Review Article. The analgesic efficacy and safety of neuraxial magnesium sulphate: a quantitative review. Summary

Transcription

1 doi: /j x Review Article The analgesic efficacy and safety of neuraxial sulphate: a quantitative review E. Albrecht, 1 K. R. Kirkham, 1 S. S. Liu 2 and R. Brull 3 1 Fellow, Department of Anaesthesia, Toronto Western Hospital, University of Toronto, Toronto, Canada 2 Chief of Regional Anaesthesia and Perioperative Pain, UW Medicine & Education Endowed Professor of Anaesthesiology, University of Washington, Seattle, WA, USA 3 Associate Professor, Departments of Anaesthesia, Toronto Western Hospital and Women s College Hospital, University of Toronto, Toronto, Canada Summary Eighteen published trials have examined the use of neuraxial as a peri-operative adjunctive analgesic since 2002, with encouraging results. However, concurrent animal studies have reported clinical and histological evidence of neurological complications with similar weight-adjusted doses. The objectives of this quantitative systematic review were to assess both the analgesic efficacy and the safety of neuraxial. Eighteen trials comparing with placebo were identified. The time to first analgesic request increased by 11.1% after intrathecal administration (mean difference: 39.6 min; 95% CI min; p = ), and by 72.2% after administration (mean difference: min; 95% CI min; p = 0.02) with doses of between 50 and 100 mg. Four trials monitored for neurological complications: of the 140 patients included, only a 4-day persistent headache was recorded. Despite promising peri-operative analgesic effect, the risk of neurological complications resulting from neuraxial has not yet been adequately defined.... Correspondence to: Eric Albrecht eric.albrecht74@gmail.com Accepted: 14 August 2012 Magnesium has analgesic properties, primarily related to the regulation of calcium influx into cells [1] and antagonism of N-methyl-D-aspartate (NMDA) receptors in the central nervous system [2, 3]. This analgesic effect was first demonstrated in humans in 1996 when was administered intravenously in the peri-operative period [4]. Since then, a number of additional trials investigating the analgesic efficacy of peri-operative intravenous have been published, with conflicting results [57]. One possible explanation for the reported variability in analgesic efficacy of administered intravenously may be limited passage of this molecule across the blood-brain barrier, which has been demonstrated even in the presence of induced systemic hypermagnesaemia [810]. Consequently, it has been suggested that an intrathecal injection would allow more effective activity at spinal cord NMDA receptors. Indeed, rat models reveal that direct intrathecal administration of enhances the antinociceptive effect of opioids for acute incisional pain [11], and suppresses nociceptive responses in neuropathic pain models [12]. The earliest clinical trials investigating intrathecal and reported an increase in the median duration of [13] and decrease in opioid consumption by 25% [14], respectively. 190 Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland

2 Eric Albrecht et al. Efficacy and safety of neuraxial However, neuraxial administration of is not without risk. Animal studies have reported histological neurotoxicity with weight-adjusted doses similar to those used in most human clinical trials to date [15], whereas two case reports have described patients suffering from disorientation [16] and continuous peri-umbilical burning pain [17] following the injection of into the neuraxis. Therefore, the dual objectives of this quantitative systematic review of the literature were to assess both the analgesic efficacy and the safety of neuraxial in the management of acute postoperative pain, to define better its role in modern clinical practice. Methods The methods of this review followed the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [18]. The authors searched the electronic databases MED- LINE (until January 2012), EMBASE (until January 2012), and the Cochrane Central Register of Controlled Clinical Trials (until January 2012) using the following population