Peripheral Nerve Blocks in Non-Operative Settings : A Review of the Evidence and Technical Commentary

This narrative review summarizes and comments the evidence derived from randomized controlled trials pertaining to the efficacy of peripheral nerve blocks in non-operative settings. The literature search was conducted using the Medline (1966-present), Embase (1980-present), Web of Science (1900-present) and Sciverse Scopus (1996-present) databases. The following search terms were used: (“peripheral nerve block” OR “brachial plexus block” OR “interscalene block” OR “supraclavicular block” OR “infraclavicular block” OR “axillary block” OR “humeral canal block” OR “lumbosacral plexus block” OR “lumbar plexus block” OR “femoral nerve block” OR “lateral femoral cutaneous block” OR “obturator nerve block” OR “sciatic nerve block”) AND (“fractures” OR “Emergency Room” OR “Emergency Department” OR “ambulance” OR “prehospital” OR “Intensive Care Unit” OR “Intensive Care”). Only randomized controlled trials were retained for analysis. Despite methodological shortcomings, the available evidence suggests that peripheral nerve blocks can provide pain control for upper and lower limb trauma in non-operative settings. For instance, brachial plexus blocks offer a useful alternative to procedural sedation for fracture manipulation in the Emergency Department. Lumbar plexus, 3-in-1 and femoral blocks can provide analgesia for patients with hip fractures. Femoral blocks also result in more comfortable ambulance transfers to the hospital for patients suffering from hip and knee trauma. Finally, in very elderly subjects, fascia iliaca blocks can decrease the incidence and duration of perioperative delirium. Published reports of randomized trials provide evidence to formulate limited recommendations regarding the use of peripheral nerve blocks in non-operative settings. Further well-designed studies are warranted. J o u r n a l o f A ne sth esia & Cical es e a r c h ISSN: 2155-6148 Journal of Anesthesia & Clinical Research Citation: Tran DQH, Bernucci F, Tiyaprasertkul W, Finlayson RJ (2014) Peripheral Nerve Blocks in Non-Operative Settings: A Review of the Evidence and Technical Commentary. J Anesth Clin Res 5: 398. doi:10.4172/2155-6148.1000398


Introduction
Increasingly, peripheral nerve blocks are being carried out in nonoperative settings [1][2][3][4][5][6][7][8][9][10][11][12][13]. For instance, Emergency department (ED) physicians and orthopedic surgeons often perform blocks to facilitate fracture manipulation and to provide analgesia for fracture-related pain [7][8]. Expectedly this has fostered heated debates between and within professional societies. In the near future, important questions will demand answers: should peripheral nerve blocks be performed by anesthesiologists only or could they be carried out by other physicians as well? What is the minimum training required to ensure proficiency and safety? Which governing body will legislate technical competency and implement quality insurance? If blocks were the sole domain of Anesthesiology departments, how best to reorganize Acute Pain Services to efficiently meet the growing needs of EDs and Intensive Care units?
Answers to these complex questions require first and foremost proof that nerve blocks improve patient care in non-operative settings. Previous review articles have attempted to succinctly summarize the available information [1][2][3][4][5][6][7][8][9][10][11][12][13]. Unfortunately, because of the various backgrounds and experience levels of the operators, the techniques and equipment employed varied substantially. Although it is paramount to ensure that the methods used do not deviate significantly from accepted technical standards, this potential shortcoming has been overlooked by all review articles so far [1][2][3][4][5][6][7][8][9][10][11][12][13]. Thus we undertook a systematic review of the literature: our goal was to analyze all level 1 evidence (randomized controlled trials) pertaining to the efficacy of peripheral nerve blocks in non-operative settings. Moreover we sought to compare the block techniques used with those advocated by the best evidence available.

