Perioperative gabapentin and post cesarean pain control: A systematic review and meta-analysis of randomized controlled trials

Review

article

Perioperative

gabapentin

and

post

cesarean

pain

control:

A

systematic

review

and

meta-analysis

of

randomized

controlled

trials

Laura

Felder

,

Gabriele

Saccone

,

Sergio

Scuotto

,

David

T.

Monks

,

Jose

C.A.

Carvalho

,

Fulvio

Zullo

,

Vincenzo

Berghella

*

a

DepartmentofObstetricsandGynecology,SidneyKimmelMedicalCollegeofThomasJeffersonUniversity,Philadelphia,PA,USA

b

DepartmentofNeuroscience,ReproductiveSciencesandDentistry,SchoolofMedicine,UniversityofNaplesFedericoII,Naples,Italy

c

DepartmentofAnesthesiology,SchoolofMedicine,UniversityofSiena,Siena,Italy

d_{DepartmentofAnesthesia,WashingtonUniversitySchoolofMedicine,St.Louis,MO,USA} e

DepartmentofAnesthesiaandPainManagement,MountSinaiHospital,UniversityofToronto,Toronto,Canada

f

DivisionofMaternal-FetalMedicine,DepartmentofObstetricsandGynecology,SidneyKimmelMedicalCollegeofThomasJeffersonUniversity,Philadelphia, PA,USA

ARTICLE INFO Articlehistory:

Received26October2018

Receivedinrevisedform26November2018 Accepted27November2018

Availableonlinexxx Keywords: Cesareandelivery Gabapentin

Postoperativepaincontrol

ABSTRACT

Cesareandeliveryoccursinroughlyonethirdofpregnancies.Effectivepostoperativepaincontrolisagoal forpatientsandphysicians.Limitingopioiduseinthisperiodisimportantassomepercentageofopioid naïveindividualswilldeveloppersistentuse.Gabapentinisanon-opioidmedicationthathasbeenused perioperativelytoimprovepostoperativepainandlimitopioidrequirements.Thegoalofthisstudyisto determinetheefficacyofperioperativegabapentininimprovingpostcesareandeliverypaincontrol.The followingdata sourcesweresearchedfromtheirinceptionthrough October2018:MEDLINE,Ovid, ClinicalTrials.gov,Sciencedirect,andtheCochraneLibraryattheCENTRALRegisterofControlledTrials.A systematicreviewoftheliteraturewasperformedtoincludeallrandomizedtrialsexaminingtheeffectof perioperativegabapentinonpostcesareandeliverypaincontrolandotherpostoperativeoutcomes.The primaryoutcomewasanalgesiceffectofgabapentinonpostcesareandeliverypain,measuredbyvisual analogscale(VAS;0–100)orNumericalRatingScale(NRS;0–10)onmovement24hpostoperative.These scores weredirectlycomparedby multiplyingall NRSscores bya factorof10.Meta-analysiswas performedusingtherandomeffectsmodelofDerSimonianandLaird,toproducesummarytreatment effectsintermsofmeandifference(MD)with95%confidenceinterval(CI).Sixplacebocontrolledtrials (n=645)wereidentifiedasrelevantandincludedinthemeta-analysis.Allstudiesincludedonlyhealthy pregnantwomen(AmericanSocietyofAnesthesiologist(ASA)physicalstatusIorII)undergoingspinal anesthesiaforcesareandeliveryatterm.Participantswererandomizedtoeither600mgoralgabapentin orplacebopreoperativelyandinonestudythemedicationswerealsocontinuedpostoperatively.Pooled datashowedthatwomenwhoreceivedgabapentinpriortocesareandeliveryhadsignificantlylowerVAS painscoresat24haftermovement(MD-11.58,95%CI-23.04to-0.12).VASpainscoresatothertime intervalsatrestoraftermovementwerenotsignificantlydifferentforthosewhoreceivedgabapentin and placebo although there was a general trend toward lower pain scores forwomen receiving gabapentin.Therewasnosignificantbetween-groupdifferenceinuseofadditionalpainmedications, supplementalopioids,andmaternalorneonatalsideeffects.Therewashigherpaincontrolsatisfactionat 12and24hinthegabapentinversusplacebogroups.

©2018ElsevierB.V.Allrightsreserved.

*Correspondingauthorat:DivisionofMaternal-FetalMedicine,DepartmentofObstetricsandGynecology,ThomasJeffersonUniversity,833ChestnutStreet,FirstFloor, Philadelphia,PA19107,USA.

E-mailaddress:[email protected](V.Berghella).

https://doi.org/10.1016/j.ejogrb.2018.11.026

0301-2115/©2018ElsevierB.V.Allrightsreserved.

