The economic impact and growing prevalence of low back pain is substantial [1,2]. An assessment of the state of US health revealed that in 2010, 3 of the 5 disorders that contributed to the most years lived with a disability were related to chronic pain, including low back pain, other muscular disorders, and neck pain [2]. Consequently, determining the appropriate strategy for managing chronic low back pain, and the disability related to it, is of utmost importance. Options range from simple exercise instructions to complex fusions [1,3,4,5,6,7,8,9,10,11,12,13,14,15,16]. The rising costs of managing spinal pain have reached approximately $100 billion per year [1].

The intervertebral disc has two distinct but inter-related mechanisms that can cause pain. These can include compression of neural structures by a herniated disc, and pathologic changes that can occur within the disc, serving as a primary pain generator [17,18,19,20]. Low back pain without disc herniation or facet joint pain, described as discogenic pain, internal disc disruption, and painful degenerative disc disease, has been identified as the primary source of pain in multiple clinical studies published over the past several decades [6,18,19,20,21,22,23,24,25,26]. Pain and disability secondary to disc herniation has been described in only a small proportion of patients. Thus, discogenic pain may be a primary source of low back pain. Even though this remains one of the greatest health care crises, it has remained poorly defined and its diagnosis and treatment continue to be controversial [6,19,20,21,22,23,24,25,26]. Malik et al. [19] concluded that despite its extensive affirmation in the literature and enormous resources regularly devoted to it, currently discogenic pain lacks clear diagnostic criteria and uniform treatment or terminology. Bogduk et al. [21], in a state-of-the-art review of lumbar discogenic pain, concluded that all of the null hypotheses that have been raised against the concept of discogenic pain and its diagnosis have each been refuted by one or more studies.

Mirza et al. [27] described that patients suffering with discogenic pain may be attracted to an expanding range of costly diagnostic and therapeutic interventions. Phillips et al. [8] concluded that the body of literature supports fusion surgery as a viable treatment option for reducing pain and improving function in patients with chronic low back pain refractory to nonsurgical care when a diagnosis of disc degeneration can be made. Bydon et al. [9] concluded that despite the significant improvement in Oswestry Disability Index (ODI) scores in the lumbar fusion groups in 3 studies, pooled data revealed no significant difference when compared to the nonoperative groups.

Lu et al. [28], in a systematic review of nonoperative management of discogenic back pain, identified 11 RCTs investigating traction therapy, injections, and ablative techniques. The results revealed that there were few high quality studies evaluating nonoperative treatments for reducing discogenic low back pain; however, the results from 5 RCTs investigating methylene blue injection, steroid injection, ramus communicans, ablation, intradiscal electrothermal therapy, and biacuplasty favored intervention over sham therapy; however, these are emerging treatments. In contrast, epidural injections are a well established therapy. Saltychev et al. [12] also reached similar conclusions with a lack of strong evidence for lumbar fusion compared to conservative treatment. Deyo et al. [29] asked for restraint from resorting to fusion surgery. Other assessments have shown a lack of significant evidence supporting multiple intradiscal therapies [6,30].

The rationale for fusion is based on the premise that if the pain generator is emanating from the disc, eliminating the painful segmental motion should cure the problem. However, there are many problems with fusion surgery. First, it could lead to increased stress on adjacent levels leading to transitional joint pains. Second, the surgery itself leads to the destruction of healthy tissue. Third, it can be complicated by intraspinal scarring. Fourth, the instrumentation alone has been associated with the development of pain. Fifth, the outcomes of fusion for pain alone have been suboptimal. Consequently, total disc replacement has been developed, which shares some of the same potential complications. Fusion surgery and disc replacement have been increasing with a lack of consensus regarding the efficacy of lumbar spinal fusion for discogenic pain. Randomized studies showed the effectiveness of fusion with good to excellent pain relief in only 39% of the patients, whereas, one study found that only 63% of the patients with discogenic pain showed any improvement after surgery [29]. Further, successful surgical fusion, which is sometimes used as a hallmark of success, does not necessarily translate into significant pain reduction or functional status improvement [31]. A Cochrane review failed to find a clinically significant difference between lumbar disc arthroplasty and fusion surgery, even though lumbar arthroplasty offers a motion sparing alternative to fusion [32]. Further, there have not been any randomized trials comparing disc arthroplasty with nonoperative management. The rising costs of managing low back pain and the costs of lumbar fusion and arthroplasty without proven efficacy have invited scrutiny from payers [27,28]. In fact, a review commissioned by the Centers for Medicare and Medicaid Services (CMS) Coverage and Advisory Committee [33], conducted by the Washington Health Care Technology Assessment program [34], concluded that lumbar fusion for degenerative disc disease lacked sufficient evidence of efficacy and safety to justify unconditional coverage.