Methods
The literature search for this review article was conducted during the first week of June 2013 using the Medline (1966 to present), Embase (1980 to present), Web of Science (1900 to present) and Sciverse Scopus (1996 to present) databases. The following search terms were used: ("peripheral nerve block" OR "nerve block" OR "peripheral block" OR "block" OR "regional block" OR "regional anesthesia" OR "local anesthesia" OR "local anesthetic" OR "brachial plexus nerve block" OR "brachial plexus block" OR "interscalene block" OR "supraclavicular block" OR "infraclavicular block" OR "axillary block" OR "humeral canal block" OR "brachial canal block" OR "midhumeral block" OR "median nerve block" OR "radial nerve block" OR "ulnar nerve block" OR "musculocutaneous nerve block" OR "lumbosacral plexus block" OR "lumbar plexus block" OR "psoas block" OR "psoas compartment block" OR "psoas sheath block" OR "femoral nerve block" OR "lateral femoral cutaneous block" OR "obturator nerve block" OR "sciatic nerve block" OR "infragluteal block" OR "subgluteal block" OR "popliteal block" OR "ankle block" OR "transversus abdominis plane block" OR "ilioinguinal nerve block" OR "iliohypogastric nerve block" OR "thoracic paravertebral block" OR "intercostal nerve block") AND ("fractures" OR "Emergency Room" OR "Emergency Department" OR "ambulance" OR "prehospital" OR "Intensive Care Unit" OR "Intensive Care").
From this initial search, only randomized controlled trials (RCTs) conducted in human subjects, published in the English language and comparing peripheral nerve blocks to a control group (conventional treatment) were retained for analysis. No RCTs were excluded based on factors such as sample size justification, statistical power, blinding, definition of intervention allocation or primary and secondary outcomes. However non-randomized studies, case reports and cohort studies were eliminated, as were RCTs published in abstract or correspondence form. In addition, trials examining hematoma block, intravenous regional block (i.e. Bier's block), intraarticular infiltration, neuraxial blocks as well as those comparing various concentrations of local anesthetic or various permutations of a given block (i.e. digital block) were excluded. A secondary search for additional material was also undertaken by examining the reference lists of the selected articles as well as our personal files.

Results
Our search criteria yielded 14 RCTs pertaining to peripheral nerve blocks in non-operative settings (Tables 1 and 2). Of these studies (average sample size = 64.9 subjects), only 9 (64.2%) and 4 (28.6%) trials provided sample size justification and blinded assessment, respectively. Most RCTs (85.7%) defined a clear primary endpoint: the latter consisted of pain, incidence of delirium, cumulative opioid consumption or length of stay in the ED. For the control group, the analgesic regimen consisted of intravenous opioids, metamizole, paracetamol or nonsteroidal anti-inflammatory drugs (NSAIDs). As for the treatment group, the peripheral nerve blocks were carried out by ED physicians or orthopedic surgeons in all but 3 trials.

Upper extremity injury
Despite the widespread use of peripheral nerve blocks to provide analgesia for upper extremity injury in the ED, to date, only 2 RCTs (combined n = 83) have compared nerve blocks to conventional treatment [14][15]. In the first trial, Kriwanek et al. [14] randomized 41 pediatric patients (8 years or older) with forearm fractures to deep sedation or axillary brachial plexus block (AXB) for fracture reduction. In the deep sedation group, the authors administered midazolam (0.1 mg/kg, up to a maximum of 2 mg) and ketamine (initial bolus: 1 mg/kg, followed by additional doses titrated to patient comfort). For AXB, the authors employed a transarterial technique and injected 0.7 mg/kg (up to 40 mL) of adrenalized lidocaine 1%. During fracture manipulation, a pediatric nurse evaluated the patient's pain and distress according to the Children's Hospital of Eastern Ontario Pain Scale (CHEOPS). Kriwanek et al. [14] found no differences in CHEOPS scores. Furthermore, orthopedic and patient satisfaction was also similar between the 2 groups. Interestingly, in the AXB group, 2 patients displayed residual motor function with presence of pain during sensory testing and 2 subjects required supplemental analgesia during fracture manipulation. In 2011, Blaivas et al. [15] randomized 42 adult patients with shoulder dislocation to procedural sedation (etomidate) or interscalene brachial plexus block (ISB). For ISB, the authors employed an ultrasound (US)guided technique and deposited 20-30 mL of adrenalized lidocaine (concentration unspecified) around the roots of the brachial plexus. Blaivas et al. [15] observed a shorter stay in the ED with ISB (100.3 ± 28.2 vs. 177.3 ± 37.9 minutes; P < 0.0001). However post reduction pain levels and patient satisfaction were similar between the 2 groups. Despite the high local anesthetic (LA) volume, the authors observed a 0%-incidence of Horner's syndrome with ISB.
In summary, the available literature suggests that, for fracture manipulation in the ED, brachial plexus blocks provide a useful alternative to conventional procedural sedation.