ContentslistsavailableatScienceDirect

European

Journal

of

Obstetrics

&

Gynecology

and

Reproductive

Biology

Contents

Introduction ... 99

Materialandmethods ... 99

Sources... 99

Studyselection ... 100

Riskofbias ... 100

Primaryandsecondaryoutcomes ... 103

Statisticalanalysis ... 103

Results ... 103

Studyselectionandstudycharacteristics ... 103

Synthesisofresults ... 105 Discussion ... 105 Conclusions... 106 Disclosure ... 106 Financialsupport ... 106 References... 106 Introduction

Cesareandeliveryoccursinapproximately32%,ofpregnancies andisoneofthemostcommonlyperformedsurgicalproceduresin theUnitedStatesamongreproductiveagedwomen[1].Itinvolvesa largeabdominalincision,sopost-operativepaincontrolisaconcern forpatientsandtheirfamilies.Poorpaincontrolinthepostpartum periodhasbeenshowntocontributetopostpartumdepressionand tobeariskfactorfordevelopingchronicpainsyndromes[2–5].

Regionalanesthesiaforcesareandeliverytypicallyconsistsof intrathecalinjectionoflocalanestheticandalipophilicopioid.The addition of intrathecal morphine has been shown to provide analgesiainthepostoperativeperiodfor 14_–36handtoreduce postoperativeopioidrequirements[5].Regimensforpostoperative pain control typically include combinations of acetaminophen, nonsteroidal anti-inflammatory drugs (NSAIDs) and systemic opioid medications. These may be administered intravenously, intramuscularly,orally,orrectally[3,5].

In addition to optimal postoperative pain control, limiting opioiduseisafurthergoalforhealthcareprovidersasapercentage ofopioidnaivewomenwillexhibitpersistentusepostoperatively [6]. As the number of individuals with opioid use disorder continuestorise,alternativeanalgesicoptionsand strategiesto limitopioidconsumptionbecomeincreasinglyimportant.

Gabapentinisananticonvulsantmedicationapprovedbythe FederalDrugAdministration(FDA)forepilepsytreatmentaswell as neuropathic pain conditions. It works in partby decreasing excitatory neurotransmitter action in the nervous system to dampen afferent nocioceptive signaling. It has also been used perioperativelyforavarietyofsurgeriestodecreasepostoperative pain and opioid consumption with varying efficacy. A meta-analysisbyAlayedetal.showedreducedpostoperativepainscores aswellasdecreasedopioidrequirementsinthefirst24hfollowing hysterectomy when used perioperatively [7]. Gabapentin is generallyconsidered safe for use in pregnancy, withonly rare reportsofneonatalwithdrawalassociatedwithprolongeduse[8]. Theaimof this systematicreviewand meta-analysiswas to determinetheefficacyofperioperativegabapentininimproving postoperativepaincontrolinhealthywomenundergoingcesarean deliveryattermunderspinalanesthesia.

Materialandmethods Sources

Thisreviewwasperformedaccordingtoaprotocoldesigneda priori and recommended for systematic review [9]. Electronic

databases (MEDLINE, Ovid, ClinicalTrials.gov, Sciencedirect, the Cochrane Libraryat theCENTRAL Register of Controlled Trials) were searched fromtheir inception untilOctober 2018. Search termsusedwerethefollowingtextwords:“cesarean,”“caesarean,” “postoperative,” “analgesia,” “pain” and “gabapentin.” No

Fig.1.Flow diagram of studies identified in the systematicreview. (Prisma template[PreferredReportingItemforSystematicReviewsandMeta-analyses]).

restrictionsforlanguageorgeographiclocationwereapplied.In addition,thereferencelistsofallidentifiedarticleswereexamined to identify studies not captured by electronic searches. The electronicsearchandtheeligibilityofthestudieswere indepen-dentlyassessedbytwoauthors(LF,GS).Differenceswerediscussed andconsensusreached.

Studyselection

Thisstudyincludedallrandomizedcontrolledtrialsexamining theeffectofperioperativegabapentinonpostcesareandelivery analgesia and other postoperative outcomes including nausea, vomiting,and pain related to spinal headache. Eligible studies includedtrialscomparingpatientsreceivingatleastonedoseof gabapentinwithacontrolgroup(eitherplaceboornotreatment).

We included only studies in which women with singleton gestations received spinal anesthesia and underwent cesarean deliveryatterm(37weeks). Trialsincludingmultiple gestations, preterm delivery, general anesthesia, studies comparing gabapentin to another drug, and quasi-randomized trials (i.e. trials in which allocationwasdoneonthebasisofapseudo-randomsequence,e.g. odd/evenhospitalnumber or date ofbirth,alternation) were excluded. Riskofbias

Theriskofbiasineachincludedstudywasassessedusingthe criteriaoutlinedintheCochraneHandbookforSystematicReviewsof Interventions.Sevendomainsrelatedtoriskofbiaswereassessed ineachincludedtrialsincethereisevidencethattheseissuesare associatedwithbiasedestimatesoftreatmenteffect:1)random

Fig.2. Assessmentofriskofbias.(A)Summaryofriskofbiasforeachtrial;Plussign:lowriskofbias;minussign:highriskofbias;questionmark:unclearriskofbias.(B)Risk ofbiasgraphabouteachriskofbiasitempresentedaspercentagesacrossallincludedstudies.

Table1

Characteristicsoftheincludedtrials.

Studylocation Samplesize*

Inclusioncriteria Exclusioncriteria Primaryoutcome Mooreetal.[11] Canada 44

(21vs23) Term Scheduledcesarean Age18 HIV Hepatitis UncontrolledHTNorDM IVdruguser

Fetalcongenitalabnormalities Useofpainmedicationinthepast

week

VASpainonmovementat 24hrspostop

Shortetal.[12] Canada 88 (44vs44) Term Scheduledcesarean Singleton Epilepsy CNSormentaldisorders Chronicpain Drugabuse

Useofneuropathicanalgesicor antiepilepticdrugs

Fetalcongenitalabnormalities

VASpainonmovementat 24hourspostop

Nofaletal.[13] Egypt 86 (42vs44) Term Scheduledcesarean Singleton Primiparous Multiparous Chronicheadaches Chronicpain

Regularanalgesicsorantiepileptic medicationuse

Fetalcongenitalabnormalities

NotreportedPost-dural punctureheadache characteristics

Memarietal.[14] Iran 200 (100vs100)

Hemoglobin>10g/dl Fasting6-8hourspriorto

surgery

Cardiovasculardisease GIdisease

Middleeardisease Vertigo

Motionsickness DM

HTN

Smokingoralcoholconsumption Nausea/vomitingbeforesurgery Fever/infection

Receivedadditionalpostoperative medicationswithin6hrs(only anti-bioticsandsedatives)

VASscorefornauseaand vomitingpostop

Monksetal.[15] Canada 197 (100vs97) Term Scheduledcesarean Ages18-55 Epilepsy Chronicpain

Useofanticonvulsantsor neuro-pathicanalgesics

OpioidorIVdrugabuser Useofantacidinprevious3hours

VASpainonmovementat 24hpostop

Hafezetal.[16] Egypt 45 (15vs15) Term Scheduledcesarean 20-40yrsold Uncomplicatedpregnancy Epilepsy

Routineuseofantiepileptic medi-cations

Kidneyorliverimpairment Alcoholism/IVdruguse HTN Oligohydramnios/polyhydramnios Antepartumhemorrhage Psychiatricdisorder Inabilitytocommunicate NRSpainscore

HIV:humanimmunodeficiencyvirus;HTN:hypertension;DM:diabetesmellitus;CNS:centralnervoussystem;GI:gastrointestinal.

*

Dataarereportedastotalnumber(numberintheinterventionvsnumberinthecontrolgroup).

Table2

Characteristicsoftheincludedwomen.

Maternalage Gestationalageatrandomization BMI(kg/m2₎

Mooreetal.[11] 355vs346 38.80.7vs38.90.8 294vs306 Shortetal.[12] 34.84.1vs35.34.8 38.7vs38.5*

30.65.6vs29.34.3 Nofaletal.[13] 32.14.8vs30.75.2 Notreported Notreported Memarietal.[14] 26.15.2vs26.15.2 38.20.5vs38.10.4 Notreported Monksetal.[15] 35.93.9vs34.74.5 38.70.8vs38.60.9 30.85.1vs31.25.6 Hafezetal.[16] 28.24.7vs27.35.5 38.31.1vs38.11.0 Notreported Dataarereportedasmeanstandarddeviation.

*

sequence generation;2) allocation concealment; 3) blindingof participantsandpersonnel;4)blindingofoutcomeassessment;5) incompleteoutcomedata;6)selectivereporting;and7)otherbias. Reviewauthors’judgmentswerecategorizedas“lowrisk”,“high risk”or“unclearrisk”ofbias[9].

Twoauthors(LF,GS)independentlyassessedinclusioncriteria, riskofbiasanddataextraction.Disagreementswereresolvedby discussion.

Allanalysesweredoneusinganintention-to-treatapproach, evaluatingwomenaccordingtothetreatmentgrouptowhichthey

Table3

Preoperativeandintraoperativepainmanagement. Intervention group Control group Timingof intervention

Anesthesiaintraoperativemedicationsreceivedbyboth interventionandcontrolgroups

Additionalintraoperativemeds asneeded

Mooreetal.[11] 600mgoral gabapentin

Lactose placebo

1hrpriorto surgery

Intrathecal:0.75%hyperbaricbupivacaine12mg,Fentanyl 10mg,Morphine100mg

Upto100mgIVfentanylas needed

Shortetal.[12] 600mgoral gabapentin

Lactose placebo

1hrpriorto surgery

Intrathecal:13.5mg0.75%hyperbaricbupivacaine,10mg fentanyl,100mgmorphine

IVfentanylatdiscretionof anesthesiologist Nofaletal.[13] 600mgoral

gabapentin

Starch placebo

2hrspriorto surgery

Intrathecal:12.5mghyperbaricbupivacaine0.5%,25mg fentanyl

None Memarietal.[14] 600mgoral

gabapentin

Placebo 1hrpriorto surgery

Spinal Notreported

Monksetal.[15] 600mgoral gabapentin

Placebo 1hrpriorto surgery

Intrathecal:1.6-1.8ml0.75%hyperbaricbupivacaine,10mg fentanyl,100mgmorphine

Upto100mgfentanylIV Hafezetal.[16] 600mgoral

gabapentin

Placebo 1hrpriorto surgery

Intrathecal:8mgbupivacaine,25mgfentanyl None

Hr:hour;IV:intravenous.

Table4

Intraoperativecharacteristics.

Skinincision Repeatedcesarean Durationofsurgery(minutes) Mooreetal.[11] Notreported 18/21(85.7%)vs15/23(65.2%) Notreported

Shortetal.[12] Pfannenstiel 30/42(71.4%)vs33/42(78.6%) Notreported Nofaletal.[13] Notreported 26/42(61.9%)vs24/44(54.5%) 55.76.9vs57.87.3 Memarietal.[14] Notreported Notreported Notreported Monksetal.[15] Pfannenstiel 76/100(76.0%)vs70/97(72.2%) Notreported Hafezetal.[16] Notreported Notreported Notreported Dataarereportedasmeanstandarddeviationorasnumberpercentage(numberintheinterventionvsnumberinthecontrolgroup).

Table5

Immediatepostoperativemedicationsusedinbothgroups. NSAID– immediately postop Acetaminophen-immediately postop

Postoperativepainmeds– standing

Postoperativepainmeds–asneeded Anti-emetics,anti-itch

Mooreetal.[11] Ketorolac 30mgIV Acetaminophen 1gperrectum Diclofenac50mgq8hpo Acetaminophen1gpoq6hfor 72hrs

PACU:Morphine2mgIVq5minsprn Ward:Morphine2mgsubcutaneous q4hrsinfirst24hrs,then5mgpoq4 hrs

Dimenhydrinate50IMq6h Diphenhydramine25IMq4h

Shortetal.[12] Ketorolac30 mgIV Acetaminophen 1300mgper rectum Diclofenac50mgq8hpo Acetaminophen1gq6hpox72 hrs

First24hrs:Morphine2mg subcu-taneousprn After:Morphine10mgpo Dimenhydrinate50mgIMq6h prn Nalbuphine5mgsubcutaneous q4hprn

Nofaletal.[13] Notreported Notreported Diclofenac50mgIM alternat-ingwithIVAcetaminophenq 12hours

Meperidine50mgIMprn Allreceivedantacids

Memarietal.[14] Notreported Notreported Notreported Notreported Notreported Monksetal.[15] Ketorolac30

mgIV Acetaminophen 1300mgper rectum Diclofenac50mgq8h Acetaminophen1gpoq6h Gabapentin200mgorplacebo q8x48hrs

PACU:Morphine2mgIVq5mins Floor:

First24h:Morphinesubcutaneousor IV2mgorHydromorphone0.4mg q1h After24h:Morphine10mgpoor Hydromorphone2mg Zofran Metoclopramide Dimenhydrinate

Hafezetal.[16] Notreported Notreported NotReported MeperidineIV1mg/kgIMinfirst24 hrs

Zofran Ranitidine

wererandomly allocatedin theoriginaltrials. Thisreview was registeredwiththePROSPEROInternationalProspectiveRegister ofSystematicReviews(RegistrationNumber:CRD42018099832). Primaryandsecondaryoutcomes

TheprimaryoutcomewastheVASpainscoreonmovementat 24hpostoperative.SecondaryoutcomeswereVASpainscoresat other time intervals both at rest or on movement following surgery, use of additional intraoperative pain medications or supplementalopioids, pain control satisfaction, persistent pain after cesarean delivery, maternal side effects, and neonatal outcomes. Maternal side effects examined include nausea, vomiting,pruritus, and sedation. Neonatal outcomes evaluated includeAPGARsat1and5min,birthweight,umbilicalcordarterial pH, neonatal intensive care unit admissions, breast feeding difficultiesinfirst24h,andneedforpositivepressureventilation. Statisticalanalysis

Thedataanalysiswascompletedindependentlybytwoauthors (LF,GS)usingReviewManagerv.5.3(TheNordicCochraneCentre, CochraneCollaboration,2014,Copenhagen,Denmark).The com-pleted analyses were then compared, and any difference was resolvedbydiscussionwithathirdreviewer(VB).

Data from each eligible study were extracted without modification of originaldataonto custom-madedatacollection forms. A 2 by 2 table was assessed for relative risk (RR); for continuousoutcomes,meansstandarddeviationwereextracted andimportedintoReviewManagerv.5.3.

Meta-analysiswasperformedusingtherandomeffectsmodel ofDerSimonianandLaird,toproducesummarytreatmenteffects in terms of either a RR or mean difference (MD) with 95% confidence interval (CI). Heterogeneity was measured using I-squared(HigginsI2_).

Potential publicationbiases wereassessed statisticallyusing Begg’sandEgger’stests.Pvalue<0.05wasconsideredstatistically significant.

The meta-analysis was reported following the Preferred Reporting Item for Systematic Reviews and Meta-analyses (PRISMA)statement[10].

