Variations in selective nerve root block technique

INTRODUCTION Much literature reports on selective nerve root blocks (SNRBs) in cases of lumbosacral radiculopathy. Unfor-tunately, authors only inconsistently reveal the exact needle tip position relative to the causative pathology at the time of injection. Different injection sites may provide different symptomatic benefits. We investigated the variation in injection techniques of practitioners working in the UK. METHODS A clinical scenario was devised depicting a patient with radiculopathy secondary to an L4/5 vertebral disc prolapse. Participants were questioned on their chosen management of this patient, focusing particularly on SNRB technique. Questionnaires were sent to spinal surgeons, pain management specialists and musculoskeletal radiologists. RESULTS A total of 100 responses were detailed enough for inclusion. The majority (83%) of respondents reported they would inject local anaesthetic and steroids, 4% would inject local anaesthetic alone and 13% would inject a different substance. Over half (53%) would target the L5 nerve root, 26% the L4 nerve root, 12% the prolapsed disc itself and 9% two separate vertebral levels. Variation was also noted in needle tip location relative to the neural sheath. CONCLUSIONS When treating lumbar radiculopathy, there are apparent variations in the use and positioning of SNRBs for a given level of disc pathology. Needle tip position may have a direct influence on clinical outcome following SNRBs. Caution is therefore required when considering the validity of previously published studies investigating SNRBs and different injectates.

Lumbosacral radiculopathy is a common disabling condition with a reported prevalence of up to 43%. 1 When a prolapsed intervertebral disc directly inflames, compresses or stretches a nerve root, the resulting symptoms are often severe enough to force patients to seek medical attention. [2][3][4] The financial and functional impact of this disabling condition make its optimal management paramount.
Patients with lumbar radiculopathy and minor neurological deficits generally recover well regardless of whether they are managed non-operatively or operatively. 5 Based on this, to avoid the potential complications of surgery, selective nerve root blocks (SNRBs), first described in 1971, 6 have become increasingly popular. SNRBs can be employed with diagnostic or therapeutic intent. By targeting local anaesthetic injections at specific anatomical sites, SNRBs allow the source of a patient's pain to be predicted. This is particularly useful in cases of multilevel spinal disease or when symptoms do not correlate with the imaging performed. The use of various injectates has also been shown to provide permanent symptomatic relief in cases of both vertebral disc prolapse and spinal stenosis. [7][8][9] Despite the widespread use of SNRBs, a lack of level one evidence means there is still debate about whether injection of local anaesthetic alone [10][11][12] or a combination of steroids and local anaesthetic 13,14 is most effective at relieving symptoms. Interestingly, many of the studies reporting on this usually fail to specify the exact position of the needle tip at the time of injection relative to the level of the causative pathology. It would seem intuitive that this is an important parameter in the efficacy of SNRBs. Consequently, if clinicians' injection sites vary for a given clinical scenario, then conclusions and comparisons drawn from these studies should be interpreted with caution. The external validity of meta-analyses and critical appraisals would also need to be questioned, given the potential for unaccounted variability in study designs. 15 In order to look at this further, a survey was performed of UK medical practitioners with an interest in the management of radicular pain. The aim was to investigate how varied their techniques are when performing an SNRB. In particular, the level of the SNRB relative to the level of the causative pathology was investigated.

EASTLEy SPITERI NEWEy
VArIAtIoNS IN SElEctIVE NErVE root BlocK tEchNIquE Methods A clinical scenario was devised depicting a patient suffering from radicular pain secondary to an L4/5 intervertebral disc prolapse. A questionnaire containing the scenario was sent to spinal surgeons, pain management specialists and musculoskeletal radiologists throughout the UK. Axial and sagittal magnetic resonance imaging showing a left-sided paracentral herniated disc and a diagram of the spinal segment related to the region were presented (Figs 1-3). Participants were asked to mark on the diagram where they would place the needle when performing an SNRB in this case. They were also asked for information about which substances they used.