Technical commentary
Axillary rachial plexus block: An enviable track record for safety explains the historical popularity of AXB. Traditional techniques such as fascial clicks, transarterial injection and elicitation of paresthesia (EP) usually result in low success rates (70-80%) [16]. Prior to the advent of US, peripheral nerve stimulation (PNS) constituted the most reliable adjunct. With PNS, the musculocutaneous, median, radial and ulnar nerves can be independently located and selectively targeted with LA. Extensive work by Sia et al. [17][18][19] suggests that, for PNS-guided AXB, a triple-stimulation technique offers the best combination of efficacy and efficiency.   [14], 2 patients displayed residual motor function and 2 subjects required supplemental analgesia during fracture manipulation. Thus one can speculate that the AXB would have been more efficacious had these authors selected PNS or US. Finally, AXB requires the patient to abduct the shoulder and flex the elbow. This may result in considerable discomfort in distal humerus/elbow trauma. In these subjects, supraclavicular (SCB) and infraclavicular (ICB) brachial plexus blocks constitute valuable alternatives [24]. However the efficacy/efficiency of SCB and ICB needs to be formally compared to conventional treatment in a non-operative context.

Interscalene brachial plexus block:
The ISB is commonly used to anesthetize the shoulder and proximal humerus. Identification of the brachial plexus in the interscalene groove has been traditionally achieved with EP or PNS [16]. Recently, US have contributed to the surge in popularity of ISB. Compared to PNS, Kapral et al. [25] found that US improved the rate of surgical anesthesia (98.8 vs 91.3%; P<0.01) as well as the onset (10 vs 22 minutes; p<0.05) and offset times (899 vs 679 minutes; P<0.05). Using ropivacaine 0.5%, McNaught et al. [26] reported that, compared to PNS, US decreased the minimal effective volume for analgesia in 50% of patients (MEV 50 ) from 5.4 to 0.9 mL (P=0.034). In contrast, after randomizing 219 patients to PNS or US, Liu et al. [27] observed no differences in performance time, surgical anesthesia and patient satisfaction. However patients in the US group required fewer needle passes (1 vs 3; P<0.001). Thus, according to the available evidence, it is not clear if US increases the success rate for ISB compared to PNS. However, US seem to provide better efficiency (fewer needle passes, decreased onset time); moreover it allows a decrease in LA volume.
In 2008, Riazi et al. [28] set out to investigate the optimal LA volume for ISB. Using a combination of PNS and US, these authors compared injectates of 5 and 20 mL of ropivacaine 0.5%. There were no differences in pain scores, sleep quality and postoperative morphine requirements up to 24 hours after surgery. However, patients receiving 5 mL presented a lower rate of diaphragmatic paralysis (45 vs 100%), fewer side effects (Horner's syndrome, hoarseness, respiratory distress) as well as smaller reductions in forced expiratory volume, forced vital    [28]. In contrast, Sinha et al. [29] compared volumes of 10 and 20 mL of adrenalized 0.5% ropivacaine and found no differences in the incidence of diaphragmatic paralysis (93%) [29].
In summary, ISB constitutes a valuable tool to provide analgesia for patients with shoulder injury. In light of the improved efficiency, preference should be given to US techniques. If an US machine is unavailable, the operator should consider employing a PNS method. Thus, in their 2011 RCT, Blaivas et al. [15] selected an appropriate technique for their ISB group. However the non-blinded assessment and patient follow-up represents a confounding variable. For instance, the authors noted a 0% incidence of Horner's syndrome despite using 20-30 mL of lidocaine. In the literature, complete absence of Horner's syndrome has been only been reported with very low volumes (5 mL) [28]. The authors' potential bias favoring ISB could have led them to under record the occurrence of adverse events and to discharge their patients more promptly from the ED. In fact, Blaivas et al. [15] opined that "it might be helpful to explore safe alternatives such as regional anesthesia to decrease the length of stay in the ED" and that "complications such as respiratory depression and aspiration can potentially be avoided with US-guided regional anesthesia". However one should not forget that, with high LA volumes (34-52 mL), ipsilateral phrenic paralysis is inevitable [30]. Moreover, as demonstrated by Riazi et al. [28], even LA volumes as low as 5 mL can block the phrenic nerve in 45% of subjects. Thus ISB should be viewed as an additional analgesic tool to efficiently manage patients with shoulder trauma and not a modality to expedite discharge by circumventing post-procedural patient monitoring. Moreover, due to its inherent risk of phrenic paralysis, ISB should be used with caution in patients with preexisting pulmonary compromise. In these subjects, combined infraclavicularsuprascapular [31] or axillary-suprascapular [32,33] blocks could be employed to anesthetize the shoulder. However their efficacy and efficiency require further validation in non-operative settings.