Results

Studyselectionandstudycharacteristics

TheflowofstudyidentificationisshowninFig.1.Sixtrialswere identifiedasrelevant,includedinthemeta-analysisandanalyzed [11–16].Publication bias, assessedstatisticallyusingBegg’s and Egger’s test, showed no significant bias (P=0.57 and P=0.52, respectively). Statistical heterogeneity within the trials ranged fromlowtomoderatewithnoinconsistency(I2=0%)forseveralof thesecondaryoutcomes,andI2_{=89%fortheprimaryoutcome.}

The quality of the RCTs included in our meta-analysis was assessed using the seven criteria outlined in the Cochrane Handbookfor Systematic ReviewsofInterventions.Mostof the includedstudieswerejudgedas“lowrisk”ofbiasinmostofthe sevenCochranedomainsrelatedtotheriskofbias.Alltheincluded studieshad“lowrisk”ofbiasin“randomsequencegeneration.”All the trials were placebo-controlled studies, and neither the

Fig.3. ForestplotfortheriskofVASpainscoresatpostoperativeonmovement.

Table6

VASPainscoresoutcomes. VAS6h movement (mm) VAS6hrest (mm) VAS12h Movement (mm) VAS12h Rest(mm) VAS24h movement (mm) VAS24h rest(mm) VAS48h movement(mm) VAS48hrest (mm)

Mooreetal.[11] NotReported NotReported 19.9516.87 vs 45.0024.39 11.8610.36vs 22.6120.23 20.9517.38vs 41.3922.08 12.0513.88vs 18.7015.53 17.2917.96vs 32.0921.67 7.911.94vs 14.4117.91 Shortetal.[12] 31.523.3vs 39.526.9 15.315.1vs 19.320.9 32.722.5vs 38.723.6 15.614.9vs 16.715.9 37.821.8vs 34.823.5 15.316.0vs 14.315.6 26.519.2vs 34.324.1 13.212.8vs 13.716.9 Nofaletal.[13] NotReported NotReported NotReported NotReported NotReported NotReported NotReported NotReported Memarietal.[14] NotReported NotReported NotReported NotReported NotReported NotReported NotReported NotReported Monksetal.[15] NotReported NotReported NotReported NotReported 40.022.5vs

46.823.1 13.116.7vs 19.118.1 34.021.5vs 36.024.5 13.015.3vs 16.017.0 Hafezetal.[16] 42.0-8.1 vs29.09.0 NotReported 30.08.0 vs23.016.0 NotReported 30.07.2 vs52.08.2

NotReported NotReported NotReported Total 36.8vs34.3 15.3vs19.3 27.5vs35.6 13.7vs19.7 32.2vs43.7 13.5vs17.4 25.9vs34.1 11.4vs14.7 I2 91% N/A 88% 64% 89% 38% 47% 0% MD(95%CI) 2.99( 17.57to 23.54) 4.00( 11.80 to3.80) 7.60( 24.97 to9.77) 5.32( 14.71to 4.06) 11.60( 23.03 to-0.16) 3.90( 8.62to 0.82) 6.90( 13.92to 0.13) 3.03( 6.45to 0.38) Dataarereportedasmeanstandarddeviationintheinterventionvscontrolgroup.

Table 7

Secondary outcomes.

Women requiring additional intraoperative pain medications

Average mg of IV morphine

equivalents used as needed in 24 hours for patients requiring additional analgesia Women requiring supplemental narcotics in first 24 hrs Pain control satisfaction 6 hrs (mm) Pain control satisfaction 12 hrs (mm) Pain control satisfaction 24 hrs (mm) Pain control satisfaction 48 hrs (mm) Persistent pain 6 weeks after delivery

Moore et al. [11] 0/21 vs 1/23 (4.3%) 4.2 2.5 vs 3.2  1.8 5/21 (23.8%) vs 5/23 (21.7%) Not reported 94.8 8.1 vs

79.1 20.0 90.15.5 vs 77.8 17.8 92.9 11.0 vs 87.3 12.4 2/16 (12.5%) vs 4/20 (20.0%) Short et al. [12] 1/42 (2.4%) vs 3/42 (7.1%) 6.7 3.6 vs 7.9  3.8 14/42 (33.3%) vs 16/42 (30.1%) 84.0 20.0 vs 82.0 21.0 84.7 17.7 vs 78.2 22.6 85.3 15.1 vs 80.5 21.6 81.0 24.0 vs 83.8 18.5 6/42 (14.3%) vs 1/42 (2.4%)

Nofal et al. [13] 0/42 vs 0/44 Not reported Not reported Not reported Not reported Not reported Not reported Not reported

Memari et al. [14] Not reported Not reported Not reported Not reported Not reported Not reported Not reported Not reported

Monks et al. [15] 1/100 (1.0%) vs 3/97 (3.1%) 8.7 5.2 vs 9.1  6.7 23/100 (23) vs 31/97 (32) Not reported Not reported 86.5 17.5

vs 77 22.9

85.5 19.3 vs 81.3 21.4

4/71(5.6%) vs 3/71 (4.2%)

Hafez et al. [16] 0/15 vs 0/15 Not reported 15/15 (100%) vs 15/15

(100%)

Not reported Not reported Not reported Not reported Not reported

Total 2/205 (1.0%) vs 7/206 (3.4%) 6.5 vs 6.7 57/178 (32.0%) vs 67/177 (37.9%) 84.0 vs 82.0 89.8 vs 78.7 87.3 vs 78.4 – 12/129 (9.3%) vs 8/ 133 (6.0%) I [2] 0% 58% 60% Not applicable 54% 0% 13% 34% RR or MD (95% CI) 0.32 (0.08 to 1.38) 0.14 (-1.48 to 1.20) 0.90 (0.62 to 1.29) 2.00 ( 6.57 to 10.57) 11.00 (1.99 to 20.02) 8.64 (4.47 to 12.81) 3.20 ( 1.09 to 7.48) 1.47 (0.45 to 4.79) Data are reported as mean standard deviation or as number (percentage) in the intervention vs control group. Boldface data, statistically significant.

MD, mean difference; CI, confidence interval; RR, relative risk.

Table 8

Maternal Side effects within 48 h.

Nausea Vomiting Pruritus Sedation

Moore et al. [11] 12/21 (57.1%) vs 8/23 (34.8%) 5/21 (23.8%) vs 3/23 (13.0%) 16/21 (76.2%) vs 22/23 (95.7%) 17/21 (80.9%) vs 17/23 (73.9) Short et al. [12] 19/42 (45.2%) vs 19/42 (45.2%) 7/42 (16.7%) vs 10/42 (23.8%) 33/42 (78.6%) vs 32/42 (80.9%) 23/42 (54.8%) vs 24/42 (57.1%) Nofal et al. [13] 7/42 (16.6%) vs 9/44 (20.5%)

Not reported Not reported 11/42 (26.2%) vs 3/44 (6.8%)

Memari et al. [14] 27/100 (27.0%) vs 41/100 (41.0%) 25/100 (25.0%) vs 38/100 (38.0%) Not reported Not reported

Monks et al. [15] 38/100 (38.0%) vs 42/97 (43.3%) 24/100 (26.0%) vs

26/97 (26.8%)

72/100 (72.0%) vs 77/97 (79.4%) 55/100 (55.0%) vs

38/97 (39.2%)

Hafez et al. [16] Not reported Not reported Not reported Not reported

Total 103/305 (33.8%) vs 119/309 (38.5%) 61/263 (23.2%) vs 77/262 (29.4%) 121/163 (74.2%) vs 131/162 (80.9%) 106/205 (51.7%) vs 82/206 (39.8%)

I [2] 30% 0% 9% 57%

RR (95% CI) 0.90 (0.69 to 1.17) 0.78 (0.58 to 1.04) 0.91 (0.81 to 1.03) 1.24 (0.90 to 1.69)

Data are reported as number in the intervention vs number in the control group. RR, relative risk; CI, confidence interval.

4 L. Felder et al. / Eur opean Journal of Obstetrics & Gynecology and Repr oductive Biology 233 (20 1 9 ) 98 – 10 6

participantsnorthe investigatorswereaware of thetreatment assignments(Fig.2).

All sixtrials included onlyhealthy pregnantwomenwith singleton gestations undergoing cesarean delivery at term underregionalanesthesia.HealthywasdefinedasASAIorII, whichincludespregnantwomenwithBMIlessthan40mg/kg, with no medical problems or well controlled diabetes/ hypertension,ormildlungdisease.Therewereatotalof320 women in the intervention group and 321 women in the control group. VAS pain score was the most used primary outcome,assessedinfourofsixtrialsandatotalof178women intheinterventiongroupand177womeninthecontrolgroup had available data for this outcome (Table 1). The mean gestationalageatrandomizationwasabout38weeksinboth groups (Table 2). All trials used 600mg oral gabapentin as intervention,andplaceboascontrol,1h(in5/6trials)before surgeryandinonestudygabapentinwascontinued postoper-ativelyfor48h(Table3).Bothgabapentinandplacebogroups receivedthesameintrathecalopioidsineachstudy(Table3). Onlyfourtrialsreported onrepeatcesarean(themajority of women),andonlytwoontypeofskinincision(Pfannenstiel) (Table4).Immediatepostoperativemedicationsusedinboth groups–an NSAID,usuallyketorolac ordiclofenac,aswellas acetaminophen-arereportedinTable5.

Synthesisofresults

Womenwhoreceived600mgoralgabapentinpriortocesarean delivery, had lower VAS pain scores at 24h postoperative on movementcomparedtothosewhoreceivedplacebo(36.4vs43.7, MD-11.60,95%CI-23.03to-0.16;Fig.3)(Table6).VASpainscores at6, 12,24,and48hatrestoraftermovementwerenotstatistically differentforwomen receivinggabapentinandplaceboalthough there was a general trend towards lower VAS pain scores for womenreceivinggabapentin(Table6).Therewasnosignificant between-groupdifferenceinuseofadditionalpainmedications, supplementalopioids,andpersistentpain6weeksaftercesarean delivery(Table7).Therewasahigherpaincontrolsatisfactionat12 and24hinthegabapentinvsplacebogroups(Table7).Therewere nosignificantbetween-groupdifferencesinmaternalsideeffects ofnausea,vomiting,pruritus,orsedation(Table8).Nodatawere

reported in any study regarding headache, hypotension, and shiveringwhichwereincludedintheinitialstudydesign.Neonatal outcomesweresimilarinthetwogroupswithrespectto1and 5min APGARs, birth weight, NICU admission, umbilical cord arterialpH,breast feedingdifficultiesinfirst24h,and needfor positivepressureventilation(Table9).

Discussion

Inhealthytermpregnantwomenundergoingcesareandelivery under spinal anesthesia with intrathecal opioids, as well as standing NSAIDs and acetaminophen postpartum, 600mg oral gabapentin1hbeforesurgerywasassociatedwithlowerVASpain scoresinthepostoperativeperiodcomparedtothosewhoreceived placebo, reaching statistical significance at 24hon movement. ThereisalsoasignificantdifferenceinVASpainscoresatisfaction at12and24halthoughthis_findingislimitedbyhavingdatafrom only2and3studiesrespectively.

Thereisconflictingdataregardingtheefficacyofperioperative gabapentin to decrease postoperative pain. In non-pregnant adults,somemeta-analyseshavefailedtoshowanimprovement inpostoperativepainscoresoropioidusefollowingperioperative gabapentin administration while others have demonstrated statistically significant improvements in these outcomes when perioperative gabapentin is used [17–21]. A meta-analysis by Alayedetalshoweddecreasingpostoperativeopioidrequirements, aswellaslowerVASpainscoresinthefirst24hfollowingtotal abdominal hysterectomywhen 400to1800mggabapentinwas dosedperioperatively [7]. Importantly,this studyincludedonly patientsundergoinggeneralanesthesiaincontrasttoourstudy, which includesonlypatientsreceivingregionalanesthesiawith intrathecal opioids where a postoperative analgesic effect is expected for upto 36h [5]. A meta-analysisby Doleman et al includedstudieswherepatientsreceivedbothgeneralandspinal anesthesia and showed a lesser effect on pain medication requirements for patients that underwent spinal anesthesia compared to general anesthesia when receiving preoperative gabapentin [20]. In other words, preoperative gabapentin was moreeffectiveindecreasingpostoperativeopioidconsumptionin patients receiving general anesthesia compared to regional anesthesia. Table9 NeonatalOutcomes. 1minAPGAR (median) 5minAPGAR (median) Umbilicalcord arterialpH

NICUadmission Breastfeeding difficultiesfirst24hrs Birthweight (grams) PositivePressure Ventilation Mooreetal.[11] 9vs 9 9vs9 7.30.1vs 7.30.1 1/21(4.76%) vs0/23 (0) 3/20 (15%)vs 5/21 (23.81%) 3520407vs 3341431 Notreported Shortetal.[12] 9vs9 9vs9 7.30.1 vs7.3 0.1 1/42(2.4%) 2/42(4.8%) 1/42(2.4%)vs 7/42(16.7%) 3505431vs 3382 553 2/42 (4.76%)vs 2/42 (4.76%) Nofaletal.[13] 9vs9 9vs9 7.30.1vs 7.30.1 2/42vs 1/44 2/42vs 1/44 3,251522vs 3,347581 Notreported Memarietal.[14] Notreported Notreported Notreported Notreported Notreported Notreported Notreported Monksetal.[15] 9vs9 9vs9 7.30.1 vs7.30.1 1/100(1) vs4/97(4.1) 16/91(17.6)vs17/88 (19.3) 3408449.7vs 3,3475.5.5 3/100(3) vs3/97(3.1) Hafezetal.[16] 9vs9 10vs10 Notreported Notreported Notreported Notreported Notreported Total – – 7.3vs7.3 5/205(2.4%)vs7/ 206(3.4%) 22/195(11.3%)vs30/ 195(15/4%) 3407.2vs3353.5 5/142(3.5%)vs 5/139(3.6%) I2 _{– –} 22% 0% 20% 0% 0% RRorMD(95%CI) – – 0.01( 0.02to 0.00) 0.77(0.23to2.62) 0.74(0.37to1.48) 65.46( 30.73to 161.66) 0.98(0.29to3.31)

Dataarereportedasnumberintheinterventionvsnumberinthecontrolgroup. MD,meandifference;CI,confidenceinterval;RR,relativerisk.

Strengthsofthismeta-analysisincludethemethodologyusedto conducttheanalysiswhichfollowedcriteriaoutlinedintheCochrane HandbookforSystematicReviewsofInterventions.Allauthorswere contacteddirectlyandadditionalunpublisheddatawascompiledfor analysis.Additionally,theincludedstudieshadlowriskofbias.

This study is limited by characteristics inherent to the individualstudiesincludedforanalysis.Onlyfouroftheincluded sixstudiesassessedtheprimaryoutcomeofVASpainscores.Two offourstudiesshowedadecreaseinpostoperativeVASpainscores andnarcotic usagefollowing pre-cesareangabapentin adminis-trationcomparedtoplacebo,whiletwoshowednodifference.Of thesefourstudies,threewereadequatelypoweredtodetectsucha differenceandonethatfailedtoshowaninterventioneffectwas not. Neuraxial medications and intraoperative/ postoperative analgesicmedicationsaswellasdoseandroutevariedbetween studies but were overall similar. Women in all studies where neuraxial medications are reported (5/6) received intrathecal bupivacaineandfentanylandin3/5studieswomenalsoreceived intrathecalmorphine.Thisstudyisalsolimitedingeneralizability asallpatientswerehealthywithASAclassIorIIandpatientswith substanceusedisordersorchronicpainsyndromeswereexcluded. Data regarding intraoperative surgical technique were not reported and different techniques have been associated with varyingpainfulness[2,22–24].Originaltrialsdidnotreportedon durationofsurgery,oronabdominaladhesions.

An optimal approach to post partum pain control remains incompletely characterized.Differentapproacheshavebeenstudied, includingantenatalandpostnatalinterventions[25–29].Limiting opioidusageinthisperiodisapriorityasdatahaveshownthat1in 300opioidnaïveindividualswillcontinuetouseopioidsafterthe postoperativeperiod[6].Currentanesthesiaguidelinesrecommend amulti-modal,step-wiseapproachthatisindividualizedtoeach patient. Women receiving regional anesthesia with intrathecal morphineshould receiveNSAIDsandacetaminophenwith addi-tionalopioidsreserved for“severebreakthroughpain”[5].While this meta-analysisis limited bya small numberof patients, it does supportaroleforprophylacticgabapentininreducingpostoperative VASpainscoresaftercesareandelivery.Importantly,nodifference wasseeninmaternalsideeffectsorneonataloutcomesforpatients receivingpreoperativegabapentinhighlightingitssafetyforusein thispatientpopulation.

Conclusions

Insummary,prophylacticuseof600mgoralgabapentinprior to cesarean delivery improves postoperative pain control in healthypatients(ASAIorII)undergoingspinalanesthesia with intrathecal opioids, as well as receiving standing NSAIDs and acetaminophenpostpartum.

Disclosure

Theauthorsreportnoconflictofinterest. Financialsupport

Nofinancialsupportwasreceivedforthisstudy. References

[1]MartinJA,HamiltonBE,OstermanMJ,CurtinSC,MatthewsTJ.Births:finaldata for2013.NatlVitalStatRep2015;64:1–68.

[2]EisenachJC,PanPH,SmileyR,Lavand’hommeP,LandauR,HouleTT.Severityof acutepainafterchildbirth,butnottypeofdelivery,predictspersistentpain andpostpartumdepression.Pain2008;140(November(1)):87–94. [3]ApfelbaumJL,HawkinsJL,AgarkarM,BucklinBA,ConnisRT,GamblingDR,

etal.Practiceguidelinesforobstetricanesthesia:anupdatedreportbythe

Americansocietyofanesthesiologiststaskforceonobstetricanesthesiaand thesocietyforobstetricanesthesiaandperinatology.Anesthesiology2016;124 (February(2)):270–300.

[4]ToledoP,MillerES,WisnerKL.Lookingbeyondthepain:caneffectivelabor analgesiapreventthedevelopmentofpostpartumdepression?2018. [5]SuttonCD,CarvalhoB.Optimalpainmanagementaftercesareandelivery.

Anesthesiologyclinics2017;35(March(1)):107–24.

[6]BatemanBT,FranklinJM,BykovK,AvornJ,ShrankWH,BrennanTA,LandonJE, RathmellJP,HuybrechtsKF,FischerMA,ChoudhryNK.Persistentopioiduse following cesareandelivery:patterns andpredictorsamong opioid-naive women.AmJObstetrGynecol2016;215(September(3))353-e1.

[7]AlayedN,AlghanaimN,TanX,TulandiT.Preemptiveuseofgabapentinin abdominalhysterectomy:a systematicreviewandmeta-analysis. Obstetr Gynecol2014;123(Junuary(6)):1221–9.

[8]Carrasco M, Rao SC, Bearer CF, Sundararajan S. Neonatal gabapentin withdrawalsyndrome.PediatrNeurol2015;53(November(5)):445–7. [9]GreenS,HigginsJ.Cochranehandbookforsystematicreviewsofinterventions. [10]MoherD, LiberatiA,TetzlaffJ, AltmanDG.Preferredreportingitemsfor systematicreviewsand meta-analyses:thePRISMAstatement.IntJSurg 2010;8(January(5)):336–41.

[11]MooreA,CostelloJ,WieczorekP,ShahV,TaddioA,CarvalhoJC.Gabapentin improