Eur J Anaesthesiol 2020; 37:212–223

ORIGINAL ARTICLE

The effect of intravenous magnesium on postoperative

morphine consumption in noncardiac surgery
A systematic review and meta-analysis with trial sequential
analysis
Ka T. Ng, Jasmine L.L. Yap, Ilyanna N. Izham, Wan Y. Teoh, Pei E. Kwok and Wen J. Koh
Downloaded from http://journals.lww.com/ejanaesthesiology by BhDMf5ePHKbH4TTImqenVHwXmMsAVI5gUpuKEr35HA4OiW6f4r944XUYoZy2St0JULvh+Mv10s0= on 02/22/2020

BACKGROUND Several studies suggest that systemic mag- P < 0.001, I2 ¼ 92%, level of evidence low). The trial sequen-
nesium reduces postoperative opioid consumption and the tial analysis for the effect of magnesium on postoperative
intensity of pain, but others report conflicting results. The morphine consumption was conclusive. Patients who
efficacy and safety profile of intravenous magnesium in received magnesium had a longer time to the first analgesia
noncardiac surgery remain uncertain. request [143 (103 to 183) min, P < 0.001, I2 ¼ 99%, level of
evidence low] and a lower incidence of shivering [0.26 (0.15
OBJECTIVES The aim of this review was to investigate the
to 0.44), P < 0.001, I2 ¼ 35%, level of evidence very low].
effect of intravenous magnesium on the consumption of
However, no significance differences were demonstrated in
postoperative morphine in the first 24 h in adults undergoing
postoperative pain scores in the first 24 h (mean difference,
noncardiac surgery.
95% CI) 0.30 (0.69 to 0.09, P ¼ 0.13, I2 ¼ 91%, level of
DESIGN Systematic review and meta-analysis with trial evidence low), bradycardia (odds ratio, 95% CI) 1.13 (0.43
sequential analysis. to 2.98, P ¼ 0.80, I2 ¼ 35%, level of evidence very low) and
postoperative nausea and vomiting (odds ratio, 95% CI)
DATA SOURCES MEDLINE, EMBASE, CENTRAL from
0.90 (0.67 to 1.22, P ¼ 0.49, I2 ¼ 25%, level of evidence
their inception until January 2019.
moderate).
ELIGIBILITY CRITERIA All randomised clinical trials com-
CONCLUSION The current meta-analysis demonstrates that
paring intravenous magnesium versus placebo in noncardiac
the use of intravenous magnesium as part of multimodal
surgery were systematically searched in the databases.
analgesia may reduce morphine consumption in the first 24 h
Observational studies, case reports, case series and non-
after surgery and delay the time to the first request for
systematic reviews were excluded.
analgesia in patients undergoing noncardiac surgery. How-
RESULTS Fifty-one trials (n¼3311) were included for quan- ever, the included studies were of low-quality with substantial
titative meta-analysis. In comparison with placebo, postop- heterogeneity.
erative morphine consumption at 24-h was significantly
TRIAL REGISTRATION CRD42018086846.
reduced in the magnesium group, with a mean difference
[95% confidence interval (CI)] of 5.6 mg (7.54 to 3.66, Published online 23 January 2020

Introduction
The magnesium ion is essential for the maintenance of body not routinely measured. Pre-operative fasting and intrave-
haemostasis.1 The incidence of hypomagnesaemia varies nous maintenance fluids lacking magnesium supplementa-
from 12% in hospitalised patients to as high as 65% in ICU tion may contribute to hypomagnesaemia.2
patients.2,3 Hypomagnesaemia is often underdiagnosed Magnesium was first introduced as a potential anaesthetic
amongst surgical patients since the serum magnesium is induction agent in the early 1900s but it failed in this role

From the Department of Anaesthesiology, University of Malaya (KTN), Department of Medicine, International Medical University, Kuala Lumpur, Malaysia (JLLY, INI, PEK),
Department of Medicine, University of Liverpool, Liverpool, UK (WYT) and Department of Dental Health, International Medical University, Kuala Lumpur, Malaysia (WJK)
Correspondence to Dr Ka T. Ng, Department of Anaesthesiology, Faculty of Medicine, University of Malaya, Jalan Universiti, 50603 Kuala Lumpur, Malaysia
E-mail: [email protected]

0265-0215 Copyright ß 2020 European Society of Anaesthesiology. All rights reserved. DOI:10.1097/EJA.0000000000001164

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

Systemic magnesium reduces postoperative morphine consumption 213

due its poor permeability through the blood–brain bar- The review questions were formulated using a Popula-
rier.4 However, the ability of magnesium to antagonise tion, Intervention, Comparison and Outcomes approach
the N-methyl-D-aspartate glutamate receptor has been (Supplementary Table S1, http://links.lww.com/EJA/
demonstrated in both animal and human studies.5–8 It is A272).
believed that magnesium inhibits central sensitisation of
pain perception by acting on the hippocampal presynap- Important changes made to the protocol after
tic calcium channels.4 Surgical procedures induce a sys- registration
temic inflammatory response by increasing sympathetic Prior to the preliminary search, our initial protocol was
activity, which imposes a metabolic stress on the body.9 mainly designed to investigate the potential neurologic
Magnesium has been shown to inhibit catecholamine effect of magnesium for the reduction of delirium or
release and to maintain haemodynamic stability in agitation in surgical patients. However, after the comple-
patients undergoing tracheal intubation.10,11 Over time, tion of our preliminary search, our team found insufficient
magnesium has been used for other therapeutic treat- data for a systematic review and meta-analysis on the
ments, namely for neuroprotection and seizure prophy- effect of magnesium on delirium or agitation. Therefore,
laxis in pre-eclampsia, as an antiarrhythmic for Torsades we decided to update the evidence of magnesium in
de Pointes, and as a bronchodilator in acute asthma.4 minimising the use of morphine consumption as there
were more than 20 additional trials published since the
In recent years, there has been substantial interest in the
last meta-analyses12–14 on this topic in 2013.
antinociceptive effect of magnesium to minimise opioid
consumption, while at the same time improving pain
Search strategy
control, in patients undergoing surgery. Magnesium is
The databases of MEDLINE, EMBASE and CENTRAL
a cheap and relatively safe drug and, as such, is more
were systematically searched from its inception until Jan-
likely to be cost-effective if its use after surgery is
uary 2019. Two trial registries (the ClinicalTrials.gov and
associated with reduced consumption of opioids and their
the World Health Organisation International Clinical
associated adverse effects. Several reviews12–14 pub-
Trials Registry Platform, International Standard Random-
lished in 2013 demonstrated that magnesium reduces
ised Controlled Trial Number Registry) were searched for
postoperative pain perception and the postoperative con-
any ongoing or unpublished studies. The search terms and
sumption of opioids in patients undergoing surgery.
strategy are outlined in the Supplementary Table S2,
However, since then many new randomised controlled
http://links.lww.com/EJA/A272.
trials15–20 (RCTs) have been published with contradic-
tory findings. Thus, a systematic review and meta-analy-
Eligibility criteria
sis of the data, with trial sequential analysis, is warranted
All RCTs comparing prophylactic intravenous magne-
to summarise the current evidence for the use of magne-
sium (pre-operative or intra-operative) and placebo in
sium to minimise morphine consumption and improve
adult patients undergoing any type of noncardiac surgery
pain control in surgical patients.
were included in this review, regardless of the type of
We hypothesised that the co-administration of intrave- postoperative analgesia and reported outcomes. We
nous magnesium, as part of a multimodal analgesic excluded cardiac studies as all these investigators exam-
regime, would reduce the cumulative dose of morphine ined mainly the systemic effects of magnesium for the
required for postoperative pain relief in noncardiac sur- reduction of postoperative arrhythmias. There were no
gery. The primary aim of this review and meta-analysis restrictions with regard to the duration of the study
was to investigate the effect of intravenous magnesium follow-up period. All prospective or retrospective obser-
on postoperative morphine consumption in the first 24 h vational studies, case–control studies, case series, case
in adults undergoing noncardiac surgery. Secondary aims reports and editorials or trials published as conference
were to examine the effects of intravenous magnesium on abstract were excluded. Articles not written in the
the time to the first request for analgesia after surgery, the English language were included if the journal provided
postoperative pain scores in the first 24 h, and the inci- an English-translation. The bibliographies of all included
dences of postoperative shivering, postoperative nausea RCTs and relevant systematic reviews were manually
and vomiting, and bradycardia. searched for additional articles. Two attempts were made
to contact the authors of articles identified in the search
Methods strategy if there was incomplete data. The primary out-
Protocol and registration come was the cumulative consumption of intravenous
The current review article was conducted and reported morphine in the first 24 h postoperatively. The secondary
according to the Cochrane Handbook of Systematic outcomes were time to request first analgesia, postopera-
Reviews of Interventions21 and the Preferred Reporting tive pain score in the first 24 h, postoperative serum
Items for Systematic Reviews and Meta-Analyses State- magnesium concentration, and the incidences of postop-
ment,22 respectively. Prior to the search, the study pro- erative nausea and vomiting, shivering and adverse
tocol was published on PROSPERO, CRD42018086846. effects of magnesium (incidence of bradycardia).