Lower extremity trauma
Except for 1 trial that tackled knee trauma, RCTs investigating the use of peripheral nerve blocks in lower extremity injury have focused exclusively on the hip joint and femur. Four nerve blocks have been studied so far: lumbar plexus, 3-in-1, fascia iliaca and femoral blocks.

Lumbar plexus block (LPB):
In 1999, Chudinov et al. [34] randomized 40 patients with femoral neck fractures to intermittent LA boluses administered through a LPB catheter (1-2 mg/kg of adrenalized bupivacaine 0.25% every 8 hours) or intramuscular meperidine (1 mg/ kg every 5 hours) combined with diclofenac (1 mg/kg for breakthrough pain). In the experimental group, the psoas compartment was identified with loss of resistance. Chudinov et al. [34] reported that patients randomized to LPB achieved lower pain scores at 8 and 16 hours during the 48-hour preoperative period (both P<0.05).

3-in-1 Block:
Three RCTs have investigated the role of 3-in-1 blocks in the setting of hip and femoral fractures [35][36][37]. In 2 trials (combined n=83) comparing single-injection 3-in-1 blocks to intravenous morphine, patients randomized to the former experienced a quicker onset of analgesia [35] and lower pain scores at 30 minutes [36]. While one study observed decreased hourly consumptions of morphine (0.49 mg/h vs. 1.17 mg/h) with 3-in-1 blocks [35], the other found no differences in 24 hour-cumulative morphine consumption or pain levels after 30 minutes [36]. In 2013, Luger et al. [37] randomized 20 very elderly patients (> 80 years) with hip fractures to an US-guided continuous 3-in-1 block or intravenous piritramide (0.05 mg/kg). The perineural catheters were bolused with 30 mL of bupivacaine 0.25% and infused with 6 mL/h of bupivacaine 0.125%. In the preoperative period, Luger et al. [37] found lower dynamic pain scores (all P<0.05) and daily paracetamol consumption (0.1 ± 0.32 vs. 1.7 ± 1.4 mg/d; P<0.05) with 3-in-1 blocks. However there were no intergroup differences in pain scores at rest or daily piritramide requirements (0.75-3.4 mg/d). Thus, compared to conventional treatment, 3-in-1 blocks seem to provide better initial pain management (especially with movement) as well as a shorter onset of analgesia. However cumulative opioid consumption may not be decreased.

