Article Text
Abstract
Background Epidural steroid injections (ESIs) are a frequently used treatment for refractory radicular spinal pain. ESIs, particularly transforaminal epidural steroid injections (TFESI), may provide pain relief and delay the need for surgery. Corticosteroid agent and diluent choices are known to impact the safety of ESIs. In particular, the risk of embolization with particulate corticosteroids has led to recommendations for non-particulate steroid use by the Multisociety Pain Workgroup. Additionally, there is in vitro evidence that ropivacaine can crystalize in the presence of dexamethasone, potentially creating a particulate-like injectate. Despite widespread use and known risk mitigation strategies, current practice trends related to steroid and diluent choices are unknown.
Objective Identify the use of particulate versus non-particulate corticosteroids for epidural steroid injections in the cervical and lumbar spine, as well as local anesthetics commonly used as diluents during these procedures.
Methods Cross-sectional survey study of 314 physician members of the Spine Interventional Society.
Results 41% and 9% of providers reported using particulate corticosteroids during lumbar TFESIs and cervical TFESI, respectively. Four per cent of providers reported the use of ropivacaine in cervical TFESIs. Forty-four per cent of respondents reported using anesthetic in cervical interlaminar ESIs. 21% of providers report using high volumes (> 4.5 mL) during cervical interlaminar ESIs.
Conclusion Current trends, as assessed by this survey study, indicate substantial variability in steroid and diluent choice for ESIs. Patterns were identified that may impact patient safety including the continued use of particulate corticosteroids for TFESIs and the use of ropivacaine during TFESIs by a subset of respondents.
- spinal/epidural injection
- chronic pain: back pain
- chronic pain: head and neck
- interventional pain management
Statistics from Altmetric.com
- spinal/epidural injection
- chronic pain: back pain
- chronic pain: head and neck
- interventional pain management
Introduction
Epidural steroid injections (ESIs) are a frequently used treatment for refractory radicular spinal pain. ESIs, particularly transforaminal epidural steroid injections (TFESI), demonstrate the ability to provide pain relief and delay and/or reduce the potential need for surgery.1 2 Injectable corticosteroids can be classified into particulate or non-particulate agents. Commercially available particulate steroids include triamcinolone, betamethasone and methylprednisolone. The primary commercially available non-particulate steroid is dexamethasone. It has been theorized that injecting particulate steroids through the transforaminal approach can lead to increased risks as this approach targets a location close the location of radiculomedullary arteries feeding the spinal cord.3 Possibly, cases of permanent neurological injuries have occurred because agents in particulate steroids larger in size than red blood cells can block the flow of these radiculomedullary arteries to the spinal cord.4 As such, in 2014, the Multisociety Pain Workgroup (MPW), a collective group representing over 100 000 members from diverse training backgrounds and medical specialty societies including anesthesiology, interventional pain medicine, radiology, physiatry and neurosurgery, among others, developed safety recommendations for the performance of ESIs. These include a recommendation for the exclusive use of non-particulate steroid during all cervical TFESIs and for initial lumbar TFESIs.5 No recommendation was provided regarding particulate versus non-particulate steroid use during interlaminar ESIs, as this approach targets locations further from these arteries. It is a common practice to include a local anesthetic in the injectate when performing ESIs. The rationale is to provide immediate analgesia for diagnostic value and also to mitigate pain associated with the temporary additional compression and/or of sensitive neural elements by the injectate volume added.6 The MPW safe use initiative did not provide specific guidelines regarding local anesthetic diluent choice or dose. Studies report using lidocaine, bupivacaine or ropivacaine in varying doses for ESIs.7 8 Intravascular injection of local anesthetics can affect the central nervous system and cardiovascular system, as well as cause an allergic response to both the local anesthetic and preparation additives.9 Additionally, ropivacaine, a weak base with poor solubility in water demonstrates crystallization at physiologic pH when combined with dexamethasone; this has prompted warning with use in the proximity of arterial blood supply to neural tissue, due to the same theoretical concern for use of particulate steroid.10 11
To date, there have been no published studies of provider practice patterns related to steroid and diluent choices during ESIs. It is also unknown whether current practice patterns are concordant with the recommendations published by the MPW. As such, the aim of this study was to describe current trends among practitioners who perform ESIs with regard to steroid and local anesthetic diluent agent choice, in the context of the MPW guidelines and recent literature pertaining to potential risks of local anesthetic use during ESIs.