search terms: OR compounds. These search results were combined using the Boolean operator AND with the terms peri-operative care OR peri-operative period and focused by combination using AND with the terms OR analgesics OR pain OR pain management OR pain measurement OR pain threshold. The following words were also searched as keywords: *, periop*, peri-op*, perop*, intraop*, intra-op*, postop*, post-op*, analg*, and pain*. Finally, reference lists from the resulting publications were manually searched for any relevant trials not identified by the above search strategy. The resulting list of publications was limited to include only randomised controlled trials, English, French and German language publications, human and adult studies and the use of sulphate in the investigation arm. Only those trials comparing the administration of intrathecal or to placebo were included in the present review. The quality of the method used in each trial included was assessed using the Jadad criteria [19] on a scale from 1 (minimum) to 5 (maximum). Two authors (EA, SA) independently extracted the data and reviewed and scored each trial using this method with differences in extracted data or scoring resolved by discussion with a third author (RB). The trial characteristics recorded included: type of surgery; type of surgical ; route of administration; dose of administered ; and type of postoperative. Specific outcomes sought from each publication were based on the American Society of Regional Anesthesia and Pain Medicine s Acute Postoperative Pain Database initiative [20]. These included the acute pain-related endpoints: (i) onset time of sensory blockade; (ii) regression time of sensory blockade; (iii) onset time of motor blockade; (iv) regression time of motor blockade; (v) time to first analgesic request after neuraxial blockade; (vi) intravenous morphine consumption at 24 h postoperatively; (vii) pain scores at rest and on movement measured at 24 h postoperatively; (viii) early postoperative (06 h) intravenous morphine consumption; (ix) early postoperative (06 h) pain scores at rest and on movement; and incidences of (x) postoperative nausea and vomiting and (xi) pruritus within the first 24 h postoperatively. If not otherwise stated, it was assumed that reported pain scores had been assessed at rest. The safety-related endpoints sought included: (i) temporary and permanent neurological complications defined based on terminology described in the American Anesthesiologists Society Closed Claims database project [21]; as well as (ii) hypotension; (iii) bradycardia; and (iv) sedation. Mean and SD values, for analysed endpoints, were extracted from the text, tables or figures of each source publication. The authors of trials that failed to report the sample size or results as a mean and SD or SEM were contacted to request the missing data. All opioid use was converted to equianalgesic doses of intravenous morphine [2224]. Similarly, pain scores reported as verbal rating or numeric rating scales were converted to a standardised analogue scale to permit quantitative evaluation. Each endpoint was analysed according to the route of administration (intrathecal,, or intrathecal and ) and a meta-analysis was conducted if two or more trials reported the endpoint of interest. Meta-analyses were performed with the assistance of RevMan (version 5.1.6; Copenhagen, The Nordic Cochrane Centre, The Cochrane Collaboration, Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland 191

3 Eric Albrecht et al. Efficacy and safety of neuraxial 2011). This software estimates the weighted mean differences for continuous or categorical data, allowing comparison between groups with an overall estimate of the pooled effect. As most data were heterogeneous, analyses were performed using a random effects model and results are presented as mean difference or relative risk (RR) with 95% CI. I 2 was determined to evaluate heterogeneity using thresholds for low (25%49%), moderate (5074%), and high (> 75%) levels [25]. Finally, the possibility of publication bias was assessed by calculating a funnel plot of standard error of the mean difference (y-axis) as