results
Questionnaires were sent to 319 practitioners and 153 responses were received. Insufficiently completed questionnaires and responses from practitioners who never perform SNRBs or would choose not to use a SNRB in the case presented were excluded. This left 100 eligible questionnaires for analysis (35 spine surgeons, 29 radiologists and 36 pain specialists).
Of the 100 respondents included in the analysis, 87% indicated they would inject a combination of local anaesthetic and steroids. Only 3% stated they would inject local anesthetic alone with the remaining 10% choosing to inject either steroids in isolation, or mixed with clonidine or radiological contrast.
Only 39% of respondents said they would inject at the level of the pathology (L4/5) while 55% reported they would inject at the next caudal level. There was some variability between specialties ( Table 1). The majority of practition- Figure 1 Axial magnetic resonance imaging showing the leftsided pathological disc herniation at L4/5 responsible for the radicular symptoms in the hypothetical case study. The disc is highlighted by an arrow Figure 3 Coronal diagram of the spinal segment affected by the prolapsed intervertebral disc responsible for the symptoms in the hypothetical case study. The disc is highlighted by an arrow. Participants were asked to mark the exact anatomical position of their needle tip at the time of injection on this diagram Figure 2 Sagittal magnetic resonance imaging of the disc herniation at L4/5 responsible for the radicular symptoms in the hypothetical case study. The disc is highlighted by an arrow ers who indicated that they would inject steroids also said that they would target the level below the pathology ( Table  2). Six per cent reported they would inject at two vertebral levels  16 A transforaminal approach to the epidural space is adopted by most practitioners. 17 In cases of far lateral disc herniation, extraforaminal injections have also been shown to be effective. 7 Injections can be placed inside the neural The majority of published studies investigating SNRBs and radicular leg pain fail to specify the precise needle tip location at the time of injection relative to the causative pathology. Although injection techniques are usually described in some part, authors often fail to define whether they target the vertebral level where pathological compression occurs or the exiting nerve root at the vertebral level below. The vertebral levels of the pathological discs and stenoses involved are also often reported inconsistently (Table 3). We surveyed how a range of practitioners would perform an SNRB when faced with one specific clinical and radiological scenario. Although our initial response rate was low at 48% (a recognised limitation of postal surveys), responses demonstrated that variation exists between their chosen techniques, including the intervertebral level targeted.
Lower back pain is a common complaint and results in a significant number of patients requiring long-term sick leave. 22 Non-operative management plays a large role in the management of those patients with radiculopathy secondary to nerve root impingement. SNRBs are used frequently for both diagnostic and therapeutic purposes. Our survey suggests that clinicians use a wide variety of SNRB injec-tion techniques for a given clinical scenario. Extrapolating this to previous studies reporting on the clinical effectiveness of SNRBs, we feel that unless authors clearly communicate their injection techniques (specifically, the injection site relative to the causative pathology), the external validity of their results has to be questioned. This is particularly important given that the most effective injectate is still debated by many.
Although the pathophysiology of radiculopathy secondary to disc herniation is not fully understood, local inflammation is thought to play a role. 4 The systemic antiinflammatory effects of steroids are well recognised. Local anti-inflammatory effects of methylprednisolone on nerve roots coated with nucleus pulposus have also been reported. 23 It would therefore seem reasonable to assume that if steroids are to have a role in SNRBs for radicular pain, they should be targeted at the site of the pathology (ie at the level of the disc lesion where the touching nerve root is potentially inflamed). This is supported by published work. 24 In our survey, out of 100 practitioners who used steroids, only 38 targeted the level of the disc lesion. Although we recognise that local anaesthetic injections given caudally to the site of inflammation may provide short-term benefit, we feel that targeting steroids at the next caudal foramen is of questionable value and comparable with treating an arthritic hip by injecting the knee!

conclusions
The results of our survey highlight a number of issues. First, as might be expected, there is variation in the technique used between clinicians for a given pathology. Some of this variation may relate to the interpretation of symptoms in terms of which nerve root is involved. However, the pathology described was present at the L4/5 level and we have some difficulty understanding why steroid injections would be aimed at the level below. This argument would apply regardless of whether or not SNRBs were being used for diagnostic or therapeutic purposes.
Second, our survey emphasises the importance of being precise when requesting an SNRB from a radiologist or other practitioner. This will ensure any subsequent decisions made on the suitability of a patient for surgery based on the therapeutic effect of an SNRB will be meaningful.
Finally, our survey highlights the need for caution when applying the results of published studies to one's own practice. There is a need for authors of future studies to clearly describe their own injection technique including the exact position of the needle tip in relation to disc pathology. This will allow readers to compare the results of published studies with their own practice in a more meaningful way.