper ESI,” including type and dose of steroids, volume of injectate, frequency of administration, approaches, methods of identifying the epidural space, or the utility of fluoroscopy.

Between 2000 and 2002, several trials evaluated the use of transforaminal fluoroscopically guided ESIs. Riew et al¹⁷ (see table 1), in an RCT of fluoroscopically guided transforaminal ESIs, allowed for as many as 4 injections, because this was what was consistent with the normal clinical routine. Patients included had radiographic evidence of nerve root compression secondary to HNP or spinal stenosis, and previous lumbar surgery was not an exclusion criterion. Of the 57 patients who received ESIs in both treatment groups (bupivacaine alone vs bupivacaine plus betamethasone), 19 subjects had more than 1 injection. This number was not broken down as to study cohort (in other words, these repeat injections were as likely to have occurred in the experimental group as the control group). Therefore, for the entire study population, on average 1.5 injections were given. The outcome of this study was prevention of progression to surgery in patients who had been considered surgical candidates and had requested operative treatment. Pain had to have lasted for at least 6 weeks, but the mean duration of symptoms was not given for either group. Riew collected information using a North American Spine Society outcome instrument that included information on whether baseline expectations had been met after treatment. There was no difference between study groups preinjection in terms of this instrument. Only 3 specific statistics were given as to the comparison of this instrument postinjection between the groups, one of which was that although there was a trend toward a decrease in neurologic symptoms from the baseline to the final follow-up evaluation in all patients who avoided operative treatment, this difference was not significant. Riew also stated that if only a measure of degree of pain relief was used as an outcome, several patients considered as outcome successes (based on not proceeding to surgery) would have been considered failures. The discussion states that some patients who avoided surgery actually had higher pain scores at the final follow-up than preinjection.

In 2001, Karppinen et al¹⁸ (see table 1) evaluated patients with unilateral pain in a dermatomal fashion in a study that included patients with evidence of a normal or disk bulge in 22% of the treatment group and 14% of the control group. Using a single injection comparing fluoroscopically guided transforaminal ESIs using either methylprednisolone and bupivacaine or normal saline, Karppinen found significant short-term effects for the steroid group for leg pain, straight-leg raising, lumbar flexion, and overall satisfaction at 2 weeks. This effect did not persist to 4 weeks, and at 3 and 6 months the saline group had a more favorable effect in terms of back pain. Karppinen believed that repeated injections could produce a more sustained effect, but the basis for this statement was not explained.

Vad et al¹⁹ (see table 1), in patients with HNP documented by MRI, reported an average of 1.7 injections (range, 1–3) over an average of 1.4 years in a prospective controlled trial that studied fluoroscopically guided transforaminal ESIs compared with trigger point injections. The patients were allowed to choose the treatment arm. The timing for the repeat injections was not given. Outcomes included the Roland-Morris Disability Questionnaire for LBP, and subject’s report of pain reduction of more than 50% at least 1 year posttreatment. Overall, 84% of the patients in the ESI group had a successful outcome and attained maximal improvement within 6 weeks of treatment. Maximal improvement was obtained 4 weeks after the final injections. This was compared with 48% of the trigger point patients, who reached a maximal improvement within 12 weeks of treatment. The difference in outcomes was statistically significant. Vad stated that fewer injections were required because of the fluoroscopic technique. Botwin et al,²⁰ in a prospective noncontrolled study of patients with spinal stenosis, reported an average of 1.9 injections in patients who had a mean duration of symptoms of 5 years and 6 months. The timing of the injections was not given. Over a 1-year follow-up, successful long-term outcome was found for 75% of the patients (>50% reduction of pain).

By 2005, McLain et al¹ (see table 2), as noted in the introduction, again reiterated the use of the time protocol suggested by Abram.² This included repeating the injection at 2-week intervals if there was partial improvement of radicular signs. The definition of partial improvement was not given. Insufficient relief of pain after a technically well-performed, fluoroscopically guided injection was thought to be sufficient reason to stop further injections. An alternative injection approach was a possible treatment option in this case versus considering the possibility of a different pain generator. Repeat injections for patients with no improvement or those with full relief of symptoms was not suggested.

In 2005, Arden et al²¹ (see table 1) evaluated patients with unilateral sciatica of between 1 and 18 months duration (defined as leg pain radiating below the knee, reduced straight-leg raising, and a positive sciatic nerve stretch test). A blind interlaminar ESI technique was performed in a prospective randomized controlled double-blinded trial that included a repeat injection at week 3 and week 6 of follow-up if there was less than 75% improvement in the ODQ score from the baseline visit. Three injections were thought to most closely mirror local practice. Arden came to the conclusion that repeating an ESI was not effective in terms of improvement of function nor was it cost effective, because at 6 weeks and for the remainder of the study, there was no significant difference found between groups on any measured outcome parameter.²² In the same year, Wilson-MacDonald et al²³ (see table 1) compared blind interlaminar ESIs with intramuscular injections of methylprednisolone and bupivacaine. Wilson-MacDonald found a significant difference in pain relief between the 2 groups that became apparent at 10 days. Repeat injections were given to both groups (16% of the ESI group). The control group was allowed to cross over to the treatment group for the second injections. There was no difference between the groups long term or for rates of subsequent surgery. There was no difference in outcome between patients with spinal stenosis and those with disk prolapse, although there were more patients with combined disk prolapse and stenosis in the control group (7% in the epidural group, 31% in the control group). This difference was not considered significant. Wilson-MacDonald believed that ESIs were a useful technique to reduce symptoms in acute stages of nerve root compression.

In 2005, Ng et al²⁴ (see table 1) performed a study that was very similar to the study by Riew et al¹⁷ in that Ng compared fluoroscopically guided transforaminal ESIs using bupivacaine and corticosteroid with bupivacaine, but they used 1 injection only (vs up to 4 in the Riew study). Patients had MRI-documented HNP or spinal stenosis. Patients in the bupivacaine-only arm had a mean duration of symptoms of 12 months; in the steroid/local arm, the duration was 16.9 months. At 3 months, Ng found no statistical difference in the groups on outcomes of changes in ODQ scores, changes in VAS, or changes in walking distance. In terms of level of satisfaction, 49% of the bupivacaine group and 45% of the bupivacaine and steroid group rated their outcomes as excellent or good at 3 months. The finding of only a modest reduction in VAS scores was attributed to the chronicity of symptoms found in both treatment groups. Ng was uncertain why corticosteroid did not provide additional benefit. Ng speculated that this might have been secondary to lack of delivery of the treatment agent (despite the confirmation with contrast agent). Ng made 2 specific points. The first was that the vast majority of patients with lumbosacral radiculopathy (as many as 90%) recovered because of the natural process of healing within approximately 1 year.²⁵ Ng also stated that multiple injections might have produced a more sustained effect, based on the trials by Lutz¹⁶ and Vad¹⁹ and colleagues.

In 2005, DePalma et al²⁶ (see table 2) recommended repeating 1 to 4 transforaminal ESIs before surgery for patients with radicular pain. This recommendation was based on the review by Riew,¹⁷ which allowed for up to 4 injections, and the specific results that 13 of the 19 patients who had repeat injections avoided surgery. DePalma did not note that this group of 19 patients in the Riew study included both patients in the experimental group (steroid plus local) and control group (local only). DePalma made the statement that “perhaps only after 4 TFESIs [transforaminal epidural steroid injections] have been performed should a patient be referred for surgical intervention to treat radicular pain or radiculopathy.”^26(p1481)

Back to Article Outline

Discussion 

This literature review shows that before the introduction of fluoroscopic guidance for ESIs there was commonly a recommendation for a second injection. Repeat injection for partial response was generally suggested, although there was little evidence of why it was thought a second injection might be helpful. There are many possibilities for why a repeat injection might be necessary, but none of them has been fully investigated. Many of the problems in performing this type of research stem from both a lack of general understanding of the underlying mechanism of radicular pain and a lack of understanding of how ESIs provide their effect. The underlying mechanism of glucocorticoid activity is not fully understood, and there remains no basis for repeat injections based on characteristics of this class of medication itself.¹, ²⁷ There has also been no evidence given that the repeat use of neural blockade based on local anesthetic is necessary to improve effectiveness of ESI delivery.

Failure of the drug to reach the affected root was posited as a rationale for treatment failure of blind injections and, therefore, a reason for recommending repeat treatment.² Several researchers have specifically stated that the efficacy of ESI injections before the introduction of fluoroscopic guidance might be limited by technical error, with the most critical errors being intrathecal or intravascular placement.¹, ²⁸, ²⁹, ³⁰, ³¹ Although Abram² did not directly suggest this possibility as an etiology of decreased effect of blind ESI treatment, he did recommend repeating injections at an adjacent level to the previously injected level or to use a caudal approach at the S1 level. Therefore, in terms of the probability that a repeat injection might be required because of the inadvertent incorrect placement of the needle while using a blind technique, the use of a repeat injection could possibly make reasonable sense if there was no response to the first injection.

Wilson-MacDonald et al²³ used repeat injections in a recent RCT of blind ESIs and found good short-term success. Conversely, the recent RCT by Arden et al²¹ tends to refute this suggestion, finding that there was no advantage to performing more than 1 injection because there was no further difference between the experimental group and the group that received interspinous ligament injections. Although it appears that there is a compelling rationale for repeating an ESI based on the possibility of incorrect needle placement secondary to blind technique, this practice continues to require further research.

After the introduction of fluoroscopic guidance for ESIs, it has been suggested that if an injection was shown to be accurate, repeat injections should not be required in patients who were nonresponders.¹, ²⁸, ³⁰ Most of the fluoroscopically guided transforaminal studies cited included repeat injections as part of their protocol.¹⁶, ¹⁷, ¹⁹, ²⁰ The reasoning behind the use of repeat injections was not given for any of these studies.

In terms of incorrect needle placement after introduction of fluoroscopic guidance, McLain et al¹ noted that there was no guarantee that steroids delivered dorsally would reach the ventral epidural space. Research³² published subsequent to this literature review period has examined the role of the inability of the injectate to reach the site of pathology; not in terms of actually missing the epidural region, but in terms of the flow of the injectate to the ventral space. Transforaminal injections were more likely to result in ventral spread of contrast (defined as dispersion between the dura and posterior longitudinal ligament) and were also significantly more likely to result in complete pain relief. On the other hand, repeat injections were found to be more effective using a caudal or translaminar approach, and this was speculated to be secondary to increased systemic uptake through the epidural veins in the posterior epidural space and blood vessels in the subarachnoid space. The characteristics of the injectate in terms of class of medication and volume might also play a role in success, but this has not been specifically investigated. Further research comparing the spread of injectate both on treatment success and on the necessity for repeat injections remains necessary.

Two commonly cited studies, by Lutz¹⁶ and Botwin²⁰ and colleagues, found long-term success using repeated fluoroscopically guided transforaminal ESIs (1.8 and 1.9 injections, respectively), but neither was a controlled study. In addition, the Botwin²⁰ trial, which only studied patients with spinal stenosis, did not give timing of the frequency of the injections, and, as reported, these repeat injections could have occurred at any point in the 1-year follow-up period. The study by Vad et al,¹⁹ again using fluoroscopically guided transforaminal injections, noted that 84% of ESI patients reported success with treatment using an average of 1.7 injections, but the timing of these injections over the 1.4 years of follow-up was not given, leaving no guidance as to timing of repeat injections.

The most commonly cited transforaminal ESI article is that by Riew et al,¹⁷ which compared methylprednisolone and bupivacaine with bupivacaine only and used an outcome of progression to surgical treatment. Up to 4 injections were allowed, and there was no breakdown given between study groups in terms of repeat injections. There was also a paucity of results that would allow for comparison with more common outcomes currently used in ESI studies.

In examining literature from the postreview period (review period, 1972–2005), we found that the Riew group was able to follow up with 21 of the 29 patients who avoided surgery at a 5-year follow-up.³³ They found there was no difference between the 2 study groups at 5 years. Because the original study showed a difference between operative rates between the groups, the discrepancy at the 5-year follow-up was thought to be due to the small number of patients in the study. Riew was confident in recommending lumbar nerve root blocks because the injections were “effective in relieving symptoms for long enough that patients in whom the pain would have resolved naturally were able to avoid surgery in the meantime.”^33(p1724) When examining this group in terms of repeat injections we noted that 9 of the 21 five-year follow-up patients had received multiple blocks (range of minimum interval between blocks, 6d–10.5mo). None of the patients received surgery. Again, there was no reported breakdown as to what group these patients were in, and there were no conclusions made as to repeat blocks.

Study Limitations 

The most pressing limitation to this study was the lack of specific research addressing our objectives. The grading system of the articles, in particular, was affected by this fact. In addition, the implication of all of the reported studies points to a continued need to evaluate the use of ESIs in general and particularly in terms of short-term efficacy for treatment of radicular pain. Another major limitation that stands in the way of research in this area is that until the underlying mechanism of radicular pain is better elucidated, the mechanism of action for the positive, generally short-term results of ESIs remains unclear.

Overall, current research continues to suggest that ESIs remain a treatment for radicular pain, with a limited duration of benefit. This treatment approach appears to improve pain while natural healing occurs.¹, ²³, ²⁵ There has been little evidence that this effect extends to improvement of function.³⁴ Repeat injections remain the norm using both a blind and fluoroscopically guided technique, but the rationale for repeating treatment has not been elucidated and research has not been provided to support this practice.

Back to Article Outline

Conclusions 

There is limited evidence to suggest guidelines for frequency and timing of ESIs, and additional RCTs are required for adequate determination of this goal. It is also impossible to suggest guidelines to define the partial response necessary to trigger a repeat ESI based on the current research. This is primarily because there has been no study to specifically evaluate these objectives. Methodologically limited research suggests that repeat injections may improve outcomes, but the evidence is limited. There does not appear to be any evidence to support the current common practice of a series of injections.

It would appear that an RCT remains necessary to evaluate the use of repeat epidurals and what sort of partial response should be a trigger of repeats. One possible recommendation for an appropriate trial would initially involve subjects with a confirmed single diagnosis (ie, HNP) with a set duration of symptomatology who were given a fluoroscopically guided transforaminal injection. Subgroups based on response (eg, 0%–19%, 20%–69%, 70%–100%) could then be randomly assigned to either an additional ESI or a control group receiving epidural saline, with the knowledge that this latter procedure itself might trigger a treatment response. Consistency between groups would be required in terms of timing and number of repeat injections as well as for any additional treatment delivered in addition to the injections. Long-term follow-up of at least 1 year would be desirable.

Back to Article Outline

Appendix 1. Classification of Evidence 

Although it was impossible to grade the articles as they applied to the outlined objectives, each article was given a class of evidence that was loosely based on a recent assessment of the use of epidural steroid injections to treat lumbosacral pain by the American Academy of Neurology³⁴ in addition to the criteria described by Koes et al.³⁰ This was performed primarily as a classification of methodologic quality.

Class I: There was evidence of a prospective, controlled clinical trial with a blinded outcome assessment for a study population that was appropriate for the objectives outlined and randomization was appropriate. There was also evidence of the following: (1) inclusion and exclusion criteria were given, (2) baseline demographics were identified and statistical adjustments made if necessary, (3) dropouts and crossover subjects were documented and statistical assessment was made for these occurrences, and (4) the outcomes were clearly identified, reported, and reproducible using validated scoring when available.

Class II: There was evidence of a prospective, controlled clinical trial with a blinded outcome assessment for a study population that was appropriate for the objectives outlined, randomization was appropriate, and 3 of the above 1–4 criteria were met.

OR

The study was not appropriately randomized but met criteria 1–4.

Class III: All other prospective controlled trials.

Class IV: All other retrospective controlled trials.

Class V: Uncontrolled studies, case series, case reports, or expert opinion.

Back to Article Outline

References 

1. McLain RF, Kapural L, Mekhail NA. Epidural steroid therapy for back and leg pain: mechanisms of action and efficacy. Spine J. 2005;5:191–201
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (297 KB)
   • CrossRef
2. Abram SE. Treatment of lumbosacral radiculopathy with epidural steroids. Anesthesiology. 1999;91:1937–1941
   • View In Article
   • MEDLINE
   • CrossRef
3. Cluff R, Mehio AK, Cohen SP, Chang Y, Sang CN, Stojanovic MP. The technical aspects of epidural steroid injections: a national survey. Anesth Analg. 2002;95:403–408
   • View In Article
   • MEDLINE
   • CrossRef
4. Abram SE. Factors that influence the decision to treat pain of spinal origin with epidural steroid injections. Reg Anesth Pain Med. 2001;26:2–4
   • View In Article
   • MEDLINE
   • CrossRef
5. Weinstein SM, Herring SA, Derby R. Contemporary concepts in spine care (Epidural steroid injections). Spine. 1995;20:1842–1846
   • View In Article
   • MEDLINE
   • CrossRef
6. Green PW, Burke AJ, Weiss CA, Langan P. The role of epidural cortisone injection in the treatment of diskogenic low back pain. Clin Orthop Relat Res. 1980;Nov-Dec(153):121–125
   • View In Article
7. Benzon HT. Epidural steroid injections for low back pain and lumbosacral radiculopathy. Pain. 1986;24:277–295
   • View In Article
   • Abstract
   • Full-Text PDF (1480 KB)
   • CrossRef
8. Fanciullo GJ, Hanscom B, Seville J, Ball PA, Rose RJ. An observational study of the frequency and pattern of use of epidural steroid injection in 25,479 patients with spinal and radicular pain. Reg Anesth Pain Med. 2001;26:5–11
   • View In Article
   • MEDLINE
   • CrossRef
9. Dilke TF, Burry HC, Grahame R. Extradural corticosteroid injection in management of lumbar nerve root compression. Br Med J. 1973;2:635–637
   • View In Article
   • MEDLINE
10. Cuckler JM, Bernini PA, Wiesel SW, Booth RE, Rothman RH, Pickens GT. The use of epidural steroids in the treatment of lumbar radicular pain (A prospective, randomized, double-blind study). J Bone Joint Surg Am. 1985;6:63–66
    • View In Article
11. Warr AC, Wilkinson JA, Burn JM, Langdon L. Chronic lumbosciatic syndrome treated by epidural injection and manipulation. Practitioner. 1972;209:53–59
    • View In Article
    • MEDLINE
12. Mathews JA, Mills SB, Jenkins VM, et al. Back pain and sciatica: controlled trials of manipulation, traction, sclerosant and epidural injections. Br J Rheumatol. 1987;26:416–423
    • View In Article
    • MEDLINE
13. Ridley MG, Kingsley GH, Gibson T, Grahame R. Outpatient lumbar epidural corticosteroid injection in the management of sciatica. Br J Rheumatol. 1988;27:295–299
    • View In Article
    • MEDLINE
14. Bush K, Hillier S. A controlled study of caudal epidural injections of triamcinolone plus procaine for the management of intractable sciatica. Spine. 1991;16:572–575
    • View In Article
    • MEDLINE
    • CrossRef
15. Carette S, Leclaire R, Marcoux S, et al. Epidural corticosteroid injections for sciatica due to herniated nucleus pulposus. N Engl J Med. 1997;336:1634–1640
    • View In Article
    • MEDLINE
    • CrossRef
16. Lutz GE, Vad VB, Wisneski RJ. Fluoroscopic transforaminal lumbar epidural steroids: an outcome study. Arch Phys Med Rehabil. 1998;79:1362–1366
    • View In Article
    • Abstract
    • Full-Text PDF (3308 KB)
    • CrossRef
17. Riew KD, Yin Y, Gilula L, et al. 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;82:1589–1593
    • View In Article
18. Karppinen J, Malmivaara A, Kurunlahti M, et al. Periradicular infiltration for sciatica: a randomized controlled trial. Spine. 2001;26:1059–1067
    • View In Article
    • MEDLINE
    • CrossRef
19. Vad VB, Bhat AL, Lutz GE, Cammisa F. Transforaminal epidural steroid injections in lumbosacral radiculopathy: a prospective randomized study. Spine. 2002;27:11–16
    • View In Article
    • CrossRef
20. Botwin KP, Gruber RD, Bouchlas CG, et al. Fluoroscopically guided lumbar transformational epidural steroid injections in degenerative lumbar stenosis: an outcome study. Am J Phys Med Rehabil. 2002;81:898–905
    • View In Article
    • MEDLINE
    • CrossRef
21. Arden NK, Price C, Reading I, et al. A multicentre randomized controlled trial of epidural corticosteroid injections for sciatica: the WEST study. Rheumatology (Oxford). 2005;44:1399–1406
    • View In Article
    • MEDLINE
    • CrossRef
22. Price C, Arden N, Coglan L, Rogers P. Cost-effectiveness and safety of epidural steroids in the management of sciatica. Health Technol Assess. 2005;9:1–58
    • View In Article
    • MEDLINE
23. Wilson-MacDonald J, Burt G, Griffin D, Glynn C. Epidural steroid injection for nerve root compression (A randomised, controlled trial). J Bone Joint Surg Br. 2005;87:352–355
    • View In Article
    • MEDLINE
    • CrossRef
24. Ng L, Chaudhary N, Sell P. The efficacy of corticosteroids in periradicular infiltration for chronic radicular pain: a randomized, double-blind, controlled trial. Spine. 2005;30:857–862
    • View In Article
    • CrossRef
25. Rydevik B, Brown MD, Lundborg G. Pathoanatomy and pathophysiology of nerve root compression. Spine. 1984;9:7–15
    • View In Article
    • MEDLINE
    • CrossRef
26. DePalma MJ, Bhargava A, Slipman CW. A critical appraisal of the evidence for selective nerve root injection in the treatment of lumbosacral radiculopathy. Arch Phys Med Rehabil. 2005;86:1477–1483
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (131 KB)
    • CrossRef
27. Manchikanti L. Role of neuraxial steroids in interventional pain management. Pain Physician. 2002;5:182–199
    • View In Article
    • MEDLINE
28. White AH, Derby R, Wynne G. Epidural injections for the diagnosis and treatment of low-back pain. Spine. 1980;5:78–86
    • View In Article
    • MEDLINE
    • CrossRef
29. Johnson BA, Schellhas KP, Pollei SR. Epidurography and therapeutic epidural injections: technical considerations and experience with 5334 cases. AJNR Am J Neuroradiol. 1999;20:697–705
    • View In Article
    • MEDLINE
30. 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;63:279–288
    • View In Article
    • Abstract
    • Full-Text PDF (784 KB)
    • CrossRef
31. Renfrew DL, Moore TE, Kathol MH, el-Khoury GY, Lemke JH, Walker CW. Correct placement of epidural steroid injections: fluoroscopic guidance and contrast administration. AJNR Am J Neuroradiol. 1991;12:1003–1007
    • View In Article
    • MEDLINE
32. Ackerman WE, Ahmad M. The efficacy of lumbar epidural steroid injections in patients with lumbar disc herniations. Anesth Analg. 2007;104:1217–1222
    • View In Article
    • CrossRef
33. Riew KD, Park JB, Cho YS, et al. Nerve root blocks in the treatment of lumbar radicular pain (A minimum 5-year follow-up). J Bone Joint Surg Am. 2006;88:1722–1725
    • View In Article
    • MEDLINE
    • CrossRef
34. Armon C, Argoff CE, Samuels J, Backonja MM Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Assessment: use of epidural steroid injections to treat radicular lumbosacral pain: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology. 2007;68:723–729
    • View In Article
    • CrossRef
35. Burn JM, Langdon L. Duration of action of epidural methyl prednisolone (A study in patients with the lumbosciatic syndrome). Am J Phys Med. 1974;53:29–34
    • View In Article
    • MEDLINE