Dexamethasone for Antiemesis in Laparoscopic.27

Dexamethasone for Antiemesis inLaparoscopic Gynecologic SurgeryA Systematic Review and Meta-Analysis

Alice Pham, MD, FRCSC, and Grace Liu, MD, FRCSC

OBJECTIVE: To estimate the beneficial and harmful

effects of dexamethasone for prevention of postopera-

tive nausea and vomiting in women undergoing laparo-

scopic gynecologic surgery.

DATA SOURCES: We searched the following bibliographic

databases: MEDLINE (from 1946 to January 2012), Embase

(from 1980 to 2012 week 3), the Cochrane Central Register

of Controlled Trials (from inception to January 2012), and

ISI Web of Knowledge (from 1950 to January 2012). We also

screened trial registries, reference lists of retrieved studies,

and other sources of unpublished literature.

METHODS OF STUDY SELECTION: Two reviewers

screened in duplicate and independently searched results

for inclusion. We included randomized controlled trials

(RCTs) comparing dexamethasone with a placebo in

patients undergoing laparoscopic gynecologic surgery.

TABULATION, INTEGRATION, AND RESULTS: Two

reviewers completed data extraction and assessed trials

for bias in duplicate and independently. We used the

Grading of Recommendations Assessment, Development,

and Evaluation methodology to assess the quality of

evidence across studies at the outcome level. A total of

13 RCTs with 1,695 patients met inclusion criteria. Data

were pooled based on the random-effects model. The use

of prophylactic dexamethasone significantly decreases the

incidence of postoperative nausea (relative risk [RR] 0.56,

95% confidence interval [CI] 0.45–0.71), postoperative

vomiting (RR 0.35, 95% CI 0.25–0.48), the need for rescue

antiemetics (RR 0.39, 95% CI 0.29–0.52), and time to meet

discharge criteria (mean 28.5 minutes, 95% CI 24.6–32.4).

The estimated number needed to treat to prevent nausea

in one patient was eight (95% CI 5–13), whereas that for

vomiting was five (95% CI 4–6). There was no observed

increase in adverse events. The quality of the evidence

ranged from very low to moderate.

CONCLUSION: This systematic review provides evidence

that dexamethasone decreases the incidence of postoper-

ative nausea and vomiting after laparoscopic gynecologic

surgery, with no observed increase in side effects.

(Obstet Gynecol 2012;120:1451–58)

DOI: http://10.1097/AOG.0b013e31827590f3

The incidence of postoperative nausea and vomitingis significant after laparoscopic gynecologic surgery,

affecting between 50% and 90% of patients.1 This islikely related to the increased intra-abdominal pressurerequired in laparoscopic surgery, and it is likely relatedto the fact that female gender confers a two-fold to four-fold increase in risk.1,2 Three other main risk factors forpostoperative nausea and vomiting have been identi-fied: history of motion sickness or postoperative nauseaand vomiting; nonsmoking status; and use of postoper-ative opioids.1,3 A variety of antiemetics with differentmechanisms of action have been used both in combi-nation and as single agents, including anticholinergics,dopamine receptor antagonists, and antihistamines.

Dexamethasone is a corticosteroid known for itsantiemetic effects, particularly in chemotherapy-inducednausea. Its mechanism of action, however, remainsunclear. A systematic review found that in patientsundergoing laparoscopic cholecystectomy, dexametha-sone, as compared with placebo, reduced the incidenceof nausea (relative risk [RR] 0.59, 95% confidenceinterval [CI] 0.48–0.72), vomiting (RR 0.41, 95% CI0.30–0.55), and postoperative nausea or vomiting(RR 0.55, 95% CI 0.78–0.98).4 Postoperative pain alsoappeared to be reduced (ratio of means 0.87, 95% CI

From the Department of Obstetrics and Gynecology, Sunnybrook Health SciencesCentre, Toronto, Ontario, Canada.

The authors thank Elie Akl for instruction, advice, and editorial support in thepreparation of this manuscript.

Corresponding author: Alice Pham, MD, FRCSC, Department of Obstetrics andGynecology, Sunnybrook Health Sciences Centre, Suite B627, 2075 BayviewAvenue, Toronto, ON, M4N 3M5, Canada; e-mail: [email protected].

Financial DisclosureThe authors did not report any potential conflicts of interest.