a function of the mean difference (x-axis). A two-sided p-value < 0.05 was considered significant. Results Eighteen trials, including a total of 1145 patients, were included in the analysis (Fig. 1). Twelve publications Included Eligibility Screening Identification Analgesia Analgesics Pain Pain management Pain measurement Pain threshold Analg* as a keyword Pain* as a keyword Magnesium Magnesium Compounds Magnesium* as a keyword AND 333 Citations 156 Citations 153 Abstracts 13 Full text articles 12 Full text articles 18 Full text articles Peri-operative care Peri-operative period Periop*as a keyword Peri-op*as a keyword Perop*as a keyword Intraop*as a keyword Intra-op*as a keyword Postop*as a keyword Post-op*as a keyword RCT Human Did not meet inclusion criteria after abstract review Did not meet inclusion criteria after article review 6 Articles retrieved from scanning bibliographies Figure 1 PRISMA flow diagram showing literature search results. Eighteen randomised controlled trials (RCTs) were ultimately used for the analysis. were identified from the literature search strategy and six from scanning bibliographies. Trial characteristics are presented in Table 1. Sixtyseven per cent of the publications achieved a Jadad score of 4 or 5. Most trials were conducted on patients undergoing orthopaedic and lower limb surgery (8 18; 44.5%) [14, 2632] or gynaecology and obstetric procedures (8 18; 44.5%) [13, 3339]. The two remaining trials were conducted on patients undergoing major abdominal surgery [40] and thoracotomy [41]. Attempts were made to contact eight authors because of incomplete data [13, 14, 28, 30, 31, 36, 38, 41], but none provided the information requested. Out of the 11 trials that examined intrathecal administration, eight administered a dose of 50 mg (73%) [13, 27, 30, 32, 3638, 41] and three a dose of approximately 100 mg (27%) [26, 29, 34]. Magnesium was administered ly as a single 50-mg bolus in three trials (50%) [31, 35, 40], as a 50-mg bolus followed by 24 h infusion in two trials (33%) [14, 33], and as a single bolus of 500 mg in one trial (17%) [39]. Finally, a single trial administered 50 mg intrathecally, followed by an infusion of 100 mg.h )1 for 24 h [28]. The total doses administered in 24 h via infusion were 147 mg [33], 150 mg [14] and 2450 mg [28]. There was significant variability between trials in anaesthetic regimens. For example, in the intrathecal space, was co-administered with levo [26], hyperbaric [27, 32, 36], isobaric [29, 30, 34, 38], lidocaine [37] or opioids without local anaesthetic [13, 41]. Table 2 presents the acute pain-related endpoints. The onset of sensory blockade was slower when was administered in the intrathecal space (mean difference: 2.2 min; 95% CI: min; p = ), whereas the opposite was true following administration (mean difference: )6.9 min; 95% CI: )8.1 to )5.7 min; p < ). With intrathecal, the duration of sensory blockade was longer (mean difference: 23.8 min; 95% CI: p = 0.01), and non-significant trends were noted towards slower onset (mean difference: 1.5 min; 95% CI: )0.3 to 3.3 min; p = 0.10) and longer duration (mean difference: 19.5 min; 95% CI: )1.2 to 40.2 min; p = 0.06) of motor blockade. The time to first analgesic request was increased by 11.1% following intrathecal administration of 192 Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland

4 Eric Albrecht et al. Efficacy and safety of neuraxial Table 1 Trial characteristics. Reference Jadad score Number of patients in group placebo group Surgery Sex Anaesthesia strategy administration of [26] Total hip and total knee arthroplasty Both Spinal [13] Labour Female Spinal + [27] Knee arthroscopy Both Spinal [34] Caesarean delivery Female Spinal [29] Lower extremity surgery Both Spinal [36] Caesarean delivery Female Spinal [41] Posterolateral thoracotomy Both General Neuraxial drugs other than 15 mg levo 8 lg sufentanil 25 lg 6 mg hyperbaric 10 lg 10 mg isobaric 10 lg l morphine 15 mg isobaric 10 mg hyperbaric 25 lg 25 lg lg morphine Bolus dose of Epidural infusion of Total dose in 24 h Postoperative 94.5 mg No 50 mg IV PCA morphine 50 mg No 50 mg Epidural infusion of + adrenaline 50 mg No 50 mg IV + oral tramadol 100 mg No 100 mg IV paracetamol 100 mg No 100 mg IV + oral tramadol 50 mg No 50 mg Intramuscular diclofenac 50 mg No 50 mg IV PCA morphine + IV paracetamol Method of data extraction Comment Graphs Four groups included: control; intrathecal; intrathecal + ; and. Data from the control and intrathecal groups were extracted Text Text tables Graphs Tables Text tables Tables graphs Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland 193

