The Effectiveness of Midazolam for Preventing ...sbmu.ac.ir/uploads/2015- the effectiveness of midazolam.pdf · Comprehensive meta-analysis software (version 2.0, Biostat, Englewood,

Copyright © 2015 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.XXX 2015 • Volume XXX • Number XXX www.anesthesia-analgesia.org 1

Copyright © 2015 International Anesthesia Research SocietyDOI: 10.1213/ANE.0000000000001062

The common and stressful adverse effects of postop-erative nausea and vomiting (PONV) are a major con-cern in 30% to 70% of high-risk groups.1 Accordingly,

numerous studies have assessed the efficacies of different drugs and treatment protocols in preventing PONV. Agents including anticholinergics, antidopaminergics, antihista-mines, steroids, neuroleptics, and serotonin antagonists have shown variable effects.2–5 Hence, PONV remains a complica-tion encountered frequently during the postoperative period.

Midazolam is a short-acting benzodiazepine that affects central benzodiazepine receptors, which contain a γ-aminobutyric acid receptor site and a chloride ion channel.6 Midazolam has a sedative effect and can be used for coinduc-tion, because it reduces the requirements for propofol and thereby lowers the incidence of the cardiovascular effects of propofol.7 Midazolam is the most commonly used sedative agent in the ambulatory setting, most likely because it has fewer depressive cardiorespiratory effects than propofol.8

Numerous studies have reported that midazolam effec-tively prevents PONV.9–12 The use of midazolam not only may reduce the incidence of PONV but may also provide an

anxiolytic effect. Moreover, midazolam may offer the ben-efits of lower cost and fewer side effects, such as headaches and extrapyramidal symptoms, that have been reported with other antiemetics.13 In several studies, such antiemet-ics did not provide superior benefits or outcomes compared with midazolam.14,15 However, the findings are variable, and the reported outcomes from several studies are conflicting, especially when midazolam has been used in combination with other antiemetics. Therefore, we aimed to determine the effectiveness of midazolam in preventing PONV by per-forming a systematic review and meta-analysis.

METHODSStudy Protocol RegistrationThis systematic review was registered in PROSPERO (CRD42014009425) and was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guidelines.16

Systematic SearchWe conducted a systematic review and meta-analysis of ran-domized controlled trials that investigated the effectiveness of midazolam in preventing PONV in comparison with controls. Studies combining midazolam and other drugs were included, provided the only difference between the groups was the use of midazolam. The MEDLINE®, Embase, and Cochrane Central Register of Controlled Trials (CENTRAL) databases were searched for all relevant articles written in English up to March 1, 2014 (inclusive). In addition, the reference lists of the full articles retrieved were searched manually. The search strat-egy, which included a combination of free text, Medical Subject Headings, and EMTREE terms, is described in the Appendix.

Study SelectionInclusion and exclusion criteria were determined before the systematic search. Two authors independently scanned the

BACKGROUND: Previous randomized controlled trials regarding the effectiveness of periopera-tive midazolam in preventing postoperative nausea and vomiting (PONV) have produced conflict-ing results. Consequently, the present systematic review was performed to assess the effect of perioperative administration of midazolam on PONV.METHODS: The MEDLINE®, Embase, and Cochrane Central Register of Controlled Trials data-bases were searched to identify all randomized controlled trials that investigated the effec-tiveness of midazolam under general anesthesia. The primary end points were defined as postoperative nausea (PON), postoperative vomiting (POV), and PONV.RESULTS: From 16 studies, 1433 patients were included in the final analysis. Compared with the control group, patients who received midazolam showed a lower overall incidence of PON (risk ratio [RR], 0.51; 95% confidence interval [CI], 0.40–0.65; I2 = 35%; number needed to treat [NNT] = 6; number of included studies [n] = 11), POV (RR, 0.46; 95% CI, 0.33–0.65; I2 = 0%; NNT = 8; n = 10), and PONV (RR, 0.45; 95% CI, 0.36–0.57; I2 = 31%; NNT = 3; n = 7).CONCLUSIONS: Perioperative administration of midazolam was effective in preventing PON, POV, and PONV. (Anesth Analg 2015;XXX:00–00)

The Effectiveness of Midazolam for Preventing Postoperative Nausea and Vomiting: A Systematic Review and Meta-AnalysisEun Jin Ahn, MD,* Hyun Kang, MD, PhD, MPH,† Geun Joo Choi, MD,† Chong Wha Baek, MD, PhD,† Yong Hun Jung, MD, PhD,† and Young Choel Woo, MD, PhD†

From the *Department of Anaesthesiology and Pain Medicine, Inje University, Seoul Paik Hospital, Seoul, Korea; and †Department of Anaesthesiology and Pain Medicine, Chung-Ang University College of Medicine, Seoul, Korea.

Accepted for publication September 14, 2015.

Funding: This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2012R1A1A1003700).

The authors declare no conflicts of interest.

LMA is a registered trade mark of The Laryngeal Mask Company Limited, an affiliate of Teleflex Incorporated.

Reprints will not be available from the authors.

Address correspondence to Hyun Kang, MD, PhD, MPH, Department of Anaesthesiology and Pain Medicine, Chung-Ang University, College of Medicine, 224-1 Heukseok-dong, Dongjak-gu, Seoul 156-755, Korea. Address e-mail to [email protected].

mailto:[email protected]

Copyright © 2015 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.2 www.anesthesia-analgesia.org aNesthesia & aNalgesia

Meta-Analysis for Midazolam and PONV

titles and abstracts of the reports identified via the search strategies described earlier. If a report was determined eli-gible from the title or abstract, the full article was retrieved. Potentially relevant studies chosen by at least one author were retrieved, and full-text versions were evaluated. Articles that met the inclusion criteria were assessed sepa-rately by 2 authors, and any discrepancies were resolved through discussion. If no agreement could be reached, a third investigator was involved to provide a resolution.

Inclusion and Exclusion CriteriaWe included randomized controlled trials that collected data on the effectiveness of midazolam as a prophylactic agent for PONV compared with a control. Control groups were com-pared with an experimental group, including subjects who did not receive a drug or who received another drug, with the only difference between the groups of midazolam administra-tion (e.g., midazolam + ondansetron versus ondansetron). The studies also included adult patients undergoing general anes-thesia involving IV administration of midazolam. Data from abstracts, posters, case reports, comments, or letters to the edi-tor, reviews, and animal studies were excluded. Studies that were not published in English were likewise excluded (Fig. 1).

Study OutcomesThe primary end points were the incidence of postoperative nausea (PON), postoperative vomiting (POV), and PONV. The severity of PON was also assessed. The use of rescue antiemetics and the incidence of adverse effects, such as headache, dizziness, and sedation, were secondary out-comes in the systematic review.

Subgroup analysis was performed regardless of whether PON, POV, and PONV incidence data were collected as pri-mary outcomes or secondary outcomes. Also, a subgroup analysis was performed to rule out the effects of propofol as an induction agent. In addition, administration times were divided into 2 phases representing the induction and con-clusion of surgery. In studies with >1 midazolam group or control group, all midazolam groups or all control groups were combined to avoid a unit of analysis error.

Validity ScoringThe quality of eligible studies was assessed independently by 2 members of the review group using the “risk of bias” tool within the Review Manager software program (version 5.2, The Cochrane Collaboration, Oxford, UK). Quality was evaluated based on potential sources of bias, including ran-dom sequence generation, allocation concealment, blind-ing of the participants, blinding of outcome assessment, incomplete outcome data, selective reporting, and overall risk bias. The data were then cross-checked. The methodol-ogy of each trial was graded as “high,” “low,” or “unclear,” to reflect a high risk or a low risk of bias, and uncertainty regarding the risk of bias, respectively.17

Data ExtractionAll interrelated data from the included studies were inde-pendently extracted and entered into a spreadsheet by 2 authors, and then cross-checked. Any discrepancy was resolved through discussion. If an agreement could not be reached, the dispute was resolved with the aid of a third investigator. The spreadsheet included the following items:

Figure 1. PRISMA flow diagram of the search, inclusion and exclusion of randomized con-trolled trials.