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
214 Ng et al.

Study selection, data items, data collection and mean SD.27 Sensitivity analyses were performed by
assessment of validity analysing only studies with a low risk of bias.
Titles and abstracts were screened for eligibility criteria
by two authors (PEK and INI) independently. All
Trial sequential analysis
screened studies were coded with ‘yes’, ‘no’ or ‘maybe’
Trial sequential analysis was conducted to assess the
by the two authors (PEK and INI) independently. Stud-
statistical reliability of the primary outcome in the cumu-
ies coded with ‘no’ were excluded. For studies coded
lative meta-analysis using the trial sequential analysis
with ‘yes’, full text articles were retrieved and screened
viewer version 0.9.5.5 Beta (Copenhagen Trial Unit,
independently by both authors (PEK and INI). The fate
2016).28 In the trial sequential analysis, the required
of articles coded with ‘maybe’ was resolved by consulting
information size and the adjusted significance threshold
a third author (KTN). The final selection of all included
for the consumption of morphine in the first 24 h post-
RCTs was discussed amongst all authors to achieve
operatively were calculated, with an anticipated clinically
a consensus.
relevant mean difference of 5 mg and low-risk bias
adjusted variance of 13.4 mg with 5% risk of type 1 error,
Risk of bias in individual studies, summary of findings
power of 80% and model variance-based heterogeneity
and GRADE assessment of evidence
correction.
All the included RCTs were assessed for risk of bias using
the Cochrane Collaboration Risk of Bias Assessment
Results
Tool by two authors (PEK and INI) independently.21
The study selection process is outlined in the PRISMA
Any conflicts were discussed with a third author (KTN)
flow chart (Fig. 1). The search strategy yielded 3686
until a consensus was achieved. The summary of findings
articles for titles and abstracts screening, of which 76
and the assessment on the certainty of evidence were
articles were retrieved for full text screening. Applying
independently performed by two authors (PEK and INI)
inclusion and exclusion criteria, 52 RCTs with a total of
using the GRADEpro/GDT software (McMaster Univer-
3341 patients were included for qualitative analysis in our
sity and Evidence Prime Inc, Ontario, Canada).23 In
review. The clinical characteristics of excluded studies
accordance with the Cochrane Handbook,21 the quality
are shown in the Supplementary Table S3, http://
of evidence in each measured outcome was assessed
links.lww.com/EJA/A272. Of the 52 RCTs, one article29
based on five criteria (risk of bias, inconsistency, indi-
did not report any of our measured outcome, leaving 51
rectness, imprecision and publication bias). We down-
RCTs with a total of 3311 patients for quantitative meta-
graded a starting rating of ‘high-quality’ evidence of RCT
analysis. Searching clinical trial registries identified three
by one level for serious concern or by two levels for very
ongoing studies (Supplementary Table S4, http://
serious concerns. Any disagreements were resolved by a
links.lww.com/EJA/A272).
third author (KTN).
The clinical characteristics of all included RCTs are
Summary measures and synthesis of results shown in Table 1. The majority of them were based
The Review Manager version 5.3 (The Cochrane Col- on general anaesthesia and eleven15,17,30–38 were of
laboration, Copenhagen, Denmark) was used for statisti- regional anaesthesia. Among all the included RCTs, 28
cal analyses.24 A two-sided P value of less than 0.05 was of them7,17,19,20,29,31,34,36,39–58 administered intravenous
denoted as statistical significance. Dichotomous findings magnesium before the induction of anaesthesia, another
were reported as odds ratio (OR) whilst continuous find- 218,15,16,18,30,32,35,37,38,59–69 administered magnesium
ings were reported as mean difference, with 95% confi- after the induction of anaesthesia and only three33,70,71
dence interval (CI). Funnel plots were performed for any gave magnesium during the induction of anaesthesia.
measured outcomes of more than 10 included RCTs to Thirty-one RCTs7,16,18,20,29,31–34,36,38–40,47,49–
assess risk of publication bias. The I2 test was used to 55,57,58,60,63–65,69,70,72
administered a bolus of intravenous
assess the heterogeneity of studies, where I2 less than magnesium followed by an infusion of magnesium, 16
40%, 40 to 60% and more than 60% was used to deter- RCTs8,17,19,30,35,37,41,42,44–46,48,59,61,62,71 adopted the strat-
mine low, moderate and substantial heterogeneity, egy of magnesium infusion only, and five RCTs43,56,66–68
respectively.25,26 A fixed-effect model (Mantel–Haens- gave a single bolus of magnesium only. The bolus dose of
zel method) was used to pool estimates of all measured magnesium used varied across all the included studies,
outcomes. If substantial heterogeneity (I2 > 60%) was ranging from 1400 to 4000 mg in an adult of 70 kg. In the
noted, a random-effect model (DerSimonian–Laird overall risk of bias assessment, 46 RCTs7,8,15–18,29–35,38–
48,50–70,72
method) was used. For measured outcomes with zero were rated as high/unclear, with an exception of
events in either arms, we followed the guidance of the six RCTs19,20,36,37,49,71 assessed as low (Supplementary
Cochrane Handbook (16.9.3) by using an OR-based Table S5, http://links.lww.com/EJA/A272). Data analyses
method as it excludes reporting bias whether or not of the primary and secondary outcomes are demonstrated
the studies are published.21 When the reported values in Table 2. The summary of findings/certainty of evi-
were presented as median [IQR] they were converted to dence is shown in Table 3.

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Systemic magnesium reduces postoperative morphine consumption 215

Fig. 1

MEDLINE Cochrane database of systematic reviews EMBASE

392 Citations 324 Citations 2970 Citations

3686 citations screened 448 double-hits removed

3238 articles for title and

abstract screening

Inclusion/Exclusion 3162 articles excluded after

Criteria Applied title/abstract screen

76 Articles retrieved
for full text screening

Inclusion/exclusion 24 articles excluded after

criteria applied full text screen

52 Articles included in
qualitative analysis

1 Article- no measured
outcomes reported

51 articles included in
Quantitative analysis

PRISMA flow diagram. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Consumption of morphine in the first 24 h of morphine consumption in the first 24 h postopera-

postoperatively (mg) tively: mean difference (95% CI) 5.6 mg (7.54 to
The median [IQR] consumption of morphine in the first 3.66), P < 0.001, quality of evidence low, trial sequen-
24 h after surgery was 17.5 [6.9 to 40.3] mg in the magne- tial analysis was conclusive. The heterogeneity across all
sium group and 24.8 [11.3 to 45.7] mg in the placebo trials was substantial (I2 ¼ 92%). In the trial sequential
group. Based on the 18 RCTs with appropriate data,15– analysis, the data exceeded the required information size
20,30,31,39,43,50,51,54,55,57,58,65,70
(899 patients), the adjunc- of 288, indicating the benefits of magnesium to minimise
tive use of magnesium significantly reduced the amount the consumption of morphine in the first 24 h after

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Table 1 Clinical characteristics of included studies

Anaesthesia Regime of Mg bolus Mg infusion Postop Overall risk Sample

Author Year strategy Timing of Mg given Mg given (mg kgS1) (mg kgS1 hS1) Control analgesia of bias Type of surgery Country size
58 1
Wilder-Smith 1997 General anaesthesia Before anaesthesia induction Bolus þ infusion 200 mg 200 mg h N.S. PCA morphine High risk Hysterectomy Switzerland 24
216 Ng et al.