5 Eric Albrecht et al. Efficacy and safety of neuraxial Table 1 Continued. Reference Jadad score Number of patients in group placebo group Surgery Sex Anaesthesia strategy [30] Lower extremity surgery Both Spinal [37] Caesarean delivery Female Spinal [32] Lower abdominal and lower limb surgery Both Spinal [38] Caesarean delivery Female Spinal + Epidural administration of [40] Major abdominal surgery Both General + [14] Total hip arthroplasty Both Spinal + [33] Abdominal hysterectomy Female General + Neuraxial drugs other than Bolus dose of Epidural infusion of Total dose in 24 h Postoperative Method of data extraction Comment 10 mg isobaric 25 lg 50 mg No 50 mg IV tramadol Tables 75 mg lidocaine 50 mg No 50 mg Opioids (not specified) Tables 15 mg hyperbaric 50 mg No 50 mg Not specified Tables 10 mg isobaric 50 mg No 50 mg Epidural bolus of Tables Epidural 30 mg +2mg morphine 50 mg No 50 mg IV morphine Text tables 7.5 mg hyperbaric 0.5% as needed 50 mg 100 mg for 24 h 150 mg Epidural + oral tramadol Text graphs None 50 mg 10 mg.h )1 during surgery + PCEA 1 mg.ml )1 for 24 h 147 mg Epidural bolus of Text tables Three groups included: control; pre-induction; and post-surgery. Data from the control and pre-induction groups were extracted. 194 Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland

6 Eric Albrecht et al. Efficacy and safety of neuraxial Table 1 Continued. Reference Jadad score Number of patients in group placebo group Surgery Sex Anaesthesia strategy [31] Lower abdominal and lower limb surgery Both Epidural [35] Abdominal hysterectomy Female Epidural [39] Caesarean delivery Female Spinal + and administration of [28] Lower limb surgery Both Spinal + IV, intravenous; PCA, patient-controlled ; PCEA, patient-controlled Neuraxial drugs other than Epidural 95 mg Epidural 95 mg 10 mg hyperbaric 25 mg 100 lg 10 mg isobaric 25 lg Bolus dose of Epidural infusion of Total dose in 24 h Postoperative Method of data extraction Comment 50 mg No 50 mg Epidural bolus of Text tables 50 mg No 50 mg Epidural bolus of Text 500 mg No 500 mg IV Text tables 50 mg 100mg.h )1 for 24 h 2450 mg Epidural Tables Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland 195

7 Eric Albrecht et al. Efficacy and safety of neuraxial Table 2 Acute pain-related endpoints. Total number of patients or number of patients with outcome total number of patients (%) p value Outcome Number of trials References Magnesium Placebo RR; 95% CI Mean difference; 95% CI I 2 ; % Heterogeneity Overall effect Onset time of sensory blockade 4 [29], [36], [30], [32] (1.03.4) 89 < Epidural 2 [31], [35] )6.9 ()8.1 to )5.7) < and 1 [28] (2.14.9) < Regression time of sensory bockade 5 [27], [29], [36], [30], [32] ( ) 96 < Epidural 2 [31], [35] ()7.0 to 19.0) and 1 [28] ( ) < Onset time of motor blockade 2 [36], [32] ()0.3 to 3.3) 95 < Epidural and 1 [28] (1.33.3) < Regression time of motor blockade 7 [27], [34], [29], [36], [30], [32] ()1.2 to 40.2) 96 < Epidural 0 and 1 [28] ()4.2 to 9.2) 0.46 Time to first analgesic request after neuraxial blockade; min 6 [27], [34], [29], [36], [30], [37] ( ) 93 < Epidural 4 [40], [31], [35], [39] ( ) 97 < and 1 [28] ( ) < Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland

8 Eric Albrecht et al. Efficacy and safety of neuraxial Table 2 Continued. Total number of patients or number of patients with outcome total number of patients (%) p value Outcome Number of trials References Magnesium Placebo RR; 95% CI Mean difference; 95% CI I 2 ; % Heterogeneity Overall effect Morphine consumption at 24 h postoperatively; mg 4 [26], [27], [29], [41] )6.0 ()10.4 to )1.6) 86 < Epidural 1 [40] )7.2 ()9.8 to )4.6) < and 0 Pain scores at 24 h postoperatively at rest 2 [26], [41] )1.3 ()5.2 to 2.6) Epidural 1 [33] )9 ()11.7 to )6.3) < and 0 Pain scores at 24 h postoperatively on movement 1 [41] )4 ()7.6 to )0.4) 0.03 Epidural 1 [33] )20 ()26.3 to )13.7) < and 0 Early postoperative morphine consumption; mg 1 [41] ()0.3 to 2.3) 0.14 Epidural 0 and 0 Early postoperative pain scores at rest 2 [26], [41] )3.3 ()5.6 to )1.0) Epidural 1 [33] )5.0 ()8.9 to )1.1) 0.01 and 0 Early postoperative pain scores on movement 1 [41] )3.0 ()6.9 to 0.9) 0.13 Epidural 1 [33] )15.0 ()22.9 to )7.2) and 0 Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland 197

9 Eric Albrecht et al. Efficacy and safety of neuraxial Table 2 Continued. Total number of patients or number of patients with outcome total number of patients (%) p value Outcome Number of trials References Magnesium Placebo RR; 95% CI Mean difference; 95% CI I 2 ; % Heterogeneity Overall effect Postoperative nausea and vomiting 9 [26], [27], [34], [29], [36], [41], [30], [37], [38] (26.3%) (15.9%) 1.11 ( ) Epidural 4 [33], [31], [35], [39] (5.6%) and 1 [28] 4 40 (10.0%) (7.2%) 9 40 (22.5%) 0.78 ( ) 0.44 ( ) Pruritus 7 [13], [34], [29], [36], [41], [30], [38] (25.6%) Epidural 3 [14], [33], [39] (7.0%) and 1 [28] (37.5%) (26.9%) (6.0%) (62.5%) 0.88 ( ) 1.15 ( ) 0.6 ( ) Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland

10 Eric Albrecht et al. Efficacy and safety of neuraxial (mean difference: 39.6 min; 95% CI: min; p = ), and by 72.2% after administration (mean difference: min; 95% CI: min; p = 0.02). Cumulative morphine consumption at 24 h postoperatively was 29.4% lower with intrathecal (mean difference: )6.0 mg; 95% CI: )10.4 to )1.6 mg; p = 0.008). There were no differences in other acute pain-related endpoints when trials were grouped for meta-analysis, with the exception of early pain scores at rest following intrathecal (mean difference: )3.3; 95% CI: )5.6 to )1.0; p = 0.005). The funnel plot for each of these outcomes was inverted, symmetric, and centred around the mean difference on the x-axis, indicating a low likelihood for publication bias. Only four trials explicitly described monitoring for signs of temporary (headaches, back and leg pain, temporary nerve injury) or permanent paraplegia, quadriplegia, peripheral nerve injuries) neurological complications [27, 30, 32, 38]. Out of the 140 patients included in these trials, only one complication was recorded: rehospitalisation and conservative management for a 4-day persistent headache [27]. There were no differences in the incidences of hypotension, bradycardia or sedation (Table 3). Of note, 39% (7 18) of trials assessed the incidence of hypotension [30, 31, 34, 3639], whereas only 17% (3 18) assessed the incidence of either bradycardia [30, 31, 38] or sedation [36, 39, 41]. Discussion This is the first meta-analysis to assess the analgesic efficacy and safety of neuraxial administration for postoperative pain management. increased the time to first analgesic request, reduced morphine consumption at 24 h postoperatively and modestly reduced early postoperative pain scores. When administered in the space, also increased the time to first analgesic request. These acute pain-related benefits reportedly occur without increased risk of hypotension, bradycardia or sedation. Although the onset of sensory blockade was statistically delayed with the use of intrathecal, the clinical significance of this result is questionable. Once sensory blockade was established, however, the duration of effect was prolonged. It is difficult to explain this phenomenon given the current understanding of s mechanism of action on the central nervous system. It would seem to imply a more direct impact of on local anaesthetic concentrations or binding within the neuraxis, which has not been previously described. Surprisingly, the opposite effect was observed with administration, resulting in accelerated onset of sensory blockade. The literature does not provide a clear explanation for this observation, but the discrepancy between sites of administration may well be a consequence of the small number of studies available to be included. In addition to the small number of source