Copyright © 2015 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.

XXX 2015 • Volume XXX • Number XXX www.anesthesia-analgesia.org 3

(1) title, (2) authors, (3) name of journal, (4) publication year, (5) study design, (6) registration of clinical trial, (7) com-peting interest, (8) country, (9) risk of bias, (10) number of patients in study, (11) dose of midazolam and control drugs, (12) sex, (13) age, (14) weight, (15) height of patients, (16) duration of anesthesia, (17) ASA physical status, (18) inclu-sion criteria, (19) exclusion criteria, (20) type of surgery, (21) airway device, (22) induction agent, (23) maintenance agent, (24) use of nitrous oxide, (25) use of an opioid during the perioperative period, (26) timing of administration of the experimental drug, (27) other drugs used during surgery, (28) timing of rescue antiemetics, (29) rescue analgesics, (30) definition of nausea, vomiting, and retching, (31) number of cases of PON, POV, and PONV overall and during the early and late postoperative phases, and (32) the need for rescue antiemetics.

The data were initially extracted from tables or text. In cases involving missing or incomplete data, an attempt was made to contact the study authors to obtain the relevant information. If the author did not respond or did not have current information, the data were extracted from available figures.

Statistical AnalysisThe review and meta-analysis were conducted using Comprehensive meta-analysis software (version 2.0, Biostat, Englewood, NJ) and Review Manager software (version 5.2, The Cochrane Collaboration, Oxford, UK). For dichotomous data, a pooled risk ratio (RR) and 95% confidence intervals (CIs) were calculated. If the 95% CI included a value of 1, the difference was not considered statistically significant. We calculated the mean difference (MD) for continuous data and reported the 95% CI. If the 95% CI included a value of 0, the difference was not considered statistically significant.

We used the χ2 test and the I2 test for heterogeneity. If the P value was <0.10 or an I2 value was >50%, the hetero-geneity was considered significant. A fixed-effects model was selected if P value for χ2 test >0.10 and the I2 value was <50%. In cases with I2 value >50%, random-effects models were used. Because all the number of combined studies that showed substantial heterogeneity were <10, t statis-tics (Hartung-Knapp-Sidik-Jonkman method) were used instead of Z test in all random-effects analysis to lower the error rate. A subgroup analysis was performed based on the timing of the PON, POV, and PONV assessments; the time of midazolam administration; the use of periopera-tive opioids; the use of propofol as an induction agent; and whether the aim of the study included PONV as a primary or secondary outcome. If heterogeneity was >50%, sensitiv-ity analysis was likewise performed by excluding one study at a time to evaluate the influence of a single study on the overall effect estimate.

We calculated the number needed to treat (NNT) using a 95% CI based on the absolute risk reduction as an esti-mate of the overall clinical impact of the intervention.18 Publication bias was identified by the Egger linear regres-sion test, and a P value of <0.05 was used to identify the presence of a publication bias, or funnel plots for each data set were visually assessed for asymmetry. If a publication bias was present, a trim-and-fill analysis was performed to evaluate the effect of the bias.

RESULTSLiterature Search and Study CharacteristicsA search of the MEDLINE, Embase, and CENTRAL data-bases returned 1296 studies that were initially evaluated. After excluding duplicates, 1197 studies remained. Another 1040 studies were excluded after the review of titles and abstracts. Studies that were not published in English were likewise excluded. Of the 151 studies that remained, 83 additional studies were excluded because 39 of the studies involved children and 44 studies involved administration of experimental drugs. Forty-two studies were excluded because they were performed under monitored or regional anesthesia care and not under general anesthesia. The full texts of the remaining 26 studies were reviewed in detail, and 11 additional studies were excluded because 8 did not include a control group,15,19–25 2 administered midazolam for therapeutic management,14,26 and 1 did not report findings for PON, POV, or PONV.27 Manual review of the references of the included studies identified 13 additional studies for assessment. However, only one met the inclusion criteria for the present review.28 Collectively, 16 studies met the inclu-sion criteria and were included in this systematic review and meta-analysis.6,8–10,12,14,26,28–38

The characteristics of the 16 studies that met the inclusion criteria are summarized in Tables 1, 2, and 3. Midazolam was administered at induction in 11 studies8–10,12,29,30,32,33,35,37,38 and at the end of the surgery in 3 studies.6,28,31 Two stud-ies compared the timing of administration,34,36 2 studies included a nausea visual analog scale for assessing the severity of PON,10,32 and 3 studies did not involve adminis-tration of perioperative opioids.12,37,38 Twelve of the 16 stud-ies6,10,12,28,29,31–34,36–38 were aimed at comparing the incidence of PONV, and 4 studies involved the collection of PONV incidence as a secondary end point.8,9,30,35

In the studies by Park et al.34 and Safavi and Honarmand,36 subjects were divided into groups based on drug adminis-tration at the induction or conclusion of surgery. The group receiving midazolam or placebo administration at induc-tion was designated Park1 and Safavi1, respectively, and the groups receiving the drugs at the end of surgery were des-ignated Park2 and Safavi2, respectively. To avoid duplicate counting, the groups were divided in both studies.

In the study by Honarmand et al.,6 4 separate cohorts received haloperidol, midazolam, haloperidol plus mid-azolam, or saline. Similarly, 4 separate cohorts in the study conducted by Makhdoom and Farid33 received dexameth-asone, midazolam, dexamethasone plus midazolam, or saline. Therefore, comparison of midazolam with a placebo was designated Honarmand1 and Makhdoom1, respec-tively, and the comparison of midazolam plus another drug with the second drug was designated Honarmand2 and Makhdoom2, respectively.

The postoperative period was divided into the overall, early, and late phases. The early phase was defined as 0 to 6 hours postoperatively, and the late phase as 6 to 24 hours postoperatively. The overall phase was included to capture the maximal number of studies that contained PON, POV, and PONV data with a variable data collection period and was defined as the first period of data collection (Table 4). Because several studies defined the early and late phase dif-ferently, we combined the earliest period of data collection



with early incidence of PON, POV, PONV, and severity of PON. For the early phase, 3 studies included postanesthesia care unit data,8,28,30 4 studies included data collected at 0 to 2 hours,32,34,36,37 and 1 study collected data at 0 to 60 minutes.35 In addition, the widest overlapping period of data collection for late incidence of PON, POV, PONV, and late PON sever-ity was combined. For the late phase, one study included data from 8 to 14 hours,10 a second study contained 2- to 24-hour data,34 3 studies reported 6- to 12-hour data,32,36,37 and 1 additional study included 6- to 24-hour data.28

Risk of BiasFifteen studies referenced the random sequence generation method,6,8–10,12,29–38 and only 2 studies described allocation concealment.32,34 In every study, outcome assessors were blinded, and no incomplete data were reported. The overall risks of bias are shown in Table 5.

Postoperative NauseaOverall, Early, and Late PONEleven studies compared the effectiveness of midazolam in preventing overall PON with those of a placebo or another drug.8,9,12,29–31,33–35,37,38 Five studies compared its effectiveness in reducing the incidence of early PON,8,30,34,35,37 and 2 stud-ies compared its effects on late PON.34,37 Analysis of the com-bined results revealed that midazolam was associated with a statistically significant reduction in the overall incidence of PON (risk ratio [RR], 0.51; 95% CI, 0.40–0.65; I2 = 35%; P

χ2 = 0.10; NNT = 6; number of included studies [n] = 11)

compared with the control groups. However, the effect of midazolam on the incidence of early and late PON compared with the control groups is unknown (RR, 0.56; 95% CI, 0.17–1.82; P

χ2 = 0.069; I2 = 51%; NNT = 7; n = 5; and RR, 0.74; 95% CI, 0.42–1.31; P

χ2 = 0.14; I2 = 49.7%; NNT = 11; n = 2, respectively).