Methods
Data were collected through administration of a survey constructed using Research Electronic Data Capture (REDCap) and were distributed via email to physician members of the Spine Intervention Society (SIS) between May and June 2018. Physicians who received the survey included current trainees, members of the American Boards of Medical Specialties or the American Osteopathic Association within the USA or Canada, or those who hold similar status in their country of residence in the fields of Anesthesiology, Physical Medicine and Rehabilitation, Radiology, Neurology, Orthopedic Surgery or Neurosurgery. All responses were anonymous, and no identifying information was collected from survey respondents. Although it is impossible to know the exact number of physicians who viewed the survey, 3123 physicians opened the email from at total of 5907 emails that were sent. The survey included questions specific to steroid choice and diluent use, as outlined in the appendix.
Supplemental material
Data analysis
Data analysis was performed by author DC with GNU PSPP (V.1.2.0, Boston, Massachusetts, USA). Means and SD were calculated for continuous data. Proportions were calculated for categorical data. Survey results were assessed using logistic regression analysis to identify associated factors, with covariates including academic versus non-academic practice setting, medical specialty, years in practice, fellowship training and the geographic location where physicians practice.
Results
Demographics of survey respondents
The survey was completed by 314 physicians for a response rate of 11%. Respondents consisted of 116 anesthesiologists (36.9%), 154 physiatrists (49.0%), 10 radiologists (3.2%) and 34 individuals from other specialties (10.8%) which may include neurology, orthopedic surgery or neurosurgery, or those who hold similar status in their country of residence to any of these specialties. Eighty per cent of respondents reported that they received their training and currently practice in the USA. Accreditation Council for Graduate Medical Education-accredited fellowship training had been attained by 30% of respondents. The most common practice setting was small group, private practice. The mean number of years of interventional pain practice was 14.02 (SD=10.01). The table 1 shows the number of providers who reported performance of each type of procedure, with lumbar TFESIs (307, (97.8%)) being most common and cervical TFESIs being the least common (129, (39.5%)).
Respondents’ steroid and diluent choices
The figure 1 illustrates the rate of use of different local anesthetic diluents based on procedure type, and figure 2 represents the total injectate volume used for each procedure type. Lidocaine was the most commonly reported, while ropivacaine was the least commonly reported local anesthetic diluent used for all procedure types. Four per cent of providers report using ropivicaine for cervical TFESI. Forty-three per cent of providers report the use of local anesthetic in cervical ILESI. The total volume of injectate varied widely based on procedure type, and of note, nearly 25% and 4% of providers reported using >5 mL of injectate for lumbar ESIs and cervical ESIs, respectively. Median volumes were cervical TFESI—2.0 mL, lumbar TFESI—3.0 mL, cervical ILESI—3.0 mL and lumbar ILESI—5.0 mL.
The table 2 illustrates practitioner use of particulate steroids, stratified by procedure type. Approximately 9% of providers report the use of particulate steroids for cervical TFESI, 41% for lumbar TFESI, 49% for cervical ILESI and 74% for lumbar ILESI. Logistic regression demonstrated no factors associated with use of particulate steroids. Physical Medicine and Rehabilitation (PM&R)-trained physicians were more likely to use lidocaine for cervical TFESIs (p=0.016, OR 7.33 (95% CI 1.46 to 36.76)), while anesthesia-trained physicians were less likely to use lidocaine for cervical TFESIs (p=0.013, OR 0.19 (95% CI 0.05 to 0.70)).
With regard to total volume of injectate, there were several statistically significant associations. First, PM&R training was associated with the use of higher volumes of injectate for cervical TFESIs (p=0.029, b=0.46 (95% CI 0.05 to 0.87)). Second, those practicing in the South (p=0.018, B=−0.47 (95% CI −0.86 to −0.08) and West (p=0.031, B=−0.43 (95% CI 0.83 to −0.04)) were less likely to use higher volumes of injectate for cervical TFESIs. Third, those identified in the ‘other’ category for primary specialty were more likely to use higher injectate volumes for lumbar TFESI when compared with anesthesia (p<0.001, b=−1.06 (95% CI −1.35 to −0.57)), PM&R (p=0.027, b=−0.56 (95% CI −1.05 to −0.06)) and radiology (p=0.035, b=−0.94 (95% CI −1.81 to −0.07)) trained physicians. Physicians practicing in the West were more likely to use lower injectate volumes for lumbar TFESIs (p=0.004, b=−0.94 (95% CI −1.81 to −0.07)). Physicians practicing in the Midwest were more likely to use higher injectate volumes for cervical ILESIs (p=0.031, B=−0.84 (95% CI −1.61 to −0.08)). Physicians practicing outside the US and those with more years of practice experience were more likely to use lower injection volumes for lumbar ILESI’s.