Koinig69 1998 General anaesthesia After anaesthesia induction Bolus þ infusion 50 mg kg1 8 mg kg1 h1 N.S. IV fentanyl High risk Arthroscopic knee Austria 46
surgery
Zarauza57 2000 General anaesthesia Before anaesthesia induction Bolus þ infusion 30 mg kg1 10 mg kg1 h1 N.S. PCA morphine High risk Colorectal surgery Spain 92
NSAIDs
IV paracetamol
IV metamizole
Schulz-Stübner68 2001 General anaesthesia After anaesthesia induction Bolus 50 mg kg1 N/A N.S. IV metamizol Unclear risk Pars plana vitrectomy Germany 50
IV nalbuphine
Levaux56 2003 General anaesthesia Before anaesthesia induction Bolus 50 mg kg1 N/A N.S. PCA piritramide Unclear risk Lumbar orthopaedics Belgium 24
surgery
Mavrommati72 2004 General anaesthesia After anaesthesia induction Bolus þ infusion 30 mg kg1 6 mg kg1 h1 N.S. IV fentanyl High risk Abdominal Greece 42
hernioplasty
55 1 1 1
Bhatia 2004 General anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg 15 mg kg h N.S. IV morphine Unclear risk Open India 50
cholecystectomy
Ayoglu7 2005 General anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg1 8 mg kg1 h1 N.S. IV alfentanil Unclear risk Abdominal surgery Europe 40
Seyhan54 2006 General anaesthesia Before anaesthesia induction Bolus þ infusion 40 mg kg1 10 mg kg1 h1 N.S. PCA morphine Unclear risk Total abdominal Turkey 60
20 mg kg1 h1 hysterectomy
Tauzin-Fin67 2006 General anaesthesia After anaesthesia induction Bolus 50 mg kg1 N/A N.S. PCA tramadol Unclear risk Radical prostatectomy France 30
PCA droperidol
53 1 1 1
Cizmeci 2007 General anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg 8 mg kg h N.S. IV meperidine Unclear risk Septorhinoplasty Turkey 60
Tramer66 2007 General anaesthesia After anaesthesia induction Bolus 4g N/A N.S. Oral/Rectal NSAID Unclear risk Ambulatory ilioinguinal Switzerland 200

Eur J Anaesthesiol 2020; 37:212–223

IV/IM opioid hernia repair/
varicose vein
operation
Mentes8 2008 General anaesthesia After anaesthesia induction Infusion N/A 50 mg kg1 h1 N.S. PCA tramadol Unclear risk Laparoscopic Turkey 83
cholecystectomy
Benhaj Amor70 2008 General anaesthesia During anaesthesia induction Bolus þ infusion 50 mg kg1 500 mg h1 N.S. PCA Unclear risk Open French 48
Morphine cholecystectomy,
gastrojejunal
surgery
Ryu52 2008 General anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg1 15 mg kg1 h1 N.S. PCA mixture, Unclear risk Abdominal South Korea 50
ketorolac and hysterectomy
morphine
Oguzhan65 2008 General anaesthesia After anaesthesia induction Bolus þ infusion 30 mg kg1 10 mg kg1 h1 N.S. PCA morphine High risk Lumbar orthopaedic Turkey 50
surgery
51 1 1
Kaya 2009 General anaesthesia Before anaesthesia induction Bolus þ infusion 30 mg kg 500 mg h N.S. PCA morphine Unclear risk Abdominal Turkey 40
hysterectomy
Dabbagh30 2009 Regional anaesthesia After anaesthesia induction Infusion N/A 8 mg kg1 h1 N.S. IV morphine Unclear risk Lower limb Iran 60
orthopaedic surgery
Hwang38 2010 Regional anaesthesia After anaesthesia induction Bolus þ infusion 50 mg kg1 15 mg kg1 h1 N.S. PCA mixture, Unclear risk Total hip arthroplasty South Korea 40
ketorolac and
morphine
Jaoua39 2010 General anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg1 10 mg kg1 h1 N.S. PCA morphine Unclear risk Gastrointestinal Tunisia 38
surgery
40 1 1 1
Saadawy 2010 General anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg 25 mg kg h N.S. PCA morphine. The Unclear risk Laparoscopic Egypt 80
actual dose was not cholecystectomy
provided
Gozdemir37 2010 Regional anaesthesia After anaesthesia induction Infusion N/A 80 mg kg1 h1 N.S. IV pethidine Low risk Transurethral Turkey 60
2 g h1 prostatectomy
64 1
Ko 2011 General anaesthesia After anaesthesia induction Bolus þ infusion 50 mg kg 15 mg kg1 h1 N.S. IV lidocaine Unclear risk Abdominal South Korea 58
PCA fentanyl hysterectomy
63 1 1 1
Song 2011 General anaesthesia After anaesthesia induction Bolus þ infusion 30 mg kg 10 mg kg h N.S. Tab Ultracet High risk Thyroidectomy South Korea 84
IM tramadol
Kiran41 2011 General anaesthesia Before anaesthesia induction Infusion N/A 50 mg kg1 h1 N.S. IV pethidine Unclear risk Inguinal surgery India 100
IM diclofenac
Olgun50 2012 General anaesthesia Before anaesthesia induction Bolus þ infusion 40 mg kg1 10 mg kg1 h1 N.S. PCA morphine Unclear risk Laparoscopic Turkey 60
cholecystectomy
62 1 1
Khafagy 2012 General anaesthesia After anaesthesia induction Infusion N/A 8 mg kg h N.S. IV meperidine Unclear risk Open Egypt 60
cholecystectomy
Kocman42 2013 General anaesthesia Before anaesthesia induction Infusion N/A 5 mg kg1 h1 N.S. IV metamizol Unclear risk Laparoscopic Croatia 60
IV diclofenac cholecystectomy
IV tramadol

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

Table 1 (continued )

Anaesthesia Regime of Mg bolus Mg infusion Postop Overall risk Sample

Author Year strategy Timing of Mg given Mg given (mg kgS1) (mg kgS1 hS1) Control analgesia of bias Type of surgery Country size
31 1 1 1
Kumar 2013 Regional anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg 10 mg kg h N.S. IV morphine Unclear risk Elective subumbilical India 60
surgery
49 1 1 1
De Oliveira 2013 General anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg 15 mg kg h N.S. IV hydromorphone Low risk Segmental USA 46
mastectomy
Ibrahim36 2014 Regional anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg1 2 mg kg1 h1 N.S. IV meperidine Low risk Lower extraperitoneal Egypt 80
and lower limb
surgery
Kahraman35 2014 Regional anaesthesia After anaesthesia induction Infusion N/A 65 mg kg1 h1 N.S. IV tramadol High risk Abdominal Turkey 38
hysterectomy
Mireskandari43 2015 General anaesthesia Before anaesthesia induction Bolus 50 mg kg1 N/A N.S. PCA morphine Unclear risk Caesarean section Iran 50
Frassanito15 2015 Regional anaesthesia After anaesthesia induction Bolus þ infusion 40 mg kg1 10 mg kg1 h1 N.S. IV paracetamol Unclear risk Total knee amputation Italy 40
IV ketorolac
PCA morphine
Taheri44 2015 General anaesthesia Before anaesthesia induction Infusion N/A 50 mg kg1 h1 N.S. IV pethidine Unclear risk Abdominal Iran 40
hysterectomy
Kim48 2015 General anaesthesia Before anaesthesia induction Infusion N/A 50 mg kg1 h1 N.S. IV ketorolac High risk Endoscopic South Korea 60
IV tramadol submucosal
dissection of gastric
neoplasm
Asadollah29 2015 General anaesthesia Before anaesthesia induction Bolus þ infusion 50 mg kg1 8 mg kg1 h1 N.S. IV pethidine High risk Hysterectomy/ Iran 30
Myomectomy
32 1 1
Shah 2016 Regional anaesthesia After anaesthesia induction Bolus þ infusion 250 mg 20 mg kg h N.S. IV tramadol Unclear risk Lower abdominal and India 108
lower limb surgery
61 1 1
Demiroglu 2016 General anaesthesia After anaesthesia induction Infusion N/A 50 mg kg h N.S. IM diclofenac Unclear risk Lumbar disc surgery Turkey 50
PCA tramadol
Ryu60 2016 General anaesthesia After anaesthesia induction Bolus þ infusion 50 mg kg1 15 mg kg1 h1 N.S. PCA fentanyl Unclear risk Laparoscopic South Korea 74
gastrectomy
Vicković47 2016 General anaesthesia Before anaesthesia induction Bolus þ infusion 30 mg kg1 10 mg kg1 h1 N.S. IV fentanyl Unclear risk Abdominal, Serbia 100
orthopaedic,
urology surgery
Muthiah17 2016 Regional anaesthesia Before anaesthesia induction Infusion N/A 150 mg h1 N.S. IV morphine Unclear risk Arthroscopic anterior India 40
cruciate ligament
repair
Shin33 2016 Regional anaesthesia During anaesthesia induction Bolus þ infusion 50 mg kg1 15 mg kg1 h1 N.S. PCA fentanyl Unclear risk Bilateral total knee South Korea 44
IV ketoprofen amputation
16 1 1 1
Sousa 2016 General anaesthesia After anaesthesia induction Bolus þ infusion 20 mg kg 2 mg kg h N.S. PCA morphine Unclear risk Laparoscopic Brazil 36
gynaecological
surgery
Gucyetmez59 2016 General anaesthesia After anaesthesia induction Infusion N/A 50 mg kg1 h1 N.S. IV tramadol Unclear risk Liver transplant Turkey 70
Shal34 2017 Regional anaesthesia Before anaesthesia induction Bolus þ infusion 30 mg kg1 10 mg kg1 h1 N.S. IV pethidine Unclear risk Arthroscopic knee Egypt 70
surgery
Sohn18 2017 General anaesthesia After anaesthesia induction Bolus þ infusion 50 mg kg1 15 mg kg1 h1 N.S. IV morphine High risk Thoracoscopic South Korea 62
IV fentanyl surgery
IV NSAIDs
Elsersy46 2017 General anaesthesia Before anaesthesia induction Infusion N/A 30 mg kg1 h1 N.S. IV pethidine Unclear risk Functional endoscopy Egypt 294
surgery
71 1 1
Haryalchi 2017 General anaesthesia During anaesthesia induction Infusion N/A 15 mg kg h N.S. IV fentanyl Low risk Total abdominal Iran 40
hysterectomy
Walia45 2018 General anaesthesia Before anaesthesia induction Infusion N/A 30 mg kg1 h1 N.S. IV propofol High risk Elective surgery India 80
Kizilcik20 2018 General anaesthesia Before anaesthesia induction Bolus þ infusion 30 mg kg1 20 mg kg1 h1 N.S. PCA morphine Low risk Sleeve gastrectomy Turkey 80
IV fentanyl
Ahmed19 2018 General anaesthesia Before anaesthesia induction Infusion N/A 50 mg kg1 h1 N.S. PCA morphine Low risk Thoracic surgery Egypt 60

IM, intramuscular; Mg, magnesium; N.S., normal saline; N/A, nonapplicable; PCA, patient-control analgesia.

Eur J Anaesthesiol 2020; 37:212–223

Systemic magnesium reduces postoperative morphine consumption 217

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

218 Ng et al.

Table 2 Data analysis of primary and secondary outcomes

Outcomes Trials n I2 (%) MD/OR (95% CI) P

1 Postoperative morphine consumption (mg) 18 899 92 5.6 (7.54 to 3.66) <0.001
Postoperative morphine consumption –
subgroup analysis by method of
administration
Bolus only 1 50 – 2.84 (3.77 to 1.91) <0.001
Infusion only 3 180 84 3.37 (6.20 to 0.54) 0.02
Bolus followed by infusion 15 715 93 6.31 (8.76 to 3.86) <0.001
Sensitivity analysis – only low risk of bias 3 210 42 3.65 (4.39 to 2.91) <0.001
2 Time to first analgesia (min) 11 824 99 143 (103 to 183) <0.001
Time to first analgesia – subgroup analysis by
method of administration
Bolus only 2 230 55 50 (88 to 187) 0.48
Infusion only 3 190 100 378 (217 to 973) 0.21
Bolus followed by infusion 6 404 98 48 (20 to 77) <0.001
Sensitivity analysis – only low risk of bias 1 46 – 13 (9 to 17) <0.001
3 Postoperative pain score (visual analogue pain 18 1232 91 0.30 (0.69 to 0.09) 0.13
score (0 to 10)
Postoperative pain score – subgroup analysis
by method of administration
Bolus only 1 50 – 0.10 (0.36 to 0.16) 0.45
Infusion only 8 681 86 0.38 (0.84 to 0.09) 0.11
Bolus followed by infusion 9 501 90 0.28 (1.01 to 0.46) 0.47
Sensitivity analysis – only low risk of bias 1 40 – 0.05 (0.41 to 0.51) 0.83
4 Postoperative nausea and vomiting 20 1214 25 1.01 (0.75 to 1.37) 0.93
Postoperative nausea and vomiting – subgroup
analysis by method of administration
Bolus only 3 274 0 1.17 (0.65 to 2.13) 0.60
Infusion only 6 297 0 0.75 (0.36 to 1.58) 0.45
Bolus followed by infusion 11 643 44 0.84 (0.57 to 1.25) 0.40
Sensitivity analysis – only low risk of bias 4 240 64 1.06 (0.22 to 5.12) 0.94
5 Incidence of shivering 7 568 35 0.26 (0.15 to 0.44) <0.001
Incidence of shivering – subgroup analysis by
method of administration
Bolus only 2 224 22 0.40 (0.15 to 1.02) 0.06
Infusion only 1 60 – 0.03 (0.00 to 0.58) 0.02
Bolus followed by infusion 4 284 41 0.26 (0.13 to 0.55) <0.001
Sensitivity analysis – only low risk of bias 2 140 74 0.18 (0.01 to 3.32) 0.25
6 Postoperative serum magnesium concentration 13 682 99 0.59 (0.39 to 0.78) <0.001
(mmol l1)
Postoperative serum magnesium concentration
– subgroup analysis by method of
administration
Bolus only 1 30 – 0.02 (0.05 to 0.01) 0.18
Infusion only 1 38 – 0.10 (0.22 to 0.42) 0.55
Bolus followed by infusion 11 614 99 0.68 (0.47 to 0.89) <0.001
Sensitivity analysis – only low risk of bias 2 126 98 0.28 (0.23 to 0.79) 0.29
7 Incidence of bradycardia 9 601 35 1.13 (0.43 to 2.98) 0.80
Incidence of bradycardia – subgroup analysis
by method of administration
Bolus only – – – – –
Infusion only 3 240 50 0.49 (0.09 to 2.72) 0.41
Bolus followed by infusion 6 361 43 1.80 (0.52 to 6.26) 0.36
Sensitivity analysis – only low risk of bias 2 140 0 0.15 (0.02 to 1.32) 0.09

CI, confidence interval; MD, mean difference; OR, odd ratio.

surgery (Fig. 2). However, the quality of evidence was [48.0 to 275.4] min: median difference (95% CI), 143 (103
graded as low due to significant inconsistency and indi- to 183) min, P less than 0.001, quality of evidence low.
rectness where different regimes of intravenous magne- Statistical heterogeneity was substantial across all the
sium were used across all included trials. included studies (I2 ¼ 99%). In the study by Gucyetmez
et al.,59 the time to first analgesia request in the magne-
Time to first analgesia request (minutes) sium group was significantly longer than in all the other
Eleven RCTs17,31,32,34,38,40,49,59,62,66,67 (824 patients) studies combined, with a mean time of 1050 150 min in
examined the time to first analgesia request after non- the magnesium group as compared with a time from all
cardiac surgery. For the magnesium group the median the other studies without Gucyetmez et al.59 of
time to the first request of analgesia after surgery was 249 315 289 min. But, even excluding that study,59 the
[63.7 to 653.5] min, and for the placebo group it was 192 time to the first request for analgesia in the magnesium

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Systemic magnesium reduces postoperative morphine consumption 219

Table 3 Summary of findings with GRADE assessment quality of evidence

Anticipated absolute effects

Outcomes No. of participants Certainty of the Relative effect Risk with Risk difference with
(studies) follow-up evidence (GRADE) (95% CI) placebo magnesium
Postoperative 945 (19 RCTs) – MD 5.41 lower (7.08 lower
morphine LOWa,b,c,d to 3.74 lower)
consumption (mg)
Time to first 824 (11 RCTs) – MD 142.88 higher
analgesia (min) LOWa,b,c,d (102.71 higher to 183.06 higher)
Incidence of shivering 568 (7 RCTs) OR 0.26 207 per 1000 144 fewer per 1000 (170
VERY LOWa,c,d,e,f (0.15 to 0.44) fewer to 104 fewer)
Pain score (visual 1232 (18 RCTs) – MD 0.3 lower (0.69 lower to
analogue score, LOWa,b,c,d 0.09 higher)
out of 10)
Postoperative serum 682 (13 RCTs) – MD 0.59 higher (0.39 higher
magnesium level MODERATEa,c,d to 0.78 higher)
(mmol l1)
Incidence of bradycardia 601 (9 RCTs) OR 1.13 21 per 1000 3 more per 1000 (12 fewer to
VERY LOWc,d,e,f,g (0.43 to 2.98) 40 more)
Incidence of 1214 (20 RCTs) OR 0.90 218 per 1000 17 fewer per 1000 (61 fewer
postoperative MODERATEb,f (0.67 to 1.22) to 36 more)
nausea and vomiting

M
The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate
certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially
different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty:
We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect. CI, confidence interval; MD, mean
difference; OR, odds ratio; RCT, randomised controlled trial. a Substantial heterogeneity (I2 > 60%). b Different regimes of intravenous magnesium (bolus only,
bolus þ infusion, infusion only) were used. c Funnel plot suggestion of publication bias. d Varied dosage of intravenous magnesium used across different trials. e It was not
the primary outcomes for most of the included trials. f Total number of events less than 300. g More than 50% of included studies were high-risk trials.

group was still significantly longer than the placebo and vomiting whether magnesium as given or not: OR
group. (95% CI), 0.90 (0.67 to 1.22), P ¼ 0.49, I2 ¼ 25%; quality of
evidence, moderate.
Postoperative pain score (visual analogue scale) in the
first 24 h Postoperative concentration of serum magnesium
Out of a total score of 10, the median [IQR] postoperative (mmol lS1) and incidence of bradycardia
pain score for the magnesium and placebo groups was 2.6 A total of 13 studies15,20,33–35,38,47,49,50,53,64,67,69 (682
[1.6 to 4.1] and 3 [1.4 to 4.6], respectively. In 18 trials, patients) measured the postoperative concentration of
there was no difference in pain scores whether patients serum magnesium. The mean serum magnesium level
received magnesium or not7,18,30,33,38,41– was 1.7 mmol l1 in the magnesium group and
44,46,48,50,55,57,60,61,63,71
(1232 patients, median difference 1.1 mmol l1 in the placebo group. Patients who received
(95% CI), 0.30 (0.69 to 0.09), P ¼ 0.13: quality of magnesium had a statistically significant higher concen-
evidence, low. A substantial degree of heterogeneity tration of serum magnesium as compared with the pla-
(I2 ¼ 91%) was detected in this pooled estimate. cebo group: mean difference (95% CI), 0.59 (0.39 to
0.78) mmol l1, P less than 0.001, I2 ¼ 99%; quality of
Incidence of shivering evidence, moderate. The incidence of bradycardia was
The incidence of shivering was 7.1% in the magnesium 2.2% in the magnesium group and 2.1% in the placebo
group and 20.7% in the placebo group. Based on the group. Based on nine RCTs36–38,41,45,51,54,57,60 (601
combined data of seven RCTs34,36,37,52,56,63,66 (568 patients), this review found no significant difference in
patients), the magnesium group had a significantly lower the incidence of bradycardia between the magnesium
incidence of shivering: OR (95% CI) 0.26 (0.15 to 0.44), P and placebo groups: OR (95% CI) 1.13 (0.43 to 2.98)%,
less than 0.001: quality of evidence, very low. The P ¼ 0.80, I2 ¼ 35%; quality of evidence, very low.
statistical heterogeneity was NS (I2 ¼ 35%) across
studies. Sensitivity analysis (low risk of bias studies)
Sensitivity analyses were performed based on trials with a
Incidence of postoperative nausea and vomiting low risk of bias for all the measured outcomes. The
The incidence of postoperative nausea and vomiting for pooled estimate effects remained unchanged for all the
the magnesium and placebo groups was 18.9 and 21.8%, measured outcomes, except for postoperative serum mag-
respectively. In 20 trials7,15,19,34,36,37,42,44,48,51– nesium concentration and the incidence of shivering
53,56,57,60,63,64,66,68,71
(1214 patients) there was no signifi- which became nonsignificant: magnesium concentration,
cant difference in the incidence of postoperative nausea two studies, low risk of bias, 126 patients, I2 ¼ 98%, the

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
220 Ng et al.

Fig. 2

Cumulative
Z-score
O’Brien-Fleming alpha-spending boundaries = 288
8

7
Z-curve
6

5
magnesium
Favours

259 Number of
–1 patients
(Linear scaled)
–2

–3
Favours
control

–4

–5

–6

–7

–8

Trial sequential analysis of postoperative morphine consumption in the first 24 h. X-axis: the number of patients randomised; Y-axis: the cumulative Z-
score; the blue cumulative Z-curve was constructed using a random-effects model. Red vertical line with diamonds: required information size of a
meta-analysis.

median difference (95% CI) was 0.28 (0.23 to 0.79), To the best of our knowledge, this is the most up to date
P ¼ 0.29); for the incidence of shivering, two studies, low review summarising the clinical evidence on the co-
risk of bias, 140 patients, I2 ¼ 74%, the median difference administration of systemic magnesium for reducing post-
(95% CI) was 0.18 (0.01 to 3.32), P ¼ 0.25). operative morphine consumption in noncardiac surgery.
We conducted an exhaustive literature search with all the
Discussions included RCTs undergoing a rigorous methodological
Our meta-analysis, in noncardiac surgery patients, dem- and quality of evidence assessment. This update has
onstrated that the adjunctive use of intravenous mag- added more than 20 trials,15–20,29,32–36,44–48,59–62,71 dou-
nesium resulted in a reduced morphine consumption bling the number of patients since the reviews by Mur-
postoperatively, with a mean difference of approxi- phy et al.,12 De Oliveira et al.13 and Albrecht et al.14 in
mately 5 mg in the first 24 h. In addition, patients 2013. In the meta-analysis by Oliveira et al.,13 positive
receiving magnesium had a longer time to the first benefits of magnesium were demonstrated, with a lower
request for postoperative analgesia, more than 2 h. dose of postoperative morphine and a lower postoperative
The incidence of shivering was significantly lower in pain score. However, their search was restricted to RCTs
the magnesium group. Although serum magnesium was comparing the use of magnesium versus placebo in
significantly higher in the magnesium group, no signifi- patients undergoing general anaesthesia only. In this
cant differences were demonstrated in the postopera- review, we have included 11 trials, which randomised
tive pain score or the incidences of postoperative nausea magnesium and placebo in patients undergoing regional
and vomiting, and bradycardia. The general quality of anaesthesia in noncardiac surgery.15,17,30–38 These posi-
evidence ranged from very low-to-moderate, due to tive findings were supported by Albrecht et al.14 How-
potential risk of bias, inconsistency, indirectness, ever, the authors admitted that, at that time, only a small
imprecision and dose–response gradient. number of trials were available for the analysis of time to

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Systemic magnesium reduces postoperative morphine consumption 221

first analgesic request.14 For this measured outcome, we injury.74 Despite the morphine-sparing effect of magne-
have included an additional five trials in our sium, it did not translate into a lower incidence of
review.17,32,34,49,59 Murphy et al.12 excluded trials which postoperative nausea and vomiting. Rahman and Beat-
used only a single bolus dose of intravenous magnesium: tie75 summarised that the use of peri-operative opioids
this introduced significant selection bias to their find- was only one of the factors causative of postoperative
ings.12 Five trials with a single bolus dose of magnesium nausea and vomiting via stimulation of chemoreceptor
were included in our current review.43,56,66–68 Lysa- trigger zone in the postrema. There are many confound-
kowski et al.73 reported no significant benefits of systemic ing factors, namely sex, age, obesity, type of surgery, type
magnesium on postoperative pain scores. However, this of anaesthesia and concurrent medications, that cause
review has to be interpreted with cautions as it was postoperative nausea and vomiting.75
underpowered and included paediatric patients. Thus,
Several studies found that surgical patients were vulner-
their findings cannot be generalised to an adult popula-
able to hypomagnesaemia due to the stress from surgical
tion.
pathology, pain, anxiety and peri-operative administra-
The effects of systemic magnesium in reducing postop- tion of fluids without magnesium supplementation.76–78
erative morphine consumption was consistent with the Surgical stress causes intracellular magnesium to move
previous meta-analysis of 20 RCTs (1257) patients.13 We into the extracellular space where it also promotes the
did not include data on morphine-equivalent analgesia: binding of magnesium to free fatty acids.77 In our review,
this was part of the measures to minimise the high degree the magnesium group was associated with a significantly
of heterogeneity, which was demonstrated in other meta- higher serum concentration of magnesium as compared to
analyses.12–14 In noncardiac surgery the trial sequential the placebo cohort. The 11 out of 13 RCTs15,20,33–
35,47,49,50,53,64,69
analysis was conclusive in demonstrating the beneficial reported that the serum magnesium
effect of magnesium in reducing morphine consumption exceeded the normal range (0.6 to 1.1 mmol l1) in the
in the first 24 h after surgery. However, the level of magnesium group. However, none of the patients had
evidence was low due to inconsistency (substantial het- documented adverse effect of magnesium, including no
erogeneity), indirectness (different regime and dosage of differences in the incidence of bradycardia.
magnesium) and publication bias. All the included trials
In this review, the incidence of shivering was significantly
used multimodal analgesia in the management of post-
lower in the magnesium group. This finding was consis-
operative pain, including intrathecal morphine or other
tent with the latest published meta-analysis,79 which
systemic analgesia (e.g. paracetamol, tramadol and ketor-
demonstrated that pre-operative and intra-operative
olac). We did not take account of such analgesia, our
magnesium is effective for the prevention of postopera-
interest was postoperative morphine consumption only,
tive shivering in surgical patients. However, our sensitiv-
and in any single study, the morphine use would be
ity analysis focusing on the two studies with low risk of
affected similarly in each of the magnesium groups. In
bias reported no significant difference in the incidence of
the current review, the benefits of magnesium for reduc-
shivering. With the limited number of high-quality RCTs
ing postoperative morphine consumption postoperatively
our findings need to be interpreted with care. The
were in line with the delay to the first request for
different regimes (bolus only, bolus followed by infusion,
analgesia. In comparison with the last reviews,13,14 our
infusion only) and the different doses used across the
current sample size for this outcome was doubled, with an
RCTs may introduce significant variance to our findings.
inclusion of an additional five RCTs.17,32,34,49,59 How-
At present, the pharmacokinetics and pharmacodynamics
ever, our findings need to be interpreted with caution as
of magnesium remain unclear in the literature. Thus,
the statistical heterogeneity remained substantial across
future RCTs investigating the dose–response effect of
all included studies.
magnesium could provide valuable information regarding
Although magnesium was shown to reduce postoperative the optimal dosage of magnesium to achieve steady-
morphine consumption, it did not significantly reduce the state concentration.
intensity of postoperative pain in the first 24 h after
noncardiac surgery. Murphy et al.12 reported a short-term Limitations
(<6 h) benefit of magnesium in reducing postoperative One of the limitations in this meta-analysis was the lack
pain, but not a long-term benefit (20 to 24 h postopera- of standardised definition for some of the measured
tively). Several reviews have documented the adjunctive outcomes, with various primary outcomes used across
use of magnesium as part of a multimodal analgesia all the included studies. A major confounding factor
regimen to reduce the intensity of postoperative pain was that all patients in the various studies were given
for 24 h after surgery.13,14,73 For its antinociception effect, different types of analgesia (nonsteroidal anti-inflamma-
it is believed that magnesium inhibits the entry of cal- tory drugs, paracetamol, fentanyl, tramadol, ketorolac,
cium into cells by acting as a noncompetitive antagonist pethidine). However, while this may affect the total
on the N-methyl-D-aspartate glutamate receptor, and doe of morphine used, both arms of the studies would
thereby preventing central sensitisation to peripheral be similarly affect. In addition, the majority of trials

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
222 Ng et al.

reporting our measured secondary outcomes were under- 14 Albrecht E, Kirkham KR, Liu SS, et al. Peri-operative intravenous
administration of magnesium sulphate and postoperative pain: a meta-
powered with each arm randomised to less than 150 analysis. Anaesthesia 2013; 68:79–90.
patients. The exclusion of trials published as abstracts 15 Frassanito L, Messina A, Vergari A, et al. Intravenous infusion of magnesium
sulfate and postoperative analgesia in total knee arthroplasty. Minerva
may introduce publication bias to our findings. Anestesiol 2015; 81:1184–1191.
16 Sousa AM, Rosado GM, Neto Jde S, et al. Magnesium sulfate improves
Conclusion postoperative analgesia in laparoscopic gynecologic surgeries: a double-
The current meta-analysis of low-quality randomised data blind randomized controlled trial. J Clin Anesth 2016; 34:379–384.
17 Muthiah T, Arora MK, Trikha A, et al. Efficacy of magnesium as an adjuvant
with substantial heterogeneity demonstrates that the to bupivacaine in 3-in-1 nerveblock for arthroscopic anterior cruciate
adjunctive use of intravenous magnesium along with multi- ligament repair. Indian J Anaesth 2016; 60:491–495.
18 Sohn HM, Jheon SH, Nam S, et al. Magnesium sulphate improves
modal analgesia may reduce the mean morphine consump- pulmonary function after video-assisted thoracoscopic surgery: a
tion in the first 24 h and delay the time to the first request for randomised double-blind placebo-controlled study. Eur J Anaesthesiol
analgesia in patients undergoing noncardiac surgery. 2017; 34:508–514.
19 Ahmed AS, Shoukry AA, Kamel MAM, et al. Analgesic potentials of
preoperative oral pregabalin, intravenous magnesium sulfate, and their
Acknowledgements relating to this article combination in acute postthoracotomy pain. Clin J Pain 2019; 35:247–251.
Assistance with the study: none. 20 Kizilcik N, Koner O. Magnesium sulfate reduced opioid consumption in
obese patients undergoing sleeve gastrectomy: a prospective, randomized
Financial support and sponsorship: none. clinical trial. Obes Surg 2018; 28:2783–2788.
21 The Cochrane Collaboration. In: Higgins JP, Green S, editors. Cochrane
Conflicts of interest: none. handbook for systematic reviews of interventions version 5.1.0. Chichester
(UK): John Wiley & Sons; 2011; https://training.cochrane.org/handbook.
Presentations: preliminary data of this study were presented as an oral [Updated March 2011; accessed 31 March 2019].
presentation at the Malaysian Medics International Medical Summit 22 Shamseer L, Moher D, Clarke M, et al. Preferred reporting items for
systematic review and meta-analysis protocols (PRISMA-P) 2015. BMJ
Research Competition 2018 (Gold award) on 18 to 19 August 2018 at
2015; 349:g4647.
Kuala Lumpur, Malaysia. The completed data of this study were 23 McMaster University and Evidence Prime Inc. GRADEpro GDT. Ontario,
accepted for e-poster presentation in the 17th International Congress Canada: McMaster University and Evidence Prime Inc; 2015; https://
of Cardiothoracic and Vascular Anesthesia, in conjunction with 13th gradepro.org/. [Accessed 25 June 2019].
Asian Society of Cardiothoracic Anesthesia at the COEX Convention 24 The Cochrane Collaboration. Review manager (RevMan) [computer
program]. Version 5.3. Copenhagen: The Nordic Cochrane Centre; 2014.
& Exhibition Centre in Seoul Korea (6 to 8 June 2019). 25 Sterne JAC, Sutton AJ, Ioannidis JPA, et al. Recommendations for
examining and interpreting funnel plot asymmetry in meta-analyses of
References randomised controlled trials. BMJ 2011; 343:d4002.
1 Fawcett WJ, Haxby EJ, Male DA. Magnesium: physiology and 26 Choi SW, Lam DMH. Heterogeneity in meta-analyses. Comparing apples
pharmacology. Br J Anaesth 1999; 83:302–320. and oranges? Anaesthesia 2017; 72:532–534.
2 Pannem RB, Manumala C. An observational study on effects of 27 Wan X, Wang WQ, Liu JM, et al. Estimating the sample mean and standard
hypomagnesemia among critically ill patients at a tertiary care hospital. Int J deviation from the sample size, median, range and/or interquartile range.
Advances Med 2018; 5: Available on: https://www.ijmedicine.com/ BMC Med Res Methodol 2014; 14:135.
index.php/ijam/article/download/1363/1063[Accessed 25 November 28 Thorlund K, Engstrom J, Wetterslev J, et al. User manual for trial sequential
2019]. analysis (TSA). 2017. Copenhagen Trial Unit.
3 Hansen BA, Bruserud O. Hypomagnesemia in critically ill patients. J 29 Asadollah S, Vahdat M, Yazdkhasti P, et al. The effect of magnesium
Intensive Care 2018; 6:1–11. sulphate on postoperative analgesia requirements in gynecological
4 George R, Condrey JM. ‘Oh Mg!’ Magnesium: a powerful tool in the surgeries. Turk J Obstet Gynecol 2015; 12:34–37.
perioperative setting. Pittsburgh, USA: American Society of Regional 30 Dabbagh A, Elyasi H, Razavi SS, et al. Intravenous magnesium sulfate for
Anesthesia and Pain Medicine; 2018; https://www.asra.com/asra-news/ postoperative pain in patients undergoing lower limb orthopedic surgery.
article/105/oh-mg-magnesium-a-powerful-tool-in-the. [Accessed 12 July Acta Anaesthesiol Scand 2009; 53:1088–1091.
2019]. 31 Kumar M, Dayal N, Rautela RS, et al. Effect of intravenous magnesium
5 McCarthy RJ, Kroin JS, Tuman KJ, et al. Antinociceptive potentiation and sulphate on postoperative pain following spinal anesthesia. A randomized
attenuation of tolerance by intrathecal co-infusion of magnesium sulfate and double blind controlled study. Middle East J Anaesthesiol 2013; 22:251–
morphine in rat. Anesth Analg 1998; 86:830–836. 256.
6 Bahar M, Chanimov M, Grinspun E, et al. Spinal anaesthesia induced by 32 Shah PN, Dhengle Y. Magnesium sulfate for postoperative analgesia after
intrathecal magnesium sulfate. Anaesthesia 1996; 51:627–633. surgery under spinal anesthesia. Acta Anaesthesiol Taiwan 2016; 54:62–
7 Ayoglu H, Karadeniz U, Kunduracilar Z, et al. The analgesic effect of 64.
magnesium sulfate and ketamine in patients undergoing laparoscopic 33 Shin HJ, Kim EY, Na HS, et al. Magnesium sulphate attenuates acute
cholecystectomy. Pain Clin 2005; 17:45–53. postoperative pain and increased pain intensity after surgical injury in
8 Mentes O, Harlak A, Yigit T, et al. Effect of intraoperative magnesium staged bilateral total knee arthroplasty: a randomized, double-blinded,
sulphate infusion on pain relief after laparoscopic cholecystectomy. Acta placebo-controlled trial. Br J Anaesth 2016; 117:497–503.
Anaesthesiol Scand 2008; 52:1353–1359. 34 Shal SM, Lofty E. Evaluation of effect of intravenous magnesium sulfate
9 Desborough JP. The stress response to trauma and surgery. Br J Anaesth infusion on tourniquet induced hypertensionand pain in arthroscopic knee
2000; 85:109–117. surgery patientsunder epidural anesthesia. Egypt J Anaesth 2017; 33:73–
10 Honarmand A, Safavi M, Badiei S, et al. Different doses of intravenous 82.
magnesium sulfate on cardiovascular changes following the laryngoscopy 35 Kahraman F, Eroglu A. The effect of intravenous magnesium sulfate infusion
and tracheal intubation: a double-blind randomized controlled trial. J Res on sensory spinal block and postoperative pain score in abdominal
Pharm Pract 2015; 4:79–84. hysterectomy. Biomed Res Int 2014; 2014:236024.
11 James MF, Beer RE, Esser JD. Intravenous magnesium sulfate inhibits 36 Ibrahim IT, Megalla SA, Khalifa OM, et al. Prophylactic vs. therapeutic
catecholamine release associated with tracheal intubation. Anesth Analg magnesium sulfate for shivering during spinal anesthesia. Egypt J Anaesth
1989; 68:772–776. 2014; 30:31–37.
12 Murphy JD, Paskaradevan J, Eisler LL, et al. Analgesic efficacy of 37 Gozdemir M, Usta B, Demircioglu RI, et al. Magnesium sulfate infusion
continuous intravenous magnesium infusion as an adjuvant to morphine for prevents shivering duringtransurethral prostatectomy with spinal
postoperative analgesia: a systematic review and meta-analysis. Middle anesthesia: a randomized, double-blinded, controlled study. J Clin Anesth
East J Anaesthesiol 2013; 22:11–20. 2010; 22:184–189.
13 De Oliveira GS, Castro-Alves LJ, Khan JH, et al. Perioperative systemic 38 Hwang JY, Na HS, Jeon YT, et al. I.V. infusion of magnesium sulphate during
magnesium to minimize postoperative pain: a meta-analysis of randomised spinal anaesthesia improves postoperative analgesia. Br J Anaesth 2010;
controlled trials. Anesthesiology 2013; 119:178–190. 104:89–93.

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Systemic magnesium reduces postoperative morphine consumption 223

39 Jaoua H, Zghidi SM, Wissem L, et al. Effectiveness of intravenous 59 Gucyetmez B, Atalan HK, Aslan S, et al. Effects of intraoperative
magnesium on postoperative pain after abdominal surgery versus magnesium sulfate administration on postoperative tramadol requirement in
placebo: double blind randomized controlled trial. Tunis Med 2010; liver transplantation: a prospective, double-blind study. Transplant Proc
88:317–323. 2016; 48:2742–2746.
40 Saadawy IM, Kaki AM, Abd El Latif AA, et al. Lidocaine vs. magnesium: 60 Ryu JH, Koo BW, Kim BG, et al. Prospective, randomized and controlled
effect on analgesia after a laparoscopic cholecystectomy. Acta trial on magnesium sulfate administration during laparoscopic gastrectomy:
Anaesthesiol Scand 2010; 54:549–556. effects on surgical space conditions and recovery profiles. Surg Endosc
41 Kiran S, Gupta R, Verma D. Evaluation of a single-dose of intravenous 2016; 30:4976–4984.
magnesium sulphate for prevention of postoperative pain after inguinal 61 Demiroglu M, Un C, Dikmen D, et al. The effect of systemic and regional use
surgery. Indian J Anaesth 2011; 55:31–35. of magnesium sulfate on postoperative tramadol consumption in lumbar
42 Kocman IB, Krobot R, Premuzić J, et al. The effect of preemptive disc surgery. Biomed Res Int 2016; 2016:3216246.
intravenous low-dose magnesium sulfate on early postoperative pain after 62 Khafagy HF, Ebied RS, Osman ES, et al. Perioperative effects of various
laparoscopic cholecystectomy. Acta Clin Croat 2013; 52:289–294. anesthetic adjuvants with TIVA guided by bispectral index. Korean J
43 Mireskandari SM, Pestei K, Nabavian O, et al. Effects of preoperative Anesthesiol 2012; 63:113–119.
magnesium sulphate on postcesarean pain, a placebo controlled double
63 Song JW, Lee YW, Yoon KB, et al. Magnesium sulfate prevents
blind study. J Family Reprod Health 2015; 9:29–33.
remifentanil-induced postoperative hyperalgesia in patients undergoing
44 Taheri A, Haryalchi K, Mansour Ghanaie M, et al. Effect of low-dose (single-
thyroidectomy. Anesth Analg 2011; 113:390–397.
dose) magnesium sulfate on postoperative analgesia in hysterectomy
patients receiving balanced general anesthesia. Anesthesiol Res Pract 64 Ko SH, Lim HR, Kim DC, et al. Magnesium sulfate does not reduce
2015; 2015:306145. postoperative analgesic requirements. Anesthesiology 2011; 95:640–646.
45 Walia C, Gupta R, Kaur B, et al. Propofol sparing effect of 65 Oguzhan N, Gunday I, Turan A. Effect of magnesium sulfate infusion on
dexmedetomidine and magnesium sulfate during BIS targeted anesthesia: sevoflurane consumption, hemodynamics, and perioperative opioid
a prospective, randomized, placebo controlled trial. J Anaesthesiol Clin consumption in lumbar disc surgery. J Opioid Manag 2008; 4:105–110.
Pharmacol 2018; 34:335–340. 66 Tramer MR, Glynn CJ. An evaluation of a single dose of magnesium to
46 Elsersy HE, Metyas MC, Elfeky HA, et al. Intraoperative magnesium supplement analgesia after ambulatory surgery: randomized controlled trial.
sulphate decreases agitation and pain in patients undergoing functional Anesth Analg 2007; 104:1374–1379.
endoscopic surgery: a randomised double-blind study. Eur J Anaesthesiol 67 Tauzin-Fin P, Sesay M, Delort-Laval S, et al. Intravenous magnesium
2017; 34:658–664. sulphate decreases postoperative tramadol requirement after radical
47 Vicković S, Pjević M, Uvelin A, et al. Magnesium sulfate as an adjuvant to prostatectomy. Eur J Anaesthesiol 2006; 23:1055–1059.
anesthesia in patients with arterial hypertension. Acta Clin Croat 2016; 68 Schulz-Stübner S, Wettmann G, Reyle-Hahn SM, et al. Magnesium as part
55:490–496. of balanced general anaesthesia with propofol, remifentanil and
48 Kim JE, Shin CS, Lee YC, et al. Beneficial effect of intravenous magnesium mivacurium: a double-blind, randomized prospective study in 50 patients.
during endoscopic submucosal dissection for gastric neoplasm. Surg Eur J Anaesthesiol 2001; 18:723–729.
Endosc 2015; 29:3795–3802. 69 Koinig H, Wallner T, Marhofer P, et al. Magnesium sulfate reduces intra- and
49 De Oliveira GS, Bialek J, Fitzgerald P, et al. Systemic magnesium to postoperative analgesic requirements. Anesth Analg 1998; 87:206–210.
improve quality of postsurgical recovery in outpatient segmental 70 Benhaj Amor M, Barakette M, Dhahri S, et al. Effect of intra and
mastectomy: a randomized, double-blind, placebo-controlled trial. Magnes postoperative magnesium sulphate infusion on postoperative pain. Tunis
Res 2013; 26:156–164. Med 2008; 86:550–555.
50 Olgun B, Oğuz G, Kaya M, et al. The effects of magnesium sulphate on
71 Haryalchi K, Abedinzade M, Khanaki K, et al. Whether preventive low dose
desflurane requirement, early recovery and postoperative analgesia in magnesium sulphate infusion has an influence on postoperative pain
laparascopic cholecystectomy. Magnes Res 2012; 25:72–78.
perception and the level of serum beta-endorphin throughout the total
51 Kaya S, Kararmaz A, Gedik R, et al. Magnesium sulfate reduces abdominal hysterectomy. Rev Esp Anestesiol Reanim 2017; 64:384–390.
postoperative morphine requirement after remifentanil-based anaesthesia.
72 Mavrommati PD, Gabopoulou ZT, Papadimos CN, et al. The perioperative
Med Sci Monit 2009; 15:PI5–PI9.
infusion of low doses of magnesium sulfate reduces analgesic requirements
52 Ryu JH, Kang MH, Park KS, et al. Effects of magnesium sulphate on
in patients undergoing abdominal hernioplasty. Arthroscopy 2004; 5:81–87.
intraoperative anaesthetic requirements and postoperative analgesia in
gynaecology patients receiving total intravenous anaesthesia. Br J Anaesth 73 Lysakowski C, Dumont L, Czarnetzki C, et al. Magnesium as an adjuvant to
2008; 100:397–403. postoperative analgesia: a systematic review of randomized trials. Anesth
53 Cizmeci P, Ozkose Z. Magnesium sulphate as an adjuvant to total Analg 2007; 104:1532–1539.
intravenous anesthesia in septorhinoplasty: a randomized controlled study. 74 Vink R. Magnesium in the central nervous system. Adelaide, South Australia:
Aesthetic Plast Surg 2007; 31:167–173. University of Adelaide Press; 2011; https://www.ncbi.nlm.nih.gov/books/
54 Seyhan TO, Tugrul M, Sungur MO, et al. Effects of three different dose NBK507245/. [Accessed 23 February 2019].
regimens of magnesium on propofol requirements, haemodynamic 75 Rahman MH, Beattie J. Postoperative nausea and vomiting. Pharm J 2018;
variables and postoperative pain relief in gynaecological surgery. Br J https://www.pharmaceutical-journal.com/learning/learning-article/
Anaesth 2006; 96:247–252. postoperative-nausea-and-vomiting/10030469.article. [Accessed 24
55 Bhatia A, Kashyap L, Pawar DK, et al. Effect of intraoperative magnesium February 2019].
infusion on perioperative analgesia in open cholecystectomy. J Clin Anesth 76 Rosanoff A, Weaver CM, Rude RK. Suboptimal magnesium status int he
2004; 16:262–265. United States: are the health consequences underestimated? Nutr Rev
56 Levaux Ch, Bonhomme V, Dewandre PY, et al. Effect of intra-operative 2012; 70:153–164.
magnesium sulphate on pain relief and patient comfort after major lumbar 77 Fulop T, Agarwal M, Keri KC. Hypomagnesemia. Medscape 2018; https://
orthopaedic surgery. Anaesthesia 2003; 58:131–135. emedicine.medscape.com/article/2038394-overview#a5. [Accessed 24
57 Zarauza R, Saez-Fernandez AN, Iribarren MH, et al. A comparative study February 2019].
with oral nifedipine, intravenous nimodipine, and magnesium sulfate in 78 Swaminathan R. Magnesium metabolism and its disorders. Clin Biochem
postoperative analgesia. Anesth Analg 2000; 91:938–943. Rev 2003; 24:47–66.
58 Wilder-Smith CH, Kn€ opfli R, Wilder-Smith OH. Perioperative magnesium 79 Hiromasa K, Daisuke N, Takahiro M, et al. Effectiveness of magnesium in
infusion and postoperative pain. Acta Anaesthesiol Scand 1997; preventing shivering in surgical patients: a systematic review and meta-
41:1023–1027. analysis. Anesth Analg 2019; 129:689–700.

Eur J Anaesthesiol 2020; 37:212–223

Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.