© 2012 by The American College of Obstetricians and Gynecologists. Publishedby Lippincott Williams & Wilkins.ISSN: 0029-7844/12

VOL. 120, NO. 6, DECEMBER 2012 OBSTETRICS & GYNECOLOGY 1451

0.78–0.98); however, there was a considerableamount of unexplained heterogeneity (I 2590.4%).The incidence of side effects (headache and dizziness)was found to be similar between the treatment groups.The objective of this systematic review was to esti-mate the beneficial and harmful effects of dexameth-asone compared with no dexamethasone in thegynecologic surgical patient population.

SOURCES

We searched the following bibliographic databases:MEDLINE (from 1946 to third week of January2012), Embase (from 1980 to third week of 2012),the Cochrane Central Register of Controlled Trials(from inception to third week of January 2012), andISI Web of Knowledge (from 1950 to third week ofJanuary 2012) for abstracts. We screened the elec-tronically available conference proceedings from theSociety of Obstetricians and Gynecologists of Canada(2006–2011), the American College of Obstetriciansand Gynecologists (2008–2011), the American Asso-ciation of Gynecologic Laparoscopists (www.aagl.org;2008–2011), the Canadian Anesthesiologists Society(www.cas.ca; 2007–2011), and the American Societyof Anesthesiologists (ASA; www.asahq.org; 2000–2011). We also screened trial registries, reference listsof included studies, and contacted clinical experts.There were no limits on language, date, or form ofpublication.

We used the following search terms to search alldatabases: laparoscopy; endoscopy; video-assisted sur-gery; gynecologic surgical procedures; steroids; anddexamethasone and all of its generic and trade names.We used a study filter for randomized controlled trials,from the Cochrane handbook, and limited our searchto “human studies.” See Appendix 1 (available onlineat http://links.lww.com/AOG/A333) for the full searchstrategy.

We developed a protocol with inclusion cri-teria and methods of analysis and registered iton PROSPERO: International prospective registerof systematic reviews.5 The registration numberis CRD42012002160 and is available online(http://www.crd.york.ac.uk/PROSPERO).

STUDY SELECTION

Two reviewers (A.P. and G.L.) screened in duplicateand independently all titles and abstracts from thesearch results. If either reviewer identified a study asbeing potentially eligible, the full-text article wasretrieved for review. Both reviewers, in duplicate andindependently, performed full-text screening for eligi-bility. We resolved all disagreements by consensus.

Agreement between the two reviewers was calculatedusing Cohen weighted kappa statistic (kw).

We reviewed published randomized controlledtrials in which the intervention was perioperativecorticosteroid administration compared with placebo inpatients undergoing laparoscopic gynecologic surgery.Quasi-randomized trials, essentially those trials lackingrandom allocation, were eligible for extraction of data onoutcomes not assessed by randomized controlled trials.6

Trials in which laparoscopic procedures were convertedto open were included in the review. Studies with a rangeof doses, route (oral, intravenous, intramuscular, inhala-tional), and timing of administration (preoperative, intra-operative, or postoperative) were included. Primaryoutcomes of interest were postoperative nausea, vomit-ing, or both. Secondary outcomes included adverseevents (infection, hyperglycemia, admission to the inten-sive care unit, reoperation, mortality), need for postop-erative rescue antiemetics, degree of postoperative pain(measured on any scale), need for postoperative analge-sics, conversion to laparotomy, quality of life (measuredon any validated scale), and hospital length of stay. Weused a piloted standardized data extraction form andinstruction manual to abstract data independently andin duplicate. We resolved disagreements through discus-sion and consensus and contacted authors if furtherinformation was required.

Data extracted from each study included relevantinformation on study identifiers and characteristics oftrial participants, including age, body mass index (BMI,calculated as weight (kg)/[height (m)]2), ASA class, his-tory of postoperative nausea and vomiting, and motionsickness; the trial’s exclusion and inclusion criteria;characteristics of the intervention (including type, dose,route, and timing); type of comparator (no interventioncompared with placebo); and type of outcome measure.

Both reviewers assessed in duplicate and indepen-dently, in an unblinded manner, the methodologicquality using the Cochrane risk of bias tool. The qualityindicators included adequacy of sequence generation,allocation concealment, blinding of participants, per-sonnel and outcome assessors, selective outcome report-ing, extent of loss to follow-up, stopping early forbenefit, and adherence to intention-to-treat analysis.

Summary measures for dichotomous data (post-operative nausea or vomiting, rates of adverse events,need for rescue antiemetics and analgesics) wereexpressed as RRs. Summary measures for continuousdata were expressed as mean differences (postoperativepain, length of hospital stay, quality of life). If differentscales were used to measure a continuous variable,the standardized mean difference was calculated. Thestandardized mean difference standardizes the results

1452 Pham and Liu Dexamethasone for Antiemesis in Gynecologic Surgery OBSTETRICS & GYNECOLOGY

to a uniform scale before the data can be pooled.6

When medians, ranges, or both were reported, thesewere converted to means and standard deviations, orboth, to pool the results for the meta-analysis.7

Statistical analyses were conducted using the pro-gram Review Manager 5.1 software.8 We tested forheterogeneity using the statistical tests x2 and I 2. TheI 2 statistic measures the percentage of the variability ineffect estimates that is attributable to heterogeneityrather than sampling error.6 Results were combinedin a meta-analysis using the random-effects modelbecause it provided a more conservative effect esti-mate. For continuous variables, the inverse variancemethod was used, which essentially calculates a pointestimate by weighting each study by the inverse vari-ance of its treatment effect estimate.6 The Mantel-Haenszel method was used to pool dichotomousvariables. Different doses of dexamethasone were com-bined into a single treatment arm to allow for pair-wisecomparison with control.6

We used the Grading of Recommendations Assess-ment, Development, and Evaluation approach to assess

the quality of evidence by outcome across studies.9 Theapproach takes into consideration the following factors:type of study design; indirectness; imprecision; incon-sistency; and publication bias. We evaluated the possi-bility of publication bias by constructing Begg andEgger inverted funnel plots, which show asymmetryin the presence of evident publication bias.10

Subgroup analyses, specified a priori, were per-formed for outcomes if significant heterogeneity wasfound to determine whether an interaction waspresent. Subgroup analyses were planned for age,ASA classification, BMI, comorbidities (specifically,history of motion sickness or postoperative nauseaand vomiting), dose, timing of administration, coin-terventions, definition of nausea, vomiting, pain, andduration of surgery. Sensitivity analyses were prespe-cified. Studies with the highest risk of bias wereexcluded and analysis was repeated to determinewhether heterogeneity could be reduced. In addition,sensitivity analysis also was performed to estimate theeffect on outcome from excluding studies that did notreport mean, standard deviation, or both.

Records identifi ed through database searching

n=621

Excluded studies: n=12Comparator not placebo, or no intervention: 8

Published only in abstract format: 3

Records identifi ed through other sources

n=3

Records screenedn=410

Records after duplicates removed

N=410

Full-text articles assessed for eligibility

n=25

Records excludedn=385

RCT not included; did not report postoperative

nausea or vomiting events in dexamethasone group

n=1

Studies included in qualitative synthesis

n=13

Studies included in quantitative synthesis

n=12

Author under investigation for alleged research misconduct: 1

Fig. 1. Flow of included studies. RCT,randomized controlled trial.

Pham. Dexamethasone for Antiemesis inGynecologic Surgery. Obstet Gynecol2012.

VOL. 120, NO. 6, DECEMBER 2012 Pham and Liu Dexamethasone for Antiemesis in Gynecologic Surgery 1453

RESULTS

The search strategy identified a total of 624 citations.After removing duplicates, 410 citations remained.Title and abstract screening excluded 384 of thesecitations because they did not meet inclusion criteria.Of the 26 trials that remained, we were not able toobtain the full text for one of these.11 The full-textscreening of the remaining 25 citations identified 14eligible studies, but one trial was excluded becausethe author is currently undergoing investigation foralleged research misconduct12,13 (Fig. 1). The agree-ment between the two reviewers was complete (Cohenkw51) for full-text screening.

Characteristics of included studies are providedin Table 1 in Appendix 2 (available online athttp://links.lww.com/AOG/A334). All 13 studieswere randomized controlled trials. Eleven werepublished in English and two included English trans-lations from Portuguese. The studies involved a total of1,695 patients undergoing laparoscopic gynecologicsurgery. All studies used dexamethasone as a prophy-lactic antiemetic, administered in a single dose. Doses(4–15 mg) and timing (preoperatively 2 hours beforesurgery to administration at the end of surgery) variedbetween studies. All studies used a placebo control.

All 13 studies measured the incidence of post-operative nausea or vomiting.14–26 One study assessedpostoperative nausea or vomiting but did not report thenumber of events in the dexamethasone group.16 Sixstudies14,15,19,20,24,26 reported on side effects, one ofwhich24 examined blood sugar elevations as well aswound infection and delayed wound healing. Postop-erative pain scores, as measured by a visual analogscale or numeric rating scale, were reported in eightstudies.15,17,19–22,24,26 Quality of life was reported inone study.17 Need for postoperative rescue antiemeticswas reported in three studies.19,21,26 Need for postoper-ative analgesia was reported in six studies.15,17,20,22,24,26

Three studies reported on conversion to laparot-omy.15,17,20 Total hospital length of stay was notreported in any study; however, time to meet dischargecriteria was reported in three studies.17,22,24

Studies varied regarding their risk of bias. Theirmethodologic quality is described in Table 1 inAppendix 2 (http://links.lww.com/AOG/A334).Sequence generation, the method used to generatethe random allocation sequence, was adequate in eightstudies.14,15,17,19,20,22,24,25 Allocation concealment wasunclear in three studies.18,21,23 Two studies16,18 did notreport blinding in any of the five groups that couldpotentially introduce bias (patients, health care pro-viders, data collectors, outcome assessors, data ana-

lysts). Selective reporting bias was difficult to assessbecause no study was included in a trial registry orhad published protocols. None of the studies clearlystated an intention-to-treat analysis and five studies fol-lowed a per-protocol analysis,15,17,19,20,25 thereby expos-ing them to further potential bias. All studies had morethan 95% follow-up and none was stopped early forbenefit.

We analyzed postoperative nausea and vomit-ing as two separate outcomes and not as a compositeto avoid the risk of double-counting. Ninetrials14,15,17–19,21,23,25,26 reported postoperative nausea.The pooled estimate showed a statistically significantreduction in nausea in patients treated with dexametha-sone compared with placebo (RR 0.56, 95% CI 0.45–0.71; I 250%) (Fig. 2). The estimated number needed totreat to prevent nausea in one patient was eight (95% CI5–13). The quality of evidence was moderate.

Ten trials14,15,17–21,23,25,26 reported postoperativevomiting. The pooled analysis showed a statisticallysignificant decrease in vomiting in patients receivingdexamethasone compared with placebo (RR 0.35,95% CI 0.25–0.48; I 250%) (Fig. 3). The estimatednumber needed to treat to prevent vomiting in onepatient was five (95% CI 4–6). The quality of evidencewas low. Subgroup analysis did not rule out a clini-cally significant reduction in the incidence of postop-erative vomiting with increasing doses ofdexamethasone (P5.68) (Fig. 4), or with administra-tion of dexamethasone at the time of induction ofanesthesia (P5.10).

Five trials15,19,21,25,26 reported on the need for a res-cue antiemetic, either ondansetron or metoclopramide,administered intravenously. The pooled analysisshowed a significant reduction in the need for a rescueantiemetic postoperatively (RR 0.39, 95% CI 0.29–0.52; I 250%). The quality of evidence was moderate.

Four trials15,17,22,26 reported on the need for a rescueanalgesic. The pooled analysis did not exclude a clini-cally significant reduction in the need for a postoperativerescue analgesic (RR 0.92, 95% CI 0.74–1.14). Therewas a high degree of heterogeneity (I 2555%; P5.08)that could not be explained by any of the planned sub-group analyses. The quality of evidence was low.

Three trials21,22,26 measured postoperative painscores, using either a visual analog scale or a numericrating scale. The pooled effect estimate suggested adecrease in the pain score of 1.03 (standardized meandifference; 95% CI 0.01–2.07), but we did not excludean absence of effect. The quality of evidence wasvery low. There was a high degree of heterogeneity(I 2593%; P,.001) that could not be explained byany of the planned subgroup analyses. Subgroup


analysis based on exclusion of patients with BMI high-er than 30–35 was statistically significant with the testfor subgroup differences (x259.30; df51; P5.002).Studies that excluded patients with BMIs higher than30–35 had a decrease in pain score of 0.51 (95% CI0.41–1.42); those studies that had no BMI exclusioncriteria had a greater decrease in pain score of 2.13(95% CI 1.63–2.62).

Two trials17,20 reported on conversion to laparot-omy. The pooled estimate did not exclude an increasein the rates of conversion (RR 0.69, 95% CI 0.05–10.27). There was a high degree of heterogeneity(I 2566%; P5.09) that could not be explored by sub-group analysis or sensitivity analysis because of thelimited number of studies with this outcome.

No trials reported total length of hospital stay.However, time to meet discharge criteria was reportedin three trials,17,22,24 and the pooled analysis showed thatpatients treated with dexamethasone decreased time to

meet discharge criteria by 28.5 minutes (95% CI 24.6 –

32.4; I 250%). The quality of evidence was moderate.Only one study assessed quality of life17 using the

Quality of Recovery 40 questionnaire, a validated toolto evaluate patients after anesthesia and surgery in thedomains of physical comfort, emotional state, physicalindependence, psychological support, and pain. Scoresranged from 40 to 200, with higher scores representinghigher quality of recovery. The mean score of patientstreated with dexamethasone was 15.5 points higherthan that of placebo patients (95% CI 13.1–18.0).

One study24 reported on hyperglycemia andwound infection, and there were zero events. Of theremaining five studies14,15,19,20,26 that reported on otherside effects, two14,15 stated that there was no differenceamong the treatment groups. One study19 foundno side effects related to the use of dexamethasone.Two studies20,26 presented numerical data and foundno increase in the incidence of adverse effects.

Fig. 2. Pooled analysis of nausea across studies.

Pham. Dexamethasone for Antiemesis in Gynecologic Surgery. Obstet Gynecol 2012.

Fig. 3. Pooled analysis of vomiting across studies.



The quality of evidence for all outcomes rangedfrom very low to moderate. A summary of findingstable based on the Grading of RecommendationsAssessment, Development, and Evaluation method-ology can be found in Table 2 in Appendix 2(http://links.lww.com/AOG/A334). The funnel plotfor postoperative vomiting suggests positive publica-tion bias (Appendix 3 [available online at http://links.lww.com/AOG/A335], which illustrates theinverted funnel plot for trials comparing the effect ofdexamethasone and placebo on postoperative vomit-ing), whereas for other outcomes there were too few

studies to allow for appropriate interpretation of publi-cation bias.

Subgroup analyses specified a priori were per-formed. There was no difference from the results ofpooled estimates. There were not enough trials withsignificant cointerventions or that defined nausea, vom-iting, or pain to perform a subgroup analysis. Mean ageacross the trials was similar in the treatment groups and,hence, a subgroup analysis was not performed.

Sensitivity analyses were prespecified. There werefour studies18,21,23,26 that did not report or that hadunclear randomization sequence generation, three18,21,23

Fig. 4. Comparison of postoperative vomiting with subgroup analysis of dose of dexamethasone.



of which had unclear allocation concealment or didnot report allocation concealment. Two studies16,18 didnot report on blinding. When those studies that did notreport or that had unclear sequence generation wereexcluded, there was no difference in results from overallpooled estimates across all outcomes. For postoperativepain score, when the studies that did not report mean,standard deviation, or both (requiring the use of estima-tion formulas to estimate the mean and variance fromthe median and range) were excluded, only one studyremained, thereby precluding the pooling of results.The results of the remaining study, however, were con-sistent with the overall results. For the outcome of timeto meet discharge criteria, there was no difference in theresults.

This systematic review identified 13 randomizedcontrolled trials that evaluated the beneficial and harm-ful effects of dexamethasone, as compared with placebo,in patients undergoing laparoscopic gynecologic sur-gery. Dexamethasone showed a substantial beneficialeffect in reducing the incidence of both postoperativenausea and vomiting. In addition, it decreased the timeto meet discharge criteria by 29 minutes and signifi-cantly decreased the need for a rescue antiemetic.Subgroup analysis did not rule out a clinically significantreduction in postoperative vomiting in patients treatedwith higher doses of dexamethasone (8 mg) comparedwith lower doses (4–5 mg). Those studies that reportedon side effects either had no events or found no increasein the incidence of adverse events among all treatmentgroups. These side effects were observed in the first 24hours postoperatively and also may have been attribut-able to a multitude of factors, such as the general anes-thetic agents.

In terms of postoperative pain scores and the needfor rescue analgesia, the data are less certain. Thissystematic review failed to demonstrate whether dexa-methasone had the potential to reduce the severity ofpostoperative pain, or whether it had any opioid-sparing effect. Dexamethasone patients had betterquality of recovery, as measured by the Quality ofRecover 40 questionnaire, in all five domains ofphysical comfort, emotional state, physical indepen-dence, psychological support, and pain, scoring almost16 points higher than the placebo group. Althoughno minimum clinically important difference could befound in the literature, it has been stated that a 10-pointdifference in the Quality of Recovery 40 questionnairescore is typical of that seen in patients with and withouta major postoperative complication, or when compar-ing minor and major surgical procedures.27

This review has several limitations, most of whichare related to the methodologic quality of the included

studies. For the outcomes of need for rescue analgesicand postoperative pain score, and conversion to lapa-rotomy, the analysis is limited by unexplained moderateto substantial heterogeneity. In addition, for postopera-tive vomiting, there was a suggestion of positivepublication bias from the funnel plot. Our efforts tocontact authors of unpublished trials and non-Englishpublications were unsuccessful. Intention-to-treat analy-sis was either unclear or not performed in any of thestudies. This systematic review also includes patientsreceiving dexamethasone in various doses, at varioustimes, and undergoing a variety of laparoscopic gyneco-logic procedures, all of different surgical duration. It alsoincludes patients of all ages. However, this does confergreater generalizability to the findings of this review.

Strengths of this review include the comprehen-sive search strategy as well as the rigorous methodol-ogy and systematic application of eligibility criteria. Athorough assessment of study quality and risk of biaswas performed, in addition to the use of standardizeddata extraction forms. A priori hypotheses to explainheterogeneity were generated and tested. Althoughthere was the possibility of publication bias for post-operative vomiting, there was a relatively smallnumber of studies involved overall.

CONCLUSION

In clinical practice, it is clear that prophylacticdexamethasone decreases the incidence of postoper-ative nausea and vomiting after laparoscopic gyneco-logic surgery, relative to placebo. Given this currentevidence, surgeons should consider routinely usingprophylactic dexamethasone in this patient popula-tion, in the absence of any contraindications.

This review did not examine the evidence for theefficacy of dexamethasone compared with those ofother commonly used antiemetics for postoperativenausea and vomiting in this surgical population.Future research should be directed at determiningwhether dexamethasone is effective in reducingpostoperative pain. Given that opioids contribute topostoperative nausea and vomiting, it also would beof clinical benefit if dexamethasone were shown todecrease the need for postoperative opioid use.Larger, high-quality, randomized controlled trialspowered to detect a difference in this outcome alsowould strengthen the evidence for using this corti-costeroid in an ambulatory surgical setting. Anotherimportant outcome to study would be whetherdexamethasone use in postmenopausal comparedwith premenopausal women leads to different overalleffects for postoperative nausea and vomiting. Theoptimal dose and timing of administration also


remain unclear. Finally, more studies examiningoverall quality of recovery are warranted.

REFERENCES1. McCracken G, Houston P, Lefebvre G. SOGC Clinical Practice

Guideline. Guideline for the management of postoperative nauseaand vomiting. J Obstet Gynaecol Can 2008;30:600–7, 608–16.

2. Raphael JH, Norton AC. Antiemetic efficacy of prophylacticondansetron in laparoscopic surgery: randomized, double-blindcomparison with metoclopramide. Br J Anaesth 1993;71:845–8.

3. Wolfgang K, Gassmayr S. Pre-operative anxiety, stress and pre-medication. Baillières Clin Anesth 1998;12:2536–4.

4. Karanicolas PJ, Smith SE, Kanbur B, Davies E, Guyatt GH. Theimpact of prophylactic dexamethasone on nausea and vomitingafter laparoscopic cholecystectomy: a systematic review andmeta-analysis. Ann Surg 2008;248:751–62.

5. PROSPERO: International prospective register of systematicreviews. 2012. Available at: www.crd.york.ac.uk/PROSPERO/.Retrieved March 16, 2012.

6. Higgins JPT, Green S, editors. Cochrane handbook for system-atic reviews of interventions. Version 5.1.0. The Cochrane Col-laboration; 2011. Available at: www.cochrane-handbook.org.Retrieved February 17, 2012.

7. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean andvariance from the median, range, and the size of a sample.BMC Med Res Methodol 2005;5:13.

8. Review Manager (RevMan) [computer program]. Version 5.1.Copenhagen (Denmark): The Nordic Cochrane Centre, TheCochrane Collaboration; 2011.

9. Schünemann H, Brozek J, Oxman A, editors. GRADE handbookfor grading quality of evidence and strength of recommendation.Version 3.2. The GRADE Working Group; 2009. Available at:http://www.cc-ims.net/gradepro. Retrieved March 9, 2012.

10. Egger M, Davey Smith G, Schneider M, Minder C. Bias inmeta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34.

11. Laiq N, Khan MN, Qureshi FA, Khan S, Jan AS. Dexametha-sone as antiemetic during gynaecological laparoscopic surgery.J Coll Physicians Surg Pak 2005;15:778–81.

12. Fujii Y, Nakayama M. Dexamethasone for reduction of nausea,vomiting and analgesic use after gynecologic laparoscopic sur-gery. Int J Gynaecol Obstet 2008;100:27–30.

13. Scott JR. Expression of concern. Available at: http://journals.lww.com/greenjournal/Documents/ExpressionOfConcern.pdf.Retrieved September 20, 2012.

14. Biswas BN, Rudra A, Mandal SK. Comparison of ondanse-tron, dexamethasone, ondansetron plus dexamethasone andplacebo in the prevention of nausea and vomiting after lapa-roscopic tubal ligation. J Indian Med Assoc 2003;101:638,640, 642.

15. Chu CC, Shieh JP, Tzeng JI, Chen JY, Lee Y, Ho ST, et al. Theprophylactic effect of haloperidol plus dexamethasone on post-operative nausea and vomiting in patients undergoing laparos-copically assisted vaginal hysterectomy. Anesth Analg 2008;106:1402–6.

16. De Abreu MP, Vieira JL, da Silva IF, Miziara LEPG, Fofano R.Efficacy of ondansetron, metoclopramide, droperidol and dexa-methasone in preventing post-gynecologic videolaparoscopynausea and vomiting in outpatient setting: comparative study.Rev Bras Anestesiol 2006;56:8–15.

17. De Oliveira GS, Ahmad S, Fitzgerald PC, Marcus RJ,Altman CS, Panjwani AS, et al. Dose ranging study on the effectof preoperative dexamethasone on postoperative quality ofrecovery and opioid consumption after ambulatory gynaecolog-ical surgery. Br J Anaesth 2011;107:362–71.

18. Ganem EM, Fukushima FB, da Silva DSM, Nakamura G,Castiglia YMM, Vianna PTG. Efficacy of propofol and propo-fol plus dexamethasone in controlling postoperative nausea andvomiting of gynecologic laparoscopy. Rev Bras Anestesiol2002;52:394–401.

19. Huang J, Shieh J, Tang C, Tzeng J, Koung-Shing C, Wang J.Low-dose dexamethasone effectively prevents postoperativenausea and vomiting after ambulatory laparoscopic surgery.Can J Anesth 2001;48:973–7.

20. Jokela RM, Ahonen JV, Tallgren MK, Marjakangas PC,Korttila KT. The effective analgesic dose of dexamethasone afterlaparoscopic hysterectomy. Anesth Analg 2009;109:607–15.

21. Lee Y, Lai H, Lin PC, Huang SJ, Lin YS. Dexamethasoneprevents postoperative nausea and vomiting more effectivelyin women with motion sickness. Can J Anaesth 2003;50:232–7.

22. Mohammadi SS. Effects of dexamethasone on early postoper-ative pain, nausea and vomiting and recovery time after ambu-latory laparoscopic surgery. J Med Sci 2007;7:1202–5.

23. Rimaitis K, Svitojute A, Macas A. The influence of dexameth-asone and ketolgan on postoperative nausea and vomiting andestimation of risk factors in women undergoing gynecologiclaparoscopic surgeries. Medicina (Kaunas) 2010;46:261–7.

24. Thangaswamy CR, Rewari V, Trikha A, DehranM, Chandralekha.Dexamethasone before total laparoscopic hysterectomy: a random-ized controlled dose-response study. J Anesth 2010;24:24–30.

25. Wang JJ, Ho ST, Liu HS, Ho CM. Prophylactic antiemeticeffect of dexamethasone in women undergoing ambulatory lap-aroscopic surgery. Br J Anaesth 2000;84:459–62.

26. Yuksek MS, Alici HA, Erdem AF, Cesur M. Comparison ofprophylactic antiemetic effects of ondansetron and dexametha-sone in women undergoing day-case gynaecological laparo-scopic surgery. J Int Med Res 2003;31:481–8.

27. Myles PS, Weitkamp B, Jones K, Melick J, Hensen S. Validityand reliability of a postoperative quality of recovery score: theQoR-40. Br J Anaesth 2000;84:11–5.