Effect of Administration Timing on Overall PONNine studies8,12,29,30,33–35,37,38 analyzed the effectiveness of mid-azolam administered at induction of surgery in preventing overall PON, whereas 2 studies31,34 assessed its effectiveness after administration at the conclusion of surgery. The com-bined results of both analyses indicated that midazolam caused a statistically significant reduction in the incidence of overall PON, regardless of the timing of administration (RR, 0.53; 95% CI, 0.39–0.71; I2 = 31%; P

χ2 = 0.16; NNT = 7; n = 9; and RR, 0.30; 95% CI, 0.14–0.66; I2 = 0%; P

χ2 = 0.61; NNT = 4; n = 2, respectively; Fig. 2).

Severity of PONTwo studies10,32 compared the effects of midazolam on the severity of PON using a visual analog scale score. The post-operative periods of the 2 studies were divided into the early phase (0–6 hours after surgery) and the late phase (6–24 hours after surgery). However, we were not able to clarify the effect on the severity of PON after administra-tion of midazolam in the early or late postoperative phases when compared with a placebo (MD, −0.38; 95% CI, −3.39 to 2.64; I2 = 77.2%; P

χ2 = 0.036; n = 2; and MD, −0.04; 95% CI, −0.71 to 0.62; I2 = 0%; P

χ2 = 0.39; n = 2, respectively).

Table 1. Data Extracted from the Included Studies

Source Risk factors for PONV ASA PS AgeDuration of

anesthesia (min) Type of surgery Induction agentBauer et al. (2004)29 ≥0 I–III 18–65 79 (6.5) Outpatient surgery No limitationGilliland et al. (1996)9 ≥2 (female, IV-PCA) I–II <75 Not mentioned Total abdominal

hysterectomyPropofol

Godsiff et al. (1995)8 ≥1 (postoperative opioid) I–IV Not mentioned 34.9 (19.7) Orthopedic, urologic, or general surgery who can use LMA™

Propofol or propofol plus midazolam

Hasani et al. (2011)30 ≥0 I–II 35–65 41 (19) Inguinal hernia surgery Propofol, fentanylHeidari et al. (2004)10 ≥1 (nonsmoking) I–II Adult Not mentioned Cholecystectomy Thiopental sodium,

fentanylHonarmand et al.

(2012)6≥1 (postoperative opioid) I–II 18–60 221.9 (23.1) Middle ear surgery Thiopental sodium,

fentanylHuh, et al. (2010)31 ≥3 (female, nonsmoking,

IV-PCA)I–II 35–60 107.1 (15.5) Total abdominal

hysterectomyPropofol, fentanyl

Jung et al. (2007)12 ≥1 (female) I–II Not mentioned 184.9 (62.4) Middle ear surgery PropofolKim et al. (2012)28 ≥3 (female, nonsmoking,

IV-PCA)I–II 20–60 109.4 (59.7) Pelviscopic gynecologic

surgeryThiopental sodium,

fentanylKim et al. (2013)32 ≥3 (female, nonsmoking,

IV-PCA)I–II 20–65 150 (126.3–173.8) Thyroidectomy Thiopental sodium,

fentanylMakhdoom and Farid

(2009)33

≥1 (female) I–II Not mentioned 136.5 Middle ear surgery Sevoflurane

Park et al. (2013)34 ≥3 (female, nonsmoking, IV-PCA)

I–II 19–64 103.5 (25.4) Laparoscopic ovarian cystectomy or hysterectomy

Thiopental

Richardson et al. (1997)35

≥2 (female, postoperative opioid)

I–II Not mentioned 46 (13) Laparoscopic tubal ligation

Propofol, fentanyl

Safavi and Honarmand (2009)36

≥1 (postoperative opioid) I–II 18–60 107.5 (7.9) Lower abdominal procedure

Propofol, fentanyl

Shirdashtzadeh et al. (2011)37

≥1 (female) I–II 15–40 44 (3.8) Appendectomy Thiopental

Yeo et al. (2009)38 ≥2 (female, IV-PCA) I 19–62 182.9 (42.4) Middle ear surgery Propofol

PONV = postoperative nausea and vomiting; PS = physical status; LMA™ = laryngeal mask airway; PCA = patient-controlled analgesia.



Tabl

e 2.

Fur

ther

Dat

a Ex

trac

ted

from

the

Inc

lude

d Stu

dies

Sou

rce

Mai

nten

ance

age

ntA

dmin

istr

atio

n ro

ute

Adm

inis

trat

ion

tim

ing

Inte

rven

tion

/co

ntro

lPer

iod

of d

ata

colle

ctio

nTi

min

g of

res

cue

an

tiem

etic

sB

auer

et

al. (2

004)2

9N

o lim

itatio

nIV

Bef

ore

indu

ctio

nM

0.0

4 m

g/kg

/pla

cebo

0–2

4 h

(PO

N, P

OV)

, PAC

U

(res

cue

antie

met

ics)

PON

V

Gill

iland

et

al. (1

996)9

Isofl

uran

e, N

2O

IVAt

indu

ctio

n, fol

low

ed

by 4

8 h

infu

sion

M 5

mg/

70 k

g bo

lus

+ 1

mg/

70

kg/h

for

48 h

/pla

cebo

0–4

8 h

(PO

N, r

escu

e

antie

met

ics)

PON

V

God

siff e

t al

. (1

995)8

Enflu

rane

, N2O

IVAt

indu

ctio

nM

2.5

–5 m

g + p

ropo

fol 3

0–1

00 m

g/

prop

ofol

60–2

00 m

gPA

CU

/0–2

4 h

(PO

N)

PON

V

Has

ani e

t al

. (2

011)3

0Pr

opof

ol, f

enta

nyl,

N2O

IVB

efor

e in

duct

ion

M 0

.05 m

g/kg

+ 1

.5 m

g/kg

dic

lofe

nac/

di

clof

enac

1.5

mg/

kgPA

CU

(PO

N, P

OV)

Not

rep

orte

d

Hei

dari

et a

l. (2

004)1

0H

alot

hane

, N2O

IVB

efor

e in

duct

ion

M 7

5 μ

g/kg

/pla

cebo

PAC

U/2

–8/8

–14/1

4–2

4 h

(N

VAS

)/0–2

4 h

(PO

V)N

ot r

epor

ted

Hon

arm

and

et a

l. (2

012)6

Isofl

uran

e, N

2O

IVEn

d of

sur

gery

2 m

g ha

lope

ridol

+ M

2 m

g/M

2 m

g/2 m

g ha

lope

ridol

/pla

cebo

0–2

/2–2

4/0

–24 h

(PO

NV)

Not

rep

orte

d

Huh

et

al. (2

010)3

1Is

oflur

ane,

N2O

IVIn

the

PAC

U fol

low

ed

by P

CA

infu

sion

5 m

g m

orph

ine

+ M

1 m

g + M

0.4

mg/

m

L + 1

mg/

mL

mor

phin

e PC

A/5 m

g m

orph

ine

+ 1

mg/

mL

mor

phin

e PC

A

0–2

4 h

(PO

N, P

OV,

PO

NV,

re

scue

ant

iem

etic

s)An

tiem

etic

s re

quire

d

for

vom

iting

Jung

et

al. (2

007)1

2Is

oflur

ane,

N2O

IVAt

indu

ctio

nM

75 μ

g/kg

/pla

cebo

0–2

4 h

(PO

N, P

OV,

PO

NV,

res

cue

antie

met

ics)

Antie

met

ics

requ

ired

for

vom

iting

Kim

et

al. (2

012)2

8S

evofl

uran

e, N

2O

IVEn

d of

sur

gery

fol

low

ed

by P

CA

infu

sion

M 5

mg

+ 1

6 m

g on

dans

etro

n in

PC

A/16 m

g on

dans

etro

n in

PC

APA

CU

/PAC

U-6

/6–2

4 h

(PO

NV)