Fascia Iliaca Block (FIB):
Because of their ease of performance, FIBs have been extensively studied in non-operative settings (4 RCTs; combined n=464) [38][39][40][41]. In adult patients with hip fractures, 3 trials have compared FIB to parenteral analgesics (opioids or NSAIDs) with mixed results. Monzon et al. [38] observed that, compared to intravenous NSAIDs, a single-injection FIB (0.3 mL/kg of bupivacaine 0.25%) resulted in a slower analgesic onset: at 15 minutes, the FIB group displayed higher pain levels (6.24 ± 0.17 vs. 2.9 ± 0.16; P<0.001). However, subsequent pain scores were similar at 2 and 8 hours. Similarly, Mouzopoulos et al. [39] found no differences in pain scores when 207 patients were randomized to daily FIB with bupivacaine or normal saline. However the authors did notice a decrease in the incidence (10.8 vs. 23.8%) and duration (5.22 4.28 vs. 10.97 ± 7.16 days) of delirum in the treatment group. In contrast, when comparing single-injection FIB (40 mL of mepivacaine 1%) to intramuscular morphine (0.1 mg/ kg), Foss et al. [40] reported improved static analgesia at 60 and 180 minutes (both P ≤ 0.03) with the former. Furthermore pain relief with a 15-degree leg lift was also superior with FIB at the 3-hour mark (P=0.04). Although subjects in the FIB group required less breakthrough morphine (P<0.01), no intergroup differences were noted in terms of oxygen saturation, sedation, nausea/vomiting and hemodynamic parameters. The analgesic benefits of FIB reported by Foss et al. [40] mirror those of an RCT conducted in pediatric patients. In 2006, Wathen et al. [41] randomized 55 children with proximal, middle or distal femoral fracture to receive FIB (0.25-0.50 mL/kg of ropivacaine 0.5%) or intravenous morphine (0.1 mg/kg). Throughout the 6-hour study period, pain scores were consistently lower in FIB subjects. The latter also experienced a longer analgesic duration (313 vs. 60 minutes) and required less breakthrough analgesia [41]. Thus, due to conflicting results, the current evidence does not permit definitive conclusions as to the analgesic benefits of FIB. Although FIB can decrease the incidence/ duration of delirium in very elderly patients, its routine use has not been shown to impact sedation or oxygen saturation.

Femoral Block (FB):
To date, 4 RCTs (combined n = 218) have investigated the role of FB in the management of lower limb trauma [42][43][44][45]. Haddad and Williams [42] and Mutty et al. [43] compared FB (bupivacaine 0.25-0.5%) to systemic analgesia (opioids/NSAIDs) in patients with extracapsular femoral neck fractures and diaphyseal/ distal femoral fractures, respectively. In both trials, FB resulted in lower pain scores during the first 1.5-2 hours. In a subsequent study, Schiferer et al. [44] randomized 62 patients with lower extremity injury (hip dislocation/fracture, femoral fracture, patellar tendon rupture) to FB (20 mL of levobupivacaine 0.5%) performed at the accident site or intravenous metamizole (1 g). Schiferer et al. [44] found that, unlike the control group, there was a significant decrease in pain and anxiety between the on-site and transport values for patients who received a FB (both P<0.001). Subsequently, the same authors proceeded to compare FB and metimazole for the ambulance transfer of patients with knee trauma [45]. Again, pain and anxiety were only decreased in patients receiving FB. Thus the available evidence suggests that, compared to conventional management, FB provides better pain control for patients suffering from hip and knee trauma.

Technical commentary
Lumbar plexus block: The hip joint is innervated by articular branches derived from the femoral, obturator, accessory obturator, superior gluteal and quadratus femoris nerves [46]. Because it anesthetizes the femoral, obturator and accessory obturator nerves with a single injection, the LPB provides an elegant analgesic option for hip fracture and trauma.
The LPB can be localized using loss of resistance or PNS. In 2011, Danelli et al. [47] set out to compare the 2 modalities. In all subjects, a bolus of 20 mL of mepivacaine 1.5% was injected after obtaining the desired endpoint. Danelli et al. [47] observed that the onset time to complete sensorimotor block was shorter with PNS (12 6 vs. 22 ± 6 minutes; P=0.03). Although more patients achieved complete block at 30 minutes with PNS (80 vs. 40%), this difference failed to reach statistical significance (P=0.113) due to the small sample size (n=30). Danelli et al. 's results seem to suggest that the analgesic difference reported by Chudinov et al. [34] (lower pain scores at 8 and 16 hours in the LPB group) might have been weakened because loss of resistance was employed instead of PNS. Therefore future trials investigating LPB in non-operative settings should rely on PNS to identify the lumbar plexus. Although US is increasingly used for LPBs [48], further investigative comparisons with PNS are required prior to its routine implementation outside of the operating room. Finally, LPBs carry an inherent risk of LA spread to the epidural space (1-16%), renal puncture and psoas hematoma [49]. Thus they should be reserved for experienced operators.

3-in-1 Block:
In 1973, Winnie et al. [50] suggested that LA injection inside the sheath of the femoral nerve, coupled with cephalad angulation of the needle and distal manual compression, would result in rostral migration of LA molecules towards the lumbar plexus. Since the 3 main branches of the latter (femoral, lateral femoral cutaneous and obturator nerves) could be anesthetized with a single injection, this technique was named "3-in-1 block" or "anterior approach to the lumbar plexus".
The unreliable obturator block seen with the 3-in-1 technique stems from the fact that, contrary to Winnie's hypothesis [50], LA anesthetizes the lateral femoral cutaneous and obturator nerves through lateral/ medial spread, deep to the fascia iliaca and not via proximal diffusion [59]. Therefore, with the 3-in-1 block, LA may distribute preferentially in a lateral direction and spare the obturator nerve [54]. Thus, the posterior approach constitutes the only reliable method to anesthetize the lumbar plexus. Terms such as "anterior lumbar plexus block" and "3-in-1 block" are therefore antiquated and should no longer be used.
In summary, since the so-called 3-in-1 blocks dependably anesthetize the femoral nerve, they do provide some analgesic effect in hip/femoral fractures, as evidenced by Fletcher [37] results. However the unreliable obturator nerve block might have led to an inconsistent reduction in breakthrough opioid requirements. For instance, if obturator blockade were achieved, hourly consumption of morphine would be decreased [35]. In contrast, an absent obturator block coupled with strong representation of obturator innervation in the hip joint (obturator and accessory obturator nerves) could explain why Graham et al. [36] and Luger et al. [37] failed to detect decreases in morphine and piritramide requirements, respectively. Thus future trials should compare LPB (or combined femoral-obturator blocks) to isolated femoral nerve block for analgesia after hip fracture. Moreover, obturator blockade should be assessed using motor rather than sensory testing. However a motor evaluation may be somewhat problematic as most patients with fractured hips would be reluctant to adduct the injured limb.

Fascia iliaca block:
In an effort to improve the obturator block seen with the 3-in-1 technique, Dalens et al. [60] introduced the FIB, a method whereby LA is injected immediately dorsal to the fascia iliaca while firm compression is applied distal to the puncture site. In 120 children randomized to a 3-in-1 block or a FIB, Dalens et al. [60] reported similar rates of complete sensory block for the femoral nerve (100%); unfortunately, the clinical test for motor blockade of the obturator nerve (elicitation of adduction at the end of surgery) did not allow for definitive conclusions. Subsequently, the same comparison was carried out in 100 adults by Capdevila et al. [61]. Again, the rates of femoral block were comparable (88-90%). However, motor blockade of the obturator nerve showed no difference (20-32%). Thus, obturator block remains elusive even with FIB.
In 2008, Dolan et al. [62] randomized 80 patients to FIB using loss of resistance or US The latter resulted in a an improved motor block of the obturator (P=0.033) and femoral (P=0.006) nerves. The authors hypothesized that the subcutaneous fascia might in fact consist of several layers separated by adipose tissue: thus blind penetration of any of these layers could have been mistaken for that of the fascia iliaca [62]. Interestingly, in the literature, the 2 RCTs that compared FIB to parenteral analgesia for hip fracture and that found no differences in pain control used an "intramuscular" or a sharp needle to identify the fascia iliaca compartment [38,39]; in contrast, the 2 trials that detected a beneficial effect with FIB employed a blunt needle [40,41]. Thus future trials investigating FIB in non-operative settings should use blunt needles as well as US to reliably detect penetration of the appropriate fascial layers (fascia lata and iliaca).

Femoral block:
In clinical practice, the femoral nerve is most commonly located using PNS or US.
In a dose finding study, the MEV 50 of ropivacaine 0.5% resulting in sensory and motor block of the femoral nerve was lower with US compared to PNS (15 vs. 26 mL; P = 0.002) [63]. A recent RCT compared US to a combination of PNS and US [64,65]. Although both techniques provided similar efficacy, the combined modalities increased the performance time (188 vs. 148 seconds; P=0.01) and number of needle passes (4.1 vs. 1.1; P<0.01).
In the literature, the 4 RCTs comparing FB to conventional treatment for lower extremity trauma have arrived at similar conclusions [42][43][44][45]. This situation seems at odds with the one afflicting 3-in-blocks and FIBs. One potential explanation lies in the predictability of FBs. Unlike their 3-in-1 counterparts, FBs do not aspire to anesthetize the obturator nerve; thus LA spread medially (towards the obturator nerve) or rostrally (towards the roots of the lumbar plexus) becomes a nonissue. More importantly, 3 out the 4 RCTs employed PNS to locate the femoral nerve [43][44][45]. Because muscular contractions were sought, the endpoint was objective. In contrast, a traditional FIB requires the operator to search double fascial click: such tactile recognition could depend on technical experience and the use of a blunt needle. Thus, one lesson that can be learned from the literature pertaining to FB/FIB and transposed to future RCTs as well as other peripheral nerve blocks is that the investigator should select blocks with predictable patterns of LA spread and techniques with reproducible endpoints (PNS, US). This may be especially important in non-operative settings where physicians perform nerve blocks only sporadically.

Conclusion
Despite multiple reports advocating the use of peripheral nerve blocks in non-operative settings, only a handful of RCTs have formally compared nerve blocks to conventional treatment at the accident site, in the ED or on the surgical wards. Unfortunately most of these trials suffered from potential observer bias due to the lack of blinded assessment. Furthermore 36% of available RCTs also failed to provide sample size justification and thus could have been statistically underpowered. More importantly, the true analgesic benefits of some nerve blocks could have been underestimated due to an unreliable technique (transarterial injection for AXB, loss of resistance for LPB, 3-in-1 block, double fascial click for FIB) or suboptimal equipment (sharp needle for FIB). The variable techniques and experience levels of the operators also constitute prohibitive obstacles for the conduct of a metaanalysis. Despite these shortcomings, the available evidence seems to suggest that peripheral nerve blocks can provide pain control for upper and lower limb trauma in non-operative settings. For instance, brachial plexus blocks offer a useful alternative to procedural sedation for fracture manipulation in the ED. Lumbar plexus blocks, 3-in-blocks and FBs can provide analgesia for patients with hip fractures. Femoral blocks also result in more comfortable ambulance transfers to the hospital for patients suffering from hip and knee trauma. Finally, in very elderly subjects, FIBs can decrease the incidence and duration of perioperative delirium.
Increasingly, anesthesiologists, ED physicians and orthopedists will be called upon to perform peripheral nerve blocks in non-operative settings. It is not the goal of this review article to determine which specialties should carry out these nerve blocks. However, for the sake of patient care, it suffices to say that residency programs should provide their trainees with sufficient exposure to basic nerve blocks. Training should encompass technique, selection of LA, post procedural monitoring as well as management of possible complications. To ensure that the learner achieves technical proficiency, strict training and testing guidelines will need to be developed. Anesthesiologists often use a cumulative experience of 50-60 blocks to denote competency [65]; however this technical benchmark remains somewhat arbitrary. Perhaps, similarly to transoesophageal echocardiography, a validated examination administered by a governing body can remedy the situation.
In summary, a critical survey of the available RCTs provides an effective tool to validate the use of peripheral nerve blocks in nonoperative settings. Despite current best evidence, many issues remain unresolved and require further elucidation (Table 3). Further welldesigned and meticulously executed RCTs are warranted. Future trials should provide a clear research hypothesis, sample size justification as well as blinded assessment. Furthermore peripheral nerve blocks should be carried out by operators with adequate training using appropriate equipment and evidenced-based techniques with a proven record of safety and efficacy. Only then can a learned dialogue be undertaken between and within professional societies to determine who should perform nerve blocks in non-operative settings.