studies included in the present review, there was significant heterogeneity, which necessarily limits the generalisability of our findings. For example, trials differed with respect to the type of neuraxial local anaesthetic used, type(s) of neuraxial opioid added, mode of administration, and patient population. Furthermore, the lack of consistency in the reporting of acute pain-related outcomes made it impossible to conduct meta-analyses on several of our pre-defined acute pain-related endpoints. The dose of neuraxial that confers optimal with the fewest possible side effects remains unclear. The earliest clinical trials investigating intrathecal selected a dose of 50 mg [13]. This was extrapolated from an animal study where 188 lg of intrathecal potentiated morphine antinociception [11], based on the relative differences between human and rat body weight and cerebrospinal fluid volume. Most of the trials reviewed herein employed doses between 50 and 100 mg, irrespective of whether administered within the intrathecal or space, without providing justification for the doses selected. Only one trial chose an alternate, larger dose of 500 mg for administration. Considering the known differences in pharmacokinetics between intrathecal and local anaesthetics and opioids, it would seem unlikely that the same dose or concentration of would have equal effects in the intrathecal and spaces. The lack of reported complications associated with neuraxial must not be interpreted as an endorsement of its safety, as only four of the 18 trials we reviewed prospectively sought to capture potential neurotoxic side-effects. Animal data have eloquently demonstrated the potential hazards of neuraxial Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland 199

11 Eric Albrecht et al. Efficacy and safety of neuraxial Table 3 Safety-related endpoints. Total number of patients or number of patients with outcome total number of patients (%) p value Outcome Number of trials References Magnesium Placebo RR; 95% CI I 2 ; % Heterogeneity Overall effect Hypotension 5 [34], [36], [30], [37], [38] (21.2%) (26.5%) 0.78 ( ) Epidural 2 [31], [39] (0.3%) (41.3%) 0.78 ( ) and 0 Bradycardia 2 [30], [38] 2 80 (2.5%) 0 80 (0%) 3.00 ( ) Epidural 1 [31] 4 30 (13.3%) 6 30 (20.0%) 0.67 ( ) 0.49 and 0 Sedation 2 [36], [41] (27.1%) (35.6%) 0.49 ( ) Epidural 1 [39] 1 45 (2.2%) 0 45 (0%) 3.00 ( ) 0.50 and 0 administration of at similar weight-based doses to the studies included herein ( mg.kg )1 based on an estimated 70-kg mass). In the only known dose-effect study, intrathecal doses of 0.3 mg.kg )1, 1 mg.kg )1, 2 mg.kg )1 and 3 mg.kg )1 inflicted pain on injection, and produced dose-dependent neurological complications and histological changes in rabbits [15]. Clinically significant motor dysfunction was observed in those rabbits receiving doses of 23 mg.kg )1 and sensory dysfunction increased significantly with increasing doses. These changes were apparent within 1 h of administration and persisted throughout the study period. Furthermore, histological evidence of direct neuronal injury was observed in 22% of animals receiving 1 mg.kg )1 or more. These changes were present 7 days after exposure, but no later observations were taken to determine the long-term consequences of this damage. However, the impact of on neurological structure and function appears inconsistent between species. doses of 4.6 mg.kg )1 and 9.2 mg.kg )1 of given every other day to rats over a 30-day period produced no adverse behavioural consequences [42]. The same animals were examined for histological signs of neurotoxicity, revealing no differences from placebo in the 4.6 mg.kg )1 dose group, but confirming a dose-related effect with moderate vacuolation of spinal grey matter ganglion cells in the 9.2 mg.kg )1 group. In a canine model of spinal cord ischaemia, none of the animals receiving intrathecal doses of 3 mg.kg )1 developed neurological or histological complications following aortic cross-clamping [43]. Taken together, these investigations have not defined the safe dose of intrathecal but have demonstrated that a risk of clinically relevant neurological injury exists. In