Discussion
This is the first survey to describe practice trends regarding steroid and diluent choices during ESIs. The study identified guideline-discordant practice trends with regard to steroid choice for TFESI in relation to the 2015 MPW recommendations for safe practice.5 The present data indicate that 41% of providers continue to use particulate steroids during lumbar TFESIs at some point, with over 25% using them in at least half of their lumbar TFESIs. While the MPW guidelines do allow for use of particulate steroid during lumbar TFESIs, use on during the first injection is not recommended. Further, these guidelines state that ‘there are situations where particulate steroids could be used in the performance of lumbar TFESIs’, implying the routine use is not recommended. The MPW guidelines are unequivocal regarding advising against the use of particulate steroid during cervical TFESI, yet, roughly 9% of providers report the use of particulates for cervical TFESI in this survey. Safeguards including live fluoroscopy with contrast injection help to mitigate risk of use of particulate steroid during cervical TFESI. However, studies have shown that even with appropriate needle placement inadvertent intra-articular injection can occur as even the slightest needle movement can lead to access to arterial flow.12–14 Particulate steroids have been implicated in several case reports as contributing to rare, but devastating neurovascular complications if injected into the vasculature.15 16 Two possible mechanisms have been described in the literature. First, red blood cell agglutination and resulting embolizations.17 18 Second, steroids containing large particles may directly obstruct arterioles if particulates are larger than the aerteriole.4 19
The reason why providers continue to use particulate agents is not clear, though it may relate to a perceived treatment benefit associated with particulate compared with non-particulate steroids. It is important to acknowledge that the literature regarding effectiveness of particulate versus non-particulate steroid remains mixed. A 2010 prospective, randomized, double blind trial demonstrated no significant difference in the proportion of participants who reported clinically meaningful pain relief when TFESI with particulate or non-particulate steroids were administered for acute radicular pain due to lumbar disc herniation. A large retrospective study (n=2634) demonstrated non-inferiority of dexamethasone compared with particulate steroid during TFESI for the treatment of lumbar radicular pain related to heterogeneous structural etiologies. However, patients treated with non-particulate steroids required more injections to maintain a similar degree of pain reduction, indicating that pain reduction occurs for longer when particulate agents are used.14 More recent studies found no significant differences in pain reduction or the number of repeat injections when comparing particulate versus non-particulate corticosteroids for lumbar transforaminal ESIs.14 20 21 Alternatively, a 2018 retrospective study showed that a significantly higher proportion of patients treated with particulate steroids reported improvement at 1 week and 1 month compared with those treated with non-particulate steroid during lumbar TFESI.22 This finding is mirrored in a randomized comparative trial of dexamethasone versus particulate steroid during cervical TFESI, which also demonstrated a significantly greater proportion of treatment responders in the particulate steroid group.23 However, the safety profile of non-particulate steroid, despite rare case reports of neurological injury24 outweigh any potential differences in clinical effectiveness compared with particulate agents during using in TFESIs, which forms the basis for the MPW’s recommendations.
Approximately 50% and 20% of providers are using dexamethasone during cervical and lumbar interlaminar ESIs, respectively. This finding is interesting as there is no evidence that non-particulate steroid use provides any safety benefit compared with particulate steroid when administered via an interlaminar route. Per literature review, there has never been a report of an embolic cord infarction from an interlaminar epidural injection as the vertebral artery (in the case of cervical interlaminar ESIs) and radiculomedullary arteries are located anterior to the needle path within the dorsal epidural space during an interlaminar approach. Again, reports of the effectiveness profile of non-particulate versus particulate remain mixed with regard to overall effectiveness of particulate versus non-particulates.
Four per cent of providers reported use of ropivacaine during cervical TFESIs. This is concerning given the risk of crystallization when mixed with dexamethasone, as identified in a recent study by Watkins et al.10 11 Thus, providers performing cervical TFESI according to MPW recommendations with non-particulate steroids may unknowingly combine medications known to crystalize and perhaps lead to an embolic event.
Forty-four per cent of providers reported local anesthetic use in cervical interlaminar ESIs. Anesthetic use in cervical interlaminar procedures may put patients at risk for high spinal block, subdural block with loss of consciousness25 or temporary upper extremity weakness and risk for falls.6 While the MPW recommendations do not comment on the use of local anesthetic during cervical interlaminar injections, the SIS guidelines do suggest that the risks of this practice likely outweigh the benefits.7
With regard to total injectate volume, the present data indicates that nearly 25% of providers use more than 5 mL of injectate during lumbar ILESIs. For cervical TFESI, the majority of providers report using 1.0–2.5 mL of total injectate volume (79%) with no providers reporting the use of >4 mL. For cervical ILESI, 21% of respondents report using >5 mL. There are no clear guidelines that address the optimal volume for ILESI. Kim et al found that 3 mL of injectate reached a median of 3.96±1.09 spinal segments in cervical interlaminar injections.26 27 However, it is not clear what total injectate volume is needed to reach the most cephalad spinal segments, and the impact of injectate volume on clinical outcomes has not been rigorously addressed. Limited literature regarding the effect of volume during lumbar TFESI suggests a possible benefit with larger volumes; specifically, Chun et al found 8 mL to more beneficial than 3 mL.28 29 It is important to note that higher volume injectate has been implicated in contributing to increased intracranial pressure and accompanying side effects such as retinal hemorrhage/Terson syndrome.30 Therefore, the risk of higher volume injections must also be balanced with perceived improved efficacy with higher volume injections.
There are limitations to our study which must be recognized. Though this is the largest study to date examining practice trends in steroid agent and diluent choice associated with ESI, the 11% response rate may not accurately reflect the patterns of all interventional providers, particularly given that only physician members of the SIS were surveyed. Additionally, the survey elicited information regarding steroid and diluent choice, but questions were not asked regarding specific combinations of injected medications and did not differentiate between 1% and 2% lidocaine or whether preservative anesthetics were used or at what specific point in the treatment process particulate steroids were used. The survey elicited information distinguishing between the interlaminar and transforaminal approach but did not specify the various routes of transforaminal access employed. The precise fluoroscopic views used during needle guidance and the type of loss of resistance technique used during interlaminar epidural access were not specifically queried in the survey. Lastly, due to a risk for response and selection bias, incidence rates of complications were not extrapolated from the present study data. As with all survey-based questions, our results are subject to recall bias.
In conclusion, this survey examined current practice trends of SIS physician members with regard to steroid and diluent choices for epidural corticosteroid injections and identified significant variability. Significant trends were identified that may present patient safety concerns. Most notably, the survey identified MPW guideline-discordant use of particulate steroid use during TFESIs, which place patients at risk for rare but catastrophic neurovascular complications. Additionally, ropivacaine use during TFESI is reported among a small number of respondents, which may further increase this risk if mixed with dexamethasone. It is unclear why these observed practice trends exist; however, it is clear that to better optimize patient safety during ESI, clear guidelines outlining the known and potential risks may need to be reviewed to ensure that preventable complications are avoided.
Acknowledgments
The authors would like to thank Shellie Cunningham, from the University of Utah for assisting with manuscript preparation. We also thank Sandra Ray, MPH for her assistance in survey distribution to the SIS membership and to those members who participated in the survey.
References
Footnotes
Contributors All authors made substantial contributions to the study design, acquisition, analysis and interpretation of data, revised critically important intellectual content, contributed to final approval of the version submitted for publishing and agreed to be accountable for all aspects of the work and that questions related to the accuracy or integrity of the work were appropriately investigated and resolved.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests ZM on the board of directors of the Spine Intervention Society. None of the other authors declare conflicts of interests.
Patient consent for publication Not required.
Ethics approval Institutional Review Board (IRB) approval for the study was acquired through the University of Utah School of Medicine.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Data are available upon reasonable request.