, 0–2

4

h (r

escu

e an

tiem

etic

s)PO

NV

Kim

et

al. (2

013)3

2S

evofl

uran

e, N

2O

IVB

efor

e in

duct

ion

M 7

5 μ

g/kg

+ 0

.3 m

g ra

mos

etro

n /0

.3 m

g ra

mos

etro

n0–2

/2–6

/6–1

2/1

2–2

4 h

(N

VAS

, PO

V, r

escu

e an

tiem

etic

s)N

VAS

> 5

Mak

hdoo

m a

nd F

arid

(2

009)3

3

Isofl

uran

eIV

Bef

ore

indu

ctio

nM

75 μ

g/kg

+ 1

0 m

g de

xam

etha

sone

/ 10 m

g de

xam

etha

sone

/M

75 μ

g/kg

/pla

cebo

0–2

4 h

(PO

N, P

OV,

PO

NV,

re

scue

ant

iem

etic

s)PO

NV

> 2

Park

et

al. (2

013)3

4S

evofl

uran

e,

rem

ifent

anil

IVR

M1: at

indu

ctio

n/R

M2:

end

of s

urge

ryM

50 μ

g/kg

/pla

cebo

0–2

/2–2

4/2

4–4

8 h

(PO

N,

POV,

res

cue

antie

met

ics)

Sev

ere

naus

ea, v

omiti

ng

epis

ode

>2

Ric

hard

son

et a

l. (1

997)3

5

Isofl

uran

e, N

2O

IVB

efor

e in

duct

ion

M 4

0 μ

g/kg

/pla

cebo

0–6

0 m

in (PO

N, r

escu

e an

tiem

etic

s)N

ause

a

Saf

avi a

nd H

onar

man

d (2

009)3

6

Isofl

uran

e, N

2O

IVM

P: b

efor

e in

duct

ion/

MI:

en

d of

sur

gery

M 3

5 μ

g/kg

/pla

cebo

0–2

/2–6

/6–1

2/1

2–1

8/1

8–2

4

h (P

ON

V), 0

–24 h

(re

scue

an

tiem

etic

s)

Nau

sea

>15 m

in,

vom

iting

Shi

rdas

htza

deh

et a

l. (2

011)3

7

Hal

otha

ne, N

2O

IVB

efor

e in

duct

ion

M 5

0 μ

g/kg

/pla

cebo

0–2

/2–6

/6–1

2/1

2–1

8/1

8–2

4 h

(P

ON

, PO

V),0

–24 h

(PO

NV)

Yeo

et a

l. (2

009)3

8Is

oflur

ane,

N2O

IVAt

indu

ctio

nM

75 μ

g/kg

+ 1

0 m

g de

xam

etha

sone

/ 10 m

g de

xam

etha

sone

/pla

cebo

0–2

4 h

(PO

N, P

OV,

PO

NV,

res

cue

antie

met

ics)

Antie

met

ics

requ

ired

for

vom

iting

PCA

= p

atie

nt-c

ontr

olle

d an

alge

sia;

N2O

= n

itrou

s ox

ide;

M =

mid

azol

am;

PON

= p

osto

pera

tive

naus

ea;

POV

= p

osto

pera

tive

vom

iting

; PO

NV

= p

osto

pera

tive

naus

ea a

nd v

omiti

ng;

PAC

U =

pos

tane

sthe

sia

care

uni

t;

NVA

S =

nau

sea

visu

al a

nalo

g sc

ale;

RM

= R

amos

etro

n pl

us m

idaz

olam

; M

P =

Mid

azol

am p

rem

edic

atio

n; M

I = M

idaz

olam

intr

aope

rativ

e.



Effect of Midazolam in Conjunction with Propofol as an Induction Agent on Overall PONSeven studies8,9,12,30,31,35,38 evaluated the effectiveness of mid-azolam in reducing overall PON with the use of propofol as an induction agent, whereas 3 studies33,34,37 assessed its effec-tiveness without propofol. The combined results of both analyses revealed that midazolam caused a statistically sig-nificant reduction in the incidence of overall PON, regard-less of propofol administration (RR, 0.64; 95%, CI 0.47–0.87; I2 = 45%; P

χ2 = 0.09; NNT = 9; n = 7; and RR, 0.39; 95% CI, 0.23–0.65; I2 = 35%; P

χ2 = 0.19; NNT = 5; n = 3, respectively).

Data Collected as Primary or Secondary OutcomesPON incidence was collected as a primary end point in 7 studies,12,29,31,33,34,37,38 whereas 4 studies8,9,30,35 collected PON incidence as a secondary end point. The combined results of both analyses regarding primary and secondary out-comes revealed that midazolam caused a statistically signifi-cant reduction in the overall incidence of PON, regardless of the aim of the study (RR, 0.46; 95% CI, 0.34–0.63; I2 = 5%; P

χ2 = 0.39; NNT = 5; n = 7; and RR, 0.64; 95% CI, 0.44–0.93; I2 = 42%; P


Postoperative VomitingOverall, Early, and Late POVTen studies assessed the effects of midazolam on the overall POV,10,12,29–34,37,38 whereas 4 studies compared the effects on the incidence of early POV,30,32,34,37 and 3 studies evaluated its effects on the incidence of late POV.32,34,37 The combined results revealed that midazolam caused a statistically sig-nificant reduction in the incidence of overall POV (RR, 0.46; 95% CI, 0.33–0.65; I2 = 0%; P

χ2 = 0.87; NNT = 8; n = 10; Fig. 3). Also, the results suggest the beneficial effect of midazolam on the incidence of early or late POV when compared with a placebo, but these did not reach the statistical significance (RR, 0.48; 95% CI, 0.22–1.01; I2 = 0%; P

χ2 = 0.54; NNT = 16; n = 4; and RR, 0.40; 95% CI, 0.13–1.27; I2 = 0%; P


Effect of Administration Timing on Overall POVEight studies10,12,29,30,32–34,37 assessed the effects of mid-azolam on the overall POV when administered at induction, and 2 studies31,34 assessed the effects of mid-azolam administered at the conclusion of surgery. The

Table 4. Effect of Midazolam on PON, POV, and PONV Versus Phases and Administration TimesPhase PON POV PONV

Recovery period Early RR, 0.56; n = 595% CI, 0.17–1.82I2 = 51%; NNT = 7

RR, 0.48; n = 495% CI, 0.22–1.01I2 = 0%; NNT = 16

RR, 0.50; n = 495% CI, 0.32–0.79I2 = 30%; NNT = 6

Late RR, 0.74; n = 295% CI, 0.42–1.31I2 = 49.7%; NNT = 3

RR, 0.40; n = 395% CI, 0.13–1.27I2 = 0%; NNT = 23

RR, 0.26; n = 295% CI, 0.12–0.59I2 = 43%; NNT = 5

Overall RR, 0.51; n = 1195% CI, 0.40–0.65I2 = 42%; NNT = 6

RR, 0.46; n = 1095% CI, 0.33–0.65I2 = 0%; NNT = 8

RR, 0.45; n = 795% CI, 0.36–0.57I2 = 31%; NNT = 3

Administration time Induction RR, 0.53; n = 995%, CI 0.39–0.71I2 = 31%; NNT = 7

RR, 0.45; n = 895% CI, 0.31–0.65I2 = 0%; NNT = 7

RR, 0.44; n = 595% CI, 0.33–0.59I2 = 24%; NNT = 3

End RR, 0.30; n = 295% CI, 0.14–0.66I2 = 0%; NNT = 4

RR, 0.49; n = 295% CI, 0.20–1.19I2 = 0%; NNT = 8

RR, 0.50; n = 395% CI, 0.31–0.78I2 = 48%; NNT = 4

PON = postoperative nausea; POV = postoperative vomiting; PONV = postoperative nausea and vomiting; RR = relative risk; CI = confidence interval; NNT = number needed to treat.

Table 3. Further Data Extracted from the Included Studies

SourceNumber of patients

Sex (male/female) Weight (kg) Height (cm)

Rescue analgesics Airway device

Bauer et al. (2004)29 88 32/56 Not reported Not reported Fentanyl, morphine No limitationGilliland et al. (1996)9 45 0/45 67 (9) Not reported Morphine Endotracheal intubationGodsiff et al. (1995)8 40 23/17 Not reported Not reported Fentanyl, NSAIDs LMA™Hasani et al. (2011)30 90 47/43 57 (11) Not reported Not reported LMA™Heidari et al. (2004)10 82 45/37 68 (14) 170 (8) Not reported Endotracheal intubationHonarmand et al. (2012)6 80 53/27 66.8 (13.8) Not reported Meperidine Endotracheal intubationHuh et al. (2010)31 90 0/90 59.3 (6.3) 159.4 (7.6) Ketorolac, morphine Endotracheal intubationJung et al. (2007)12 90 0/90 58.7 (6.3) 159.1 (6.2) Ketorolac Endotracheal intubationKim et al. (2012)28 90 0/90 57.1 (7.8) 159.0 (4.1) Ketorolac Endotracheal intubationKim et al. (2013)32 94 0/94 60.7 (8.5) 159.3 (8.4) Ketorolac Endotracheal intubationMakhdoom and Farid (2009)33 80 0/80 52.3 Not reported Paracetamol Endotracheal intubationPark et al. (2013)34 126 0/126 58.8 (9.4) 160.6 (5.2) Ketorolac Endotracheal intubationRichardson et al. (1997)35 30 0/30 66 (13) Not reported Morphine sulfate Endotracheal intubationSafavi and Honarmand (2009)36 60 34/26 68.2 (8.7) 166.0 (4.7) Meperidine Endotracheal intubationShirdashtzadeh et al. (2011)37 75 0/75 55 (5.6) Not reported Not reported Endotracheal intubationYeo et al. (2009)38 120 0/120 57.6 (8.3) 161.1 (4.2) Ketorolac Endotracheal intubation

Values of weight and height are mean (SD).LMA™ = laryngeal mask airway; NSAIDs = nonsteroidal antiinflammatory drugs.



combined results revealed that midazolam administered at induction caused a statistically significant reduction in the incidence of overall POV (RR, 0.45; 95% CI, 0.31–0.65; I2 = 0%; P

χ2 = 0.61; I2 = 0%; NNT = 7; n = 8). Also, the effect of midazolam administered at the end of sur-gery showed the tendency of reduction in the incidence of overall POV (RR, 0.49; 95% CI, 0.20–1.19; I2 = 0%; P

χ2 = 0.56; NNT = 8; n = 2).

Effect of Midazolam in Conjunction with Propofol as an Induction Agent on Overall POVFour studies12,30,31,38 assessed the effectiveness of midazolam administered with propofol as an induction agent in pre-venting overall PON, whereas 5 studies14,32–34,37 assessed its effectiveness without propofol. The combined results of both analyses revealed that midazolam caused a sta-tistically significant reduction in the incidence of overall PON, regardless of propofol administration (RR, 0.37; 95% CI, 0.20–0.68; I2 = 0%; P

χ2 = 0.42; NNT = 8; n = 4; and RR, 0.54; 95% CI, 0.35–0.84; I2 = 0%; P


Data Collected as Primary or Secondary OutcomesNine studies10,12,29,31–34,37,38 collected the incidence of POV as a primary end point, whereas only 1 study30 collected POV incidence as a secondary end point. The results of the com-bined analysis revealed that midazolam caused a statisti-cally significant reduction in the overall incidence of POV as a primary outcome (RR, 0.48; 95% CI, 0.34–0.68; I2 = 0%; P

χ2 = 0.90; NNT = 8; n = 9). However, the analysis of POV incidence as a secondary end point could not be performed because of the small number of studies included.

Postoperative Nausea and VomitingOverall, Early, and Late PONVSeven studies assessed the effects of midazolam on overall PONV,6,12,31,33,36–38 4 investigations compared the effects on the incidence of early PONV,6,32,34,36 and 2 studies evaluated the effects on the incidence of late PONV.28,36 The combined results revealed that midazolam caused a statistically significant reduction in the incidence of overall, early, and late PONV (RR, 0.45; 95% CI, 0.36–0.57; I2 = 31%; P

χ2 = 0.16; NNT = 3; n = 7; RR, 0.50; 95% CI, 0.32–0.79; I2 = 30%; P

χ2 = 0.22; NNT = 6; n = 4;

Figure 2. Forest plot for midazolam administered at induction of overall postoperative nausea. The figure depicted individual trials as filled squares with relative size of sample size and solid line as the 95% confidence interval (CI) of the difference. The diamond shape indicates the pooled estimate and uncertainty for the combined effect.

Table 5. Risk of Bias in the Included Randomized Controlled Trials

Biases/references

Random sequence generation

Allocation concealment

Blinding of outcome assessment

Incomplete outcome data

Selective reporting Other bias

Overall risk of bias

Bauer et al. (2004)29 Low risk Unclear Low risk Low risk Low risk Low risk Low riskGilliland et al. (1996)9 Low risk Unclear Low risk Low risk Low risk Low risk UnclearGodsiff et al. (1995)8 Low risk Unclear Low risk Low risk Low risk Low risk UnclearHasani et al. (2011)30 Low risk Unclear Low risk Low risk Low risk Low risk UnclearHeidari et al. (2004)10 Low risk Unclear Low risk Low risk Low risk Low risk UnclearHonarmand et al. (2012)6 Low risk Unclear Low risk Low risk Low risk Low risk UnclearHuh et al. (2010)31 Low risk Unclear Low risk Low risk Low risk Low risk UnclearJung et al. (2007)12 Low risk Unclear Low risk Low risk Low risk Low risk UnclearKim et al. (2012)28 Unclear Unclear Low risk Low risk Low risk Low risk UnclearKim et al. (2013)32 Low risk Low risk Low risk Low risk Low risk Low risk Low riskMakhdoom and Farid (2009)33 Low risk Unclear Low risk Low risk Low risk Low risk UnclearPark et al. (2013)34 Low risk Low risk Low risk Low risk Low risk Low risk Low riskRichardson et al. (1997)35 Low risk Unclear Low risk Low risk Low risk Low risk UnclearSafavi and Honarmand (2009)36 Low risk Unclear Low risk Low risk Low risk Low risk UnclearShirdashtzadeh et al. (2011)37 Low risk Unclear Low risk Low risk Low risk Low risk UnclearYeo et al. (2009)38 Low risk Unclear Low risk Low risk Low risk Low risk Unclear

Method of estimating overall risk of bias: If all results or the above items were “low risk,” the overall risk of bias of the trial was deemed to be low risk of bias. If more than one of the above items were “unclear” or “ high risk,” the overall risk of bias of the trial was deemed to be unclear risk of bias or high risk of bias, respectively.High risk = high risk of bias; low risk = low risk of bias; unclear risk = unclear risk of bias because of lack of detailed reports.



and RR, 0.26; 95% CI, 0.12–0.59; I2 = 43%; Pχ2= 0.17; NNT = 5;

n = 2, respectively; Fig. 4).

Effect of Administration Timing on Overall PONVFive studies12,33,36–38 assessed the effects of midazolam on overall PONV after administration at surgical induction, and 3 investigations6,28,36 assessed its effectiveness at the end of surgery. The combined results indicated that midazolam administration at both the induction and the conclusion of surgery caused statistically significant reductions in the incidence of overall PONV (RR, 0.44; 95% CI, 0.33–0.59; I2 = 24%; P

χ2 = 0.25; NNT = 3; n = 5; and RR, 0.50; 95% CI, 0.31–0.78; I2 = 48%; P


Effects on Overall PONV With or Without Perioperative OpioidsFive studies6,28,31,33,36 compared the effectiveness of mid-azolam in reducing the overall PONV in combination with perioperative opioids, and 3 studies12,37,38 compared its effectiveness without perioperative opioids. The combined results revealed that midazolam administered with periop-erative opioids caused a statistically significant reduction in the incidence of overall PONV (RR, 0.43; 95% CI, 0.32–0.59; I2 = 1%; P

χ2 = 0.42; NNT = 3; n = 5). Also, midazolam admin-istered without perioperative opioids showed beneficial effect on the incidence of overall PONV, but this did not

reach the statistical significance (RR, 0.46; 95% CI, 0.10–2.12; I2 = 66.4%; P

χ2= 0.05; NNT = 3; n = 3).

Effect of Midazolam in Conjunction with Propofol as an Induction Agent on Overall PONVFour studies12,31,36,38 analyzed the effectiveness of midazolam in reducing the overall PONV when used in conjunction with propofol as an induction agent, whereas 4 separate studies6,28,33,38 assessed its effectiveness without the admin-istration of propofol as an induction agent. The combined results of both analyses revealed that midazolam caused a statistically significant reduction in the incidence of over-all PON independent of propofol administration (RR, 0.48; 95% CI, 0.36–0.65; I2 = 9%; P

χ2 = 0.35; NNT = 3; n = 4; and RR, 0.41; 95% CI, 0.29–0.57; I2 = 42%; P


Rescue AntiemeticsEleven studies9,12,28,29,31–36,38 compared the requirement for rescue antiemetics among midazolam recipients and controls. The combined results revealed statistically sig-nificant differences in the requirement for rescue antiemet-ics, but with substantial heterogeneity (RR, 0.52; 95% CI, 0.37–0.74; I2 = 53%; P

χ2 = 0.009; NNT = 5; n = 11). Exclusion of the study by Gilliland et al.9 before a sensitivity analy-sis revealed a decrease in heterogeneity and a statistically

Figure 4. Forest plot for overall postoperative nausea and vomiting. The figure depicted individual trials as filled squares with relative size of sample size and solid line as the 95% confidence interval (CI) of the difference. The diamond shape indicates the pooled estimate and uncertainty for the combined effect.

Figure 3. Forest plot for overall postoperative vomiting. The figure depicted individual trials as filled squares with relative size of sample size and solid line as the 95% confidence interval (CI) of the difference. The diamond shape indicates the pooled estimate and uncertainty for the combined effect.



significant reduction in the requirement for rescue anti-emetics (RR, 0.44; 95% CI, 0.34–0.56; I2 = 0%; P

χ2 = 0.67; NNT = 4; n = 10). In addition, the same outcome was observed when the study by Yeo et al.38 was excluded (RR, 0.57; 95% CI, 0.45–0.71; I2 = 47%; P

χ2 = 0.03; NNT = 5; n = 10).

Effect of Midazolam in Conjunction with Propofol as an Induction Agent in the Requirement for Rescue AntiemeticsThe effectiveness of midazolam with propofol as an induc-tion agent in cases requiring rescue antiemetics was ana-lyzed in 6 studies.9,12,31,35,36,38 Four studies28,32–34 assessed the effectiveness of midazolam without the use of propofol as an induction agent. The combined results of combined analyses indicated that midazolam with propofol as an induction agent showed no evidence of difference in the use of rescue antiemetics (RR, 0.54; 95% CI, 0.024–1.09; P

χ2 = 0.002; I2 = 72%; NNT = 7; n = 6). Conversely, the effects of midazolam on the requirement for rescue antiemetics with-out propofol revealed statistically significant reductions without heterogeneity (RR, 0.50; 95% CI, 0.31–0.79; I2 = 0%; P

χ2 = 0.84; NNT = 7; n = 4).

Data Collected as Primary or Secondary OutcomesSeven studies6,12,31,33,36–38 collected the incidence of PONV as a primary end point, but there were no studies that collected PONV incidence as a secondary end point. The results of the combined analysis revealed that midazolam caused a statistically significant reduction in the overall incidence of PONV as a primary outcome (RR, 0.45; 95% CI, 0.36–0.57; I2 = 31%; P

χ2 = 0.16; NNT = 3; n = 7).

Safety AnalysisHeadacheThe incidence of headaches among midazolam recipi-ents was compared with controls in 4 studies.6,12,34,38 The analysis of the combined findings showed a neutral

result (RR, 1.18; 95% CI, 0.56–2.49; I2 = 0%; Pχ2 = 0.86;

NNT = 1045; n = 4).

DizzinessThe overall effect of midazolam on the incidence of diz-ziness was assessed in 4 studies,12,31,34,38 but there was no statistically significant effect of midazolam relative to the placebo (RR, 0.90; 95% CI, 0.46–1.77; I2 = 0; P

χ2 = 0.79; NNT = 53; n = 4).

SedationThe overall effect of midazolam on the incidence of postop-erative sedation was assessed in 4 studies,9,12,31,34 but there were no statistically significant effect (RR, 1.33; 95% CI, 0.65–2.71; I2 = 58%; P

χ2 = 0.04; NNT = 15; n = 4).

Publication BiasEvidence of a publication bias was detected by the Egger linear regression test (P < 0.05; Table 4). Similarly, the fun-nel plot analyses for these values demonstrated asymme-try (Fig. 5). However, the funnel plots became symmetrical after adjustment using the trim-and-fill method (Fig. 5). The results before and after elimination of the effect of publica-tion bias are shown in Table 6. The significance of the RR with 95% CIs did not change after the elimination of the publication bias.

DISCUSSIONThe results of the current study suggested that midazolam reduced the overall incidence of PON, POV, and PONV, with an NNT of 3 for overall PONV. Furthermore, the results indicated that midazolam treatment can prevent nausea and vomiting in approximately 1 in 3 patients who would otherwise continue to suffer from PONV if adminis-tered a placebo.18 In addition, the need for rescue antiemet-ics was reduced among patients in the midazolam groups

Figure 5. Funnel for overall postoperative nausea. White circles, comparisons included; black circles, inputted comparisons using the trim-and-fill method; white diamond, pooled observed log risk ratio; black diamond, pooled inputted log risk ratio.



compared with control groups, and no evidence found that midazolam was associated with adverse effects, such as headache, dizziness, or sedation.

Midazolam was shown to have a preventive effect on PONV in the early, late, and overall recovery period and was also effective in preventing overall PON and POV. However, we were not able to clarify the effects of mid-azolam on the early and late recovery period of PON or POV. The combined results of early and late PON were considerably heterogeneous, otherwise there was no sub-stantial heterogeneity in POV. It seems that subjective judg-ments of feeling nausea lead to a large variation among collected data, unlike vomiting.

The combined results of 2 studies indicated that mid-azolam appeared to be effective in preventing late PONV. However, 1 of the 2 studies28 administered midazolam as a continuous patient-controlled analgesia infusion, which might have adversely affected the results. Therefore, fur-ther studies are needed to evaluate the duration of the anti-emetic effect. In addition, many patients currently receive antiemetics as opposed to no treatment. Therefore, further studies or a systematic review comparing midazolam with other antiemetics would be needed to confirm the potency of the antiemetic effects of midazolam.

Preoperative anxiety may increase gastric acid secretion, lower gastric pH, and delay gastric emptying, which leads to an increase in PONV.29 However, several studies reported that anxiety was not related to the incidence of PONV.39–41 In addition, the consensus guidelines for the management of PONV42 state that in adults, anxiety is not a proven risk factor and may be of limited clinical relevance. In the study conducted by Safavi and Honarmand,36 the incidence of PONV was significantly lower among subjects who received midazolam at induction compared with those who received the agent at the end of surgery. However, Park et al.,34 reported that no significant differences were observed between administration of midazolam at induction or at the end of surgery. In this meta-analysis, a statistically significant reduction in the incidence of PON, POV, and PONV relative to controls was observed after midazolam administration at induction. However, only the incidence of PON was significantly reduced after administration of midazolam at the end of surgery when compared with the control group. This would imply that the anxiolytic effect of midazolam may be important in its PONV efficacy, which is not in accord with the consensus guidelines.

In the study by Splinter et al.,43 a 50 to 75 μg/kg mid-azolam dose was recommended for prophylactic anti-emetic use. In the current meta-analysis, the doses of midazolam in the reviewed studies ranged from 35 to 75 μg/kg. However, the 35 to 75 μg/kg dosing range might be an excess of quantity for preventing PONV with the risk

of sedation. Therefore, when using midazolam, which is a hypnotic agent, the sedative effects, respiratory depres-sion, and prolonged recovery time might be a major safety concern. However, no significant differences were appar-ent between midazolam and control groups after a safety analysis, but the different sedation scoring scales in the studies limited the combination of data. Fifteen of the 16 relevant studies reported that there were no significant dif-ferences between midazolam and control groups. Among 3 studies in which midazolam was administered continu-ously,9,28,31 only 1 study31 reported that the incidence of mild sedation in the midazolam group was higher than levels in the control group. However, the incidence of moderate to severe sedation or respiratory depression was not observed in any group.

A subgroup analysis was performed to evaluate the effects of midazolam administration on PONV when used in conjunction with propofol as an induction agent. The results of the analysis indicated that midazolam had a sta-tistically significant effect on overall PON, POV, and PONV whether propofol was administered or not. However, the requirement for rescue antiemetics showed no evidence when propofol was used in conjunction with midazolam as an induction agent. Conversely, the requirement for rescue antiemetics revealed a significant reduction compared with the control group when only midazolam was used as an induction agent without propofol. We suggest that the use of propofol, which has been reported to decrease the inci-dence of early PONV,44,45 might have affected the require-ment for rescue antiemetics.

Nonetheless, an apparent publication bias was evi-dent in several comparisons of midazolam and controls in the current study and may have influenced the results. Nevertheless, the significance of the study findings was not altered after consideration of the apparent publication bias. Consequently, we had confidence in our findings, despite the omission of data from unpublished studies.

To include maximum data regarding the antiemetic effects of midazolam, several studies contained PONV data as an inevitable secondary end point, which might have increased the likelihood that confounding variables influ-enced the results. Therefore, we performed a subgroup analysis based on whether PON, POV, and PONV incidence data were collected as primary or secondary outcomes. The combined results of the analyses revealed that midazolam caused a statistically significant reduction in the incidence of overall PON, regardless of the aim of the studies.

The current study has several limitations. First, only pub-lished studies were included in our meta-analysis. Second, only 2 studies included the late effectiveness of midazolam on PONV, one of which included the continuous infusion of midazolam using a patient-controlled analgesia device.

Table 6. Publication BiasEgger test (P value) RR (95% CI) Adjusted RR (95% CI)

Overall PON 0.0200 0.53 (95% CI, 0.42–0.68) 0.73 (95% CI, 0.59–0.91)Early PON 0.0195 0.26 (95% CI, 0.17–1.82) 0.95 (95% CI, 0.38–2.34)Overall POV 0.0107 0.46 (95% CI, 0.33–0.65) 0.45 (95% CI, 0.30–0.68)Early POV 0.0012 0.48 (95% CI, 0.22–1.01) 0.84 (95% CI, 0.42–1.67)Rescue antiemetics 0.0151 0.52 (95% CI, 0.37–0.74) 0.76 (95% CI, 0.63–0.93)

RR = relative risk; PON = postoperative nausea; POV = postoperative vomiting; CI = confidence interval.



Third, many studies either did not assess side effects or adverse events or approached them in different ways. Hence, the late effectiveness of midazolam on PONV and the results of safety analyses should be interpreted cau-tiously. Despite its limitations, the present meta-analysis demonstrated strength through the application of rigor-ous methodology and provided the first systematic review assessing the prophylactic effect of midazolam on the inci-dence of PONV. Moreover, added strengths of our study included the large sample size, the thorough identification of articles, the data extraction, and the resolution of all dis-crepancies by 3 independent investigators. Furthermore, our subgroup and sensitivity analyses yielded stable and robust findings.

In summary, the prophylactic administration of mid-azolam reduced the incidence of overall PON, POV, and PONV. Furthermore, the effects were more evident during the early postoperative phase than during the late phase, and the reductions in the incidence of PON, POV, and PONV were greater when midazolam was administered at the induction of surgery. E

APPENDIXMedline

1. randomized controlled trial.pt2. randomized controlled trial$.mp3. controlled clinical trial.pt4. controlled clinical trial$.mp5. random allocation.mp6. exp double-blind method/7. double-blind.mp8. 8.exp single-blind method/9. single-blind.mp

10. or/1–911. clinical trial.pt12. clinical trial$.mp13. exp clinical trial/14. (clin$ adj25 trial$).mp15. ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind$ or

mask$)).mp16. random$.mp17. exp research design/18. research design.mp19. or/11–1820. 10 or 1921. Case report.tw.22. Letter.pt.23. Historical article.pt.24. Review.pt.25. or/21–2426. 20 not 2527. exp postoperative nausea and vomiting/28. vomit* OR nausea* OR PONV.mp29. (postoperative adj6 nausea adj6 vomiting).mp30. or/27–2931. 26 and 3032. exp midazolam/33. midazolam.mp34. or/32–3335. 31 and 34

EMBASE1. randomi?ed controlled trial$.mp.2. ‘controlled clinical trial’/exp3. controlled AND clinical AND trials4. controlled clinical trial$.mp.5. ‘randomization’/exp6. ‘random allocation’/exp7. random allocation.mp.8. double-blind.mp.9. single-blind.mp.

10. #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9

11. ‘clinical trial (topic)’/exp12. clinical AND trial$.mp.13. random$.mp.14. rct15. #11 OR #12 OR #13 OR #1416. #10 OR #1517. ‘case study’/exp18. ‘case report’/exp19. ‘abstract report’/exp20. ‘letter’/exp21. #17 OR #18 OR #19 OR #2022. #16 NOT #2123. ‘postoperative nausea and vomiting’/exp24. ‘postoperative nausea’/exp25. ‘postoperative vomiting’/exp26. Nausea or vomiting or PONV.mp.27. #23 OR #24 OR #25 OR #2628. #22 AND #2729. ‘midazolam’/exp30. midazolam.mp31. #29 OR #3032. #28 AND #31

DISCLOSURESName: Eun Jin Ahn, MD.Contribution: This author helped conduct the study, analyze the data, and write the manuscript.Attestation: Eun Jin Ahn has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.Name: Hyun Kang, MD, PhD, MPH.Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.Attestation: Hyun Kang has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.Name: Geun Joo Choi, MD.Contribution: This author helped analyze the data and write the manuscript.Attestation: Geun Joo Choi has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.Name: Chong Wha Baek, MD, PhD.Contribution: This author helped conduct the study and ana-lyze the data.Attestation: Chong Wha Baek has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.Name: Yong Hun Jung, MD, PhD.



Contribution: This author helped design the study and write the manuscript.Attestation: Yong Hun Jung has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.Name: Young Choel Woo, MD, PhD.Contribution: This author helped conduct the study and ana-lyze the data.Attestation: Young Choel Woo has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.This manuscript was handled by: Tong J. Gan, MD, MHS, FRCA.

REFERENCES 1. Gan TJ. Postoperative nausea and vomiting—can it be elimi-

nated? JAMA 2002;287:1233–6 2. Tramèr MR, Reynolds DJ, Moore RA, McQuay HJ. Efficacy,

dose-response, and safety of ondansetron in prevention of postoperative nausea and vomiting: a quantitative systematic review of randomized placebo-controlled trials. Anesthesiology 1997;87:1277–89

3. Henzi I, Walder B, Tramèr MR. Dexamethasone for the preven-tion of postoperative nausea and vomiting: a quantitative sys-tematic review. Anesth Analg 2000;90:186–94

4. Khalil S, Philbrook L, Rabb M, Wells L, Aves T, Villanueva G, Amhan M, Chuang AZ, Lemak NA. Ondansetron/prometha-zine combination or promethazine alone reduces nausea and vomiting after middle ear surgery. J Clin Anesth 1999;11:596–600

5. Henzi I, Walder B, Tramèr MR. Metoclopramide in the preven-tion of postoperative nausea and vomiting: a quantitative sys-tematic review of randomized, placebo-controlled studies. Br J Anaesth 1999;83:761–71

6. Honarmand A, Safavi M, Khalili G, Mohammadnejad F. Prophylactic administration of haloperidol plus midazolam reduces postoperative nausea and vomiting better than using each drug alone in patients undergoing middle ear surgery. Saudi J Anaesth 2012;6:145–51

7. Yoon SH. Concerns of the anesthesiologist: anesthetic induction in severe sepsis or septic shock patients. Korean J Anesthesiol 2012;63:3–10

8. Godsiff L, Magee L, Park GR. Propofol versus propofol with midazolam for laryngeal mask insertion. Eur J Anaesthesiol Suppl 1995;12:35–40

9. Gilliland HE, Prasad BK, Mirakhur RK, Fee JP. An investiga-tion of the potential morphine sparing effect of midazolam. Anaesthesia 1996;51:808–11

10. Heidari SM, Saryazdi H, Saghaei M. Effect of intravenous mid-azolam premedication on postoperative nausea and vomiting after cholecystectomy. Acta Anaesthesiol Taiwan 2004;42:77–80

11. Hvarfner A, Hammas B, Thörn SE, Wattwil M. The influence of propofol on vomiting induced by apomorphine. Anesth Analg 1995;80:967–9

12. Jung JS, Park JS, Kim SO, Lim DG, Park SS, Kwak KH, Cho JD, Jeon YH. Prophylactic antiemetic effect of midazolam after mid-dle ear surgery. Otolaryngol Head Neck Surg 2007;137:753–6

13. Kranke P, Eberhart LH. Possibilities and limitations in the phar-macological management of postoperative nausea and vomit-ing. Eur J Anaesthesiol 2011;28:758–65

14. Heidari SM, Rahimi M, Saeb F, Shafa A. Comparison of the effect of intravenous midazolam, metoclopramide and their combination on the treatment of postoperative nausea and vomiting: a double blinded randomized clinical trial. Pak J Med Sci 2011;27:847–50

15. Lee Y, Wang JJ, Yang YL, Chen A, Lai HY. Midazolam vs ondan-setron for preventing postoperative nausea and vomiting: a randomised controlled trial. Anaesthesia 2007;62:18–22

16. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol 2009;62:e1–34

17. Wu SJ, Xiong XZ, Lin YX, Cheng NS. Comparison of the efficacy of ondansetron and granisetron to prevent postoperative nau-sea and vomiting after laparoscopic cholecystectomy: a system-atic review and meta-analysis. Surg Laparosc Endosc Percutan Tech 2013;23:79–87

18. Naing C, Aung K, Mak JW. Reporting ‘number needed to treat’ in meta-analyses: a cross-sectional study. J Evid Based Med 2012;5:232–7

19. Clyburn P, Kay NH, McKenzie PJ. Effects of diazepam and midazolam on recovery from anaesthesia in outpatients. Br J Anaesth 1986;58:872–5

20. Crawford ME, Carl P, Bach V, Ravlo O, Mikkelsen BO, Werner M. A randomized comparison between midazolam and thio-pental for elective cesarean section anesthesia. I. Mothers. Anesth Analg 1989;68:229–33

21. El-Deeb A, Ali Y, Rashdy H. Evaluation of combination anti-emetic prophylaxis in high risk emetogenic patients undergo-ing thyroid surgery: a randomized double-blind study. Egypt J Anaesth 2011;27:203–6

22. Fragen RJ, Caldwell NJ. Awakening characteristics following anesthesia induction with midazolam for short surgical proce-dures. Arzneimittelforschung 1981;31:2261–3

23. Puri GD, Palvannan S, Chari P, Singh H, Gujral JS. Midazolam as an induction agent in mitral stenosis patients for closed mitral commissurotomy. Indian J Med Res 1993;98:174–7

24. Quario Rondo L, Thompson C. Efficacy of propofol compared to midazolam as an intravenous premedication agent. Minerva Anestesiol 2008;74:173–9

25. Sanjay OP, Tauro DI. Midazolam: an effective antiemetic after cardiac surgery—a clinical trial. Anesth Analg 2004;99:339–43

26. Di Florio T, Goucke CR. The effect of midazolam on persistent postoperative nausea and vomiting. Anaesth Intensive Care 1999;27:38–40

27. Lohakare R, Sangawar AV, Ghosh AA. Influence of intravenous fentanyl and / or midazolam on induction of anaesthesia with thiopentone. J Anaesthesiol Clin Pharmacol 2004;20:273–8

28. Kim DS, Koo GH, Kang H, Baek CW, Jung YH, Woo YC, Kim JY, Park SG. The antiemetic effect of midazolam or/and ondansetron added to intravenous patient controlled anal-gesia in patients of pelviscopic surgery. Korean J Anesthesiol 2012;62:343–9

29. Bauer KP, Dom PM, Ramirez AM, O’Flaherty JE. Preoperative intravenous midazolam: benefits beyond anxiolysis. J Clin Anesth 2004;16:177–83

30. Hasani A, Maloku H, Sallahu F, Gashi V, Ozgen SU. Preemptive analgesia with midazolam and diclofenac for hernia repair pain. Hernia 2011;15:267–72

31. Huh BK, Jung S, White W, Jeon Y. Anti-emetic effect of mid-azolam added to morphine patient-controlled analgesia after total abdominal hysterectomy. Anaesth Intensive Care 2010;38:481–5

32. Kim WJ, Kang H, Shin HY, Baek CW, Jung YH, Woo YC, Kim JY, Koo GH. Ramosetron, midazolam, and combination of ramo-setron and midazolam for prevention of postoperative nausea and vomiting: a prospective, randomized, double-blind study. J Int Med Res 2013;41:1203–13

33. Makhdoom NK, Farid MF. Prophylactic antiemetic effects of midazolam, dexamethasone, and its combination after middle ear surgery. Saudi Med J 2009;30:504–8

34. Park EY, Lee SK, Kang MH, Lim KJ, Kim YS, Choi E, Park YH. Comparison of ramosetron with combined ramosetron and midazolam for preventing postoperative nausea and vomiting in patients at high risk following laparoscopic gynaecological surgery. J Int Med Res 2013;41:654–63

35. Richardson MG, Wu CL, Hussain A. Midazolam premedication increases sedation but does not prolong discharge times after brief outpatient general anesthesia for laparoscopic tubal steril-ization. Anesth Analg 1997;85:301–5

36. Safavi MR, Honarmand A. Low dose intravenous midazolam for prevention of PONV, in lower abdominal surgery—pre-operative vs intraoperative administration. Middle East J Anaesthesiol 2009;20:75–81

37. Shirdashtzadeh N, Eshraghi N, Eshraghi A. Comparison of par-enteral promethazine versus midazolam effect as a preoperative



medication on postoperative nausea and vomiting after appen-dectomy. Caspian J Intern Med 2011;2:270–3

38. Yeo J, Jung J, Ryu T, Jeon YH, Kim S, Baek W. Antiemetic effi-cacy of dexamethasone combined with midazolam after mid-dle ear surgery. Otolaryngol Head Neck Surg 2009;141:684–8

39. Haavik PE, Søreide E, Hofstad B, Steen PA. Does preopera-tive anxiety influence gastric fluid volume and acidity? Anesth Analg 1992;75:91–4

40. Lydon A, McGinley J, Cooke T, Duggan PF, Shorten GD. Effect of anxiety on the rate of gastric emptying of liquids. Br J Anaesth 1998;81:522–5

41. Wang SM, Kain ZN. Preoperative anxiety and postopera-tive nausea and vomiting in children: is there an association? Anesth Analg 2000;90:571–5

42. Gan TJ, Diemunsch P, Habib AS, Kovac A, Kranke P, Meyer TA, Watcha M, Chung F, Angus S, Apfel CC, Bergese SD, Candiotti

KA, Chan MT, Davis PJ, Hooper VD, Lagoo-Deenadayalan S, Myles P, Nezat G, Philip BK, Tramèr MR; Society for Ambulatory Anesthesia. Consensus guidelines for the man-agement of postoperative nausea and vomiting. Anesth Analg 2014;118:85–113

43. Splinter WM, MacNeill HB, Menard EA, Rhine EJ, Roberts DJ, Gould MH. Midazolam reduces vomiting after tonsillectomy in children. Can J Anesth 1995;42:201–3

44. Tramèr M, Moore A, McQuay H. Propofol anaesthesia and post-operative nausea and vomiting: quantitative systematic review of randomized controlled studies. Br J Anaesth 1997;78:247–55

45. Visser K, Hassink EA, Bonsel GJ, Moen J, Kalkman CJ. Randomized controlled trial of total intravenous anesthesia with propofol versus inhalation anesthesia with isoflurane-nitrous oxide: postoperative nausea with vomiting and eco-nomic analysis. Anesthesiology 2001;95:616–26

Documents

The Effectiveness of Midazolam for Preventing ...sbmu.ac.ir/uploads/2015- the effectiveness of midazolam.pdf · Comprehensive meta-analysis software (version 2.0, Biostat, Englewood,