summary, data on the neuraxial administration of as an analgesic adjunct in the peri-operative setting are promising. However, these results should be interpreted with caution. The lack of neurotoxicity assessment in most clinical trials, combined with current limited animal data, has not yet defined the risk of neurological complications resulting from neuraxial. Additional dose-effect studies would strengthen our understanding of the safety profile of neuraxial before translation into routine clinical practice can be supported. Acknowledgement We are grateful to Dr Saravanan Ankichetty for his assistance in the preparation of this article. 200 Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland

12 Eric Albrecht et al. Efficacy and safety of neuraxial Competing interests EA has received grants from the Swiss Academy for Anaesthesia Research (SACAR), Lausanne, Switzerland and from the Foundation SICPA, Prilly, Switzerland. References 1. Iseri LT, French JH. Magnesium: nature s physiologic calcium blocker. American Heart Journal 1984; 108: Woolf CJ, Thompson SW. The induction and maintenance of central sensitization is dependent on N-methyl-D-aspartic acid receptor activation; implications for the treatment of post-injury pain hypersensitivity states. Pain 1991; 44: Feria M, Abad F, Sanchez A, Abreu P. Magnesium sulphate injected subcutaneously suppresses autotomy in peripherally deafferented rats. Pain 1993; 53: Tramer MR, Schneider J, Marti RA, Rifat K. Role of sulfate in postoperative. Anesthesiology 1996; 84: Jaoua H, Zghidi SM, Wissem L, et al. Effectiveness of intravenous on postoperative pain after abdominal surgery versus placebo: double blind randomized controlled trial. Tunisie Médicale 2010; 88: Tramer MR, Glynn CJ. An evaluation of a single dose of to supplement after ambulatory surgery: randomized controlled trial. Anesthesia and Analgesia 2007; 104: Zarauza R, Saez-Fernandez AN, Iribarren MJ, et al. A comparative study with oral nifedipine, intravenous nimodipine, and sulfate in postoperative. Anesthesia and Analgesia 2000; 91: McKee JA, Brewer RP, Macy GE, Borel CO, Reynolds JD, Warner DS. Magnesium neuroprotection is limited in humans with acute brain injury. Neurocritical Care 2005; 2: Thurnau GR, Kemp DB, Jarvis A. Cerebrospinal fluid levels of in patients with preeclampsia after treatment with intravenous sulfate: a preliminary report. American Journal of Obstetrics and Gynecology 1987; 157: Ko SH, Lim HR, Kim DC, Han YJ, Choe H, Song HS. Magnesium sulfate does not reduce postoperative analgesic requirements. Anesthesiology 2001; 95: Kroin JS, McCarthy RJ, Von Roenn N, Schwab B, Tuman KJ, Ivankovich AD. Magnesium sulfate potentiates morphine antinociception at the spinal level. Anesthesia and Analgesia 2000; 90: Xiao WH, Bennett GJ. Magnesium suppresses neuropathic pain responses in rats via a spinal site of action. Brain Research 1994; 666: Buvanendran A, McCarthy RJ, Kroin JS, Leong W, Perry P, Tuman KJ. prolongs : a prospective, randomized, controlled trial. Anesthesia and Analgesia 2002; 95: Bilir A, Gulec S, Erkan A, Ozcelik A. Epidural reduces postoperative analgesic requirement. British Journal of Anaesthesia 2007; 98: Saeki H, Matsumoto M, Kaneko S, et al. Is intrathecal sulfate safe and protective against ischemic spinal cord injury in rabbits? Anesthesia and Analgesia 2004; 99: Goodman EJ, Haas AJ, Kantor GS. Inadvertent administration of sulfate through the catheter: report and analysis of a drug error. International Journal of Obstetric Anesthesia 2006; 15: Dror A, Henriksen E. Accidental sulfate injection. Anesthesia and Analgesia 1987; 66: Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Medicine 2009; 6: e Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Controlled Clinical Trials 1996; 17: Liu SS, Wu CL, Ballantyne JC, et al. Where in the world is ASRA AcutePOP? Regional Anesthesia and Pain Medicine 2009; 34: Lee LA, Posner KL, Domino KB, Caplan RA, Cheney FW. Injuries associated with regional anesthesia in the 1980s and 1990s: a closed claims analysis. Anesthesiology 2004; 101: Skaer TL. Practice guidelines for transdermal opioids in malignant pain. Drugs 