In contrast to fusion and arthroplasty, the role of epidural injections for axial discogenic pain has not been addressed with the same rigor, subjecting the therapy to a systematic review [6,35,36,37,38]. Considering the presence of numerous modalities of treatments with intradiscal therapies, apart from surgical fusion and arthroplasty, it may be worthwhile to consider epidural injections - a simple and common procedure frequently performed in managing low back and lower extremity pain with or without steroids [6,35,36,37,38,39,40,41,42,43]. Consequently, this systematic review is undertaken to determine the comparative efficacy of lumbar fusion and epidural injections utilizing all 3 anatomical approaches in the treatment of lumbar discogenic pain.

The methodology utilized in this systematic review followed the widely accepted review process derived from evidence-based systematic reviews and meta-analysis of randomized trials [44,45].

Only randomized controlled trials (RCTs) of epidural injections and fusion and/or disc arthroplasty were utilized, either placebo- or active-controlled. The trials were eligible if the assessment was performed for discogenic pain. The duration of symptoms of the study participants was chronic pain of more than 3 months. For this evaluation, the studies including disc herniation, radiculitis, central or foraminal stenosis, or post surgery syndrome were not included.

All trials providing appropriate management and with outcome evaluations of 3 months or longer, statistical evaluations, and at least 25 patients were reviewed.

The primary outcome measure was pain relief. The secondary outcome measure was functional status improvement.

A search for literature published from 1966 through October 2014 was performed utilizing data from PubMed, Cochrane library, the US National Guideline Clearinghouse (NGC), previous systematic reviews, and cross references.

The search strategy emphasized low back and lower extremity pain, discogenic pain, pain treated with either lumbar fusion, lumbar disc arthroplasty, caudal, lumbar interlaminar, or lumbar transforaminal epidural injections in the lumbosacral spine. Search terms included: (epidural injection) OR epidural steroid) OR epidural perineal injection) OR interlaminar epidural) OR intraarticular corticosteroid) OR nerve root blocks) OR intraarticular injection) OR periradicular infiltration) OR saline injection) OR transforaminal injection) OR corticosteroid) OR methyl prednisolone) OR (surgical [Title/Abstract]) OR surgery [Title/Abstract]) OR fusion [Title/Abstract]) AND (meta-analysis [pt] OR randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [mh] OR random allocation [mh] OR double-blind method [mh] OR single-blind method [mh] OR clinical trial [pt] OR clinical trials [mh] OR ("clinical trial" [tw]) OR (single [tw] OR double [tw] OR trebl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw]) OR (placebos [mh] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp]) NOT (animals [mh] NOT human [mh]) AND (Lumbar) AND ("back pain" [Title/Abstract]) OR spondylosis [Title/Abstract]) OR DDD [Title/Abstract]) OR "disc degeneration" [Title/Abstract]) OR "degenerative disk disease" [Title/Abstract]) OR "degenerative disc disease" [Title/Abstract])

The quality of each individual article used in this analysis was assessed by Cochrane review criteria for randomized trials as shown in Appendix 1 [44]. Only randomized trials meeting the inclusion criteria with at least 5 of 12 Cochrane criteria were utilized for analysis.

Meta-analysis was considered if more than 2 randomized trials were homogeneous initially with clinical assessment followed by a meta-analysis.

At least 2 of the review authors independently, in an unblinded standardized manner, performed each search and methodological quality assessment. The primary authors of assessed manuscripts were not involved in the methodological quality assessment. All searches were combined to obtain a unified strategy. Any disagreements between reviewers were resolved by a third author and consensus.

An analysis of the evidence was performed based on modified grading of evidence which was developed from Cochrane criteria of evidence synthesis and multiple other criteria including the US Preventive Services Task Force (USPSTF) analysis of evidence criteria as shown in Table 1 [46].

Summary measures included a 50% or more reduction in pain in at least 50% of the patients or at least a 3-point decrease in pain scores and a relative risk of adverse events, including side effects.

Randomized trials were judged to be positive if the intervention (fusion/disc arthroplasty or epidural injections) was clinically relevant and effective, either with a placebo control or active control, with a difference in effect for the primary outcome measure in a statistically significant manner at the conventional 5% level. Any improvement of less than 6 months was considered as short-term and 6 months or longer was considered as long-term for injection therapy and less than 12 months was considered as short-term for fusion. Since epidural injections can have a short-term benefit, repeat injections were allowed. Furthermore, the outcomes were judged at the reference point with positive or negative results reported at one month, 3 months, 6 months, 1 year, and 2 years.

Fig. 1 shows a flow diagram of the study selection as recommended by Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [45].

Of the multiple trials for consideration [11,47,48,49,50,51,52,53,54,55,56], 5 trials of lumbar fusion [47,48,49,50,51] and 2 trials of epidural injections [53,55] were considered for quality assessment. There were no trials available comparing total disc arthroplasty with conservative management. Of the 5 trials of lumbar fusion [47,48,49,50,51], 3 trials met the inclusion criteria of fusion surgery versus nonsurgical therapy [47,48,49,50] after exclusion of duplicates [47,48], and one trial [51] assessed fusion after disc excision. These 3 trials [47,48,49,50] and 2 trials of epidural injections [53,55] were considered for inclusion.

This systematic review comparing epidural injections and fusion in managing lumbar discogenic pain, based on a high quality methodological quality assessment and qualitative evidence synthesis of 3 trials comparing lumbar fusion with conservative management, and 2 trials utilizing epidural injections for management of discogenic pain, shows that caudal and lumbar interlaminar epidural injections with or without steroids provide effective and significant improvement in pain and function in lumbar discogenic pain with long-term results, with Level II evidence for caudal and interlaminar approaches. However, the evidence for fusion appears to be Level III-IV, based on significant improvements of 50% or more at the end of 2 years, applying the same criteria as epidural injections with a lack of efficacy demonstrated by RCTs comparing fusion with conservative management. However, considering a low 15% improvement as success, the evidence may be considered Level III-IV based on 3 randomized trials of moderate quality, with only one trial providing marginally better results than conservative management.

There were no direct comparative trials comparing fusion with epidural injections or disc arthroplasty with conservative management in managing discogenic pain. Thus, it appears that epidural injections may be superior to surgical fusion based on the available evidence with demonstration of cost utility [56]. Further, it is essential to take into consideration not only the costs, but also the complications of surgical interventions, including reoperation. Additional studies should include an economic analysis of costs, risks, and success of the various options for discogenic back pain. A rationale algorithm would include costs, outcomes, and risks of the proposed therapies. In this analysis, given the high success of epidural injections, the low costs and risks, when compared to surgical intervention epidurals should be considered early in a treatment continuum.

The evidence in this systematic review, while similar to previous systematic reviews, also is contradictory to multiple systematic reviews [8,9,12,29,57,58].

Multiple systematic reviews conducted in the past have sought to determine whether the fusion of the lumbar spine is superior to nonoperative management for the improvement of discogenic back pain. Mirza and Deyo [58] concluded that surgery may be more efficacious than unstructured nonsurgical care for chronic back pain, but may not be more efficacious than structured cognitive-behavioral therapy. Saltychev et al. [12], in their systematic review, concluded that there was strong evidence that lumbar fusion was not more effective than conservative treatment in reducing perceived disability because of chronic low back pain among patients with degenerative spinal diseases. Phillips et al. [8], in a systematic review of lumbar spine fusion for chronic low back pain due to degenerative disc disease, compiled and analyzed the currently available published literature on fusion as of July 2011, on fusion for chronic back pain with underlying disc degeneration, updating the evidence with recent studies, and broadening the scope of prior reviews to include a range of study designs beyond RCTs. In this extensive assessment, they included a total of 3,060 patients with a weighted average improvement and visual analog scale of back pain of 36.8 of 100, ODI of 22.2, with average satisfaction of 71.1 across the studies. They also showed radiographic fusion rates averaged 89.1% with a reoperation rate of 12.5%. They concluded that this body of literature supports fusion surgery as a viable treatment option for reducing pain and improving function in patients with chronic low back pain refractory to nonsurgical care when a diagnosis of disc degeneration can be made. However, multiple deficiencies may exist in this systematic review due to the inclusion of duplicate studies, thus increasing the number of patients assessed, as well as the inclusion of variable designs with highly variable surgical interventions, and conservative management with homogeneity. Bydon et al. [9], in contrast, concluded that despite the significant improvement in ODI in the lumbar fusion group in 3 studies, pooled data revealed no significant difference when compared with a nonoperative group. They included 5 RCTs meeting inclusion criteria; however, it appears that they may have included duplicate trials. Further, these trials lacked homogeneity to conduct a metaanalysis. Even then, the results were still equal between lumbar fusion and nonoperative management. They also showed that there was an overall improvement of 7.39 points in the ODI in favor of lumbar fusion [superior], however, it was unclear that this minuscule change in ODI would have led to a clinically significant difference. They also concluded that prospective randomized trials comparing a specific surgical technique versus a structured physical therapy program may improve evidence quality. Surprisingly, they also concluded that until then, either operative intervention by lumbar fusion or nonoperative management and physical therapy remained 2 acceptable treatment methods for intractable low back pain.

However, there are no systematic reviews, randomized trials, or observational studies comparing epidural injections as part of nonsurgical management in conjunction with other conservative modalities.

There was a single systematic review of nonoperative management of discogenic back pain [29]. In this assessment they identified 11 RCTs investigating traction therapy, injections, and ablative techniques. Results from 5 RCTs investigating methylene blue injection, steroid injection, ramus communicans ablation, intradiscal electrothermal therapy, and biacuplasty favored intervention over sham therapy. In this assessment, the trial by Manchikanti et al. [52] of caudal epidural injections was utilized and the level of evidence was 1.

Epidural injections are not only clinically effective, but also have been shown to be cost effective with caudal epidural injections with a cost utility assessment of $2,136 per quality-adjusted life-year (QALY) [56].

The limitations of this review include a paucity of literature without trials of disc arthroplasty comparing epidural injections to surgical trials, comparing to nonsurgical trials, and the inability to perform a metaanalysis due to a lack of homogeneity, either among epidural injections or lumbar fusion trials. All of the evidence is obtained from active control trials with epidural injections as well as for fusion. In this study, we utilized strict methodological quality assessment criteria and also had strict inclusion criteria with at least 25 patients in each group. Some may consider this as a deficiency as we have not eliminated one RCT from fusion and 2 RCTs from epidurals; however, size and quality are important conclusions and also provide strength to the systematic review.

In accordance with our objective of determining comparative efficacy, we have shown Level II evidence for epidural injections for long-term efficacy in managing chronic lumbar discogenic pain without facet joint or sacroiliac joint pain and also without disc herniation or radiculitis utilizing controlled diagnostic blocks and imaging in addition to symptomatology and physical findings. The continued debate in reference to the efficacy of epidural injections may be based on improper assessment utilizing local anesthetic as placebo and performing a meta-analysis on these trials without homogeneity which ultimately yielded inappropriate results [6,35,36,37,42,57,58,59,60,61].

In conclusion, fluoroscopically-directed epidural injections have been shown to be effective with Level II evidence with or without steroids, whereas, lumbar fusion, based on one moderate quality trial considered as borderline, showed effectiveness at Level III-IV.