2004; 64: Silvasti M, Svartling N, Pitkanen M, Rosenberg PH. Comparison of intravenous patient-controlled with tramadol versus morphine after microvascular breast reconstruction. European Journal of Anaesthesiology 2000; 17: Dopfmer UR, Schenk MR, Kuscic S, Beck DH, Dopfmer S, Kox WJ. A randomized controlled double-blind trial comparing piritramide and morphine for after hysterectomy. European Journal of Anaesthesiology 2001; 18: Higgins JP, Thompson SG. Quantifying heterogeneity in a metaanalysis. Statistics in Medicine 2002; 21: Arcioni R, Palmisani S, Tigano S, et al. Combined intrathecal and sulfate supplementation of spinal anesthesia to reduce post-operative analgesic requirements: a prospective, randomized, double-blind, controlled trial in patients undergoing major orthopedic surgery. Acta Anaesthesiologica Scandinavica 2007; 51: Dayioglu H, Baykara ZN, Salbes A, Solak M, Toker K. Effects of adding to and for spinal anesthesia in knee arthroscopy. Journal of Anesthesia 2009; 23: El-Kerdawy H. Analgesic requirements for patients undergoing lower extremity orthopedic surgery the effect of combined spinal and. Middle East Journal of Anesthesiology 2008; 19: Khalili G, Janghorbani M, Sajedi P, Ahmadi G. Effects of adjunct intrathecal sulfate to for spinal anesthesia: a randomized, double-blind trial in patients undergoing lower extremity surgery. Journal of Anesthesia 2011; 25: Ozalevli M, Cetin TO, Unlugenc H, Guler T, Isik G. The effect of adding intrathecal sulphate to - spinal. Acta Anaesthesiologica Scandinavica 2005; 49: Ghatak T, Chandra G, Malik A, Singh D, Bhatia VK. Evaluation of the effect of sulphate vs. clonidine as adjunct to. Indian Journal of Anaesthesia 2010; 54: Shukla D, Verma A, Agarwal A, Pandey HD, Tyagi C. Comparative study of intrathecal dexmedetomidine with intrathecal sulfate used as adjuvants to. Journal of Anaesthesiology Clinical Pharmacology 2011; 27: Farouk S. Pre-incisional provides preemptive and preventive in patients undergoing abdominal hysterectomy. British Journal of Anaesthesia 2008; 101: Ghrab BE, Maatoug M, Kallel N, et al. Does combination of intrathecal sulfate and morphine improve postcae- Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland 201

13 Eric Albrecht et al. Efficacy and safety of neuraxial sarean section? Annales Françaises d Anesthésie et de Réanimation 2009; 28: Hussien N. A comparative study between sulphate and clonidine as adjuvants to anesthesia in patients undergoing abdominal hysterectomies. Ain Shams Journal of Anesthesiology 2011; 4: Malleeswaran S, Panda N, Mathew P, Bagga R. A randomised study of sulphate as an adjuvant to intrathecal in patients with mild preeclampsia undergoing caesarean section. International Journal of Obstetric Anesthesia 2010; 19: Shoeibi G, Sadegi M, Firozian A, Tabassomi F. The additional effect of sulfate to lidocaine in spinal anesthesia for cesarean section. International Journal of Pharmacology 2007; 3: Unlugenc H, Ozalevli M, Gunduz M, et al. Comparison of intrathecal,, or placebo combined with 0.5% for parturients undergoing elective cesarean delivery. Acta Anaesthesiologica Scandinavica 2009; 53: Yousef AA, Amr YM. The effect of adding sulphate to and in elective caesarean section using combined spinal- : a prospective double blind randomised study. International Journal of Obstetric Anesthesia 2010; 19: Abdel-halim J. The effect of preoperative single shot dose of sulphate or dexamethesone as adjuvants to local anesthesia. Ain Shams Journal of Anesthesiology 2011; 4: Ouerghi S, Fnaeich F, Frikha N, et al. The effect of adding intrathecal sulphate to morphine- spinal after thoracic surgery. A prospective, double-blind, placebo-controlled research study. Annales Françaises d Anesthésie et de Réanimation 2011; 30: Chanimov M, Cohen ML, Grinspun Y, et al. Neurotoxicity after spinal induced by serial intrathecal injections of sulphate. An experimental study in a rat model. Anaesthesia 1997; 52: Simpson JI, Eide TR, Schiff GA, et al. sulfate protects the spinal cord from ischemic injury during thoracic aortic cross-clamping. Anesthesiology 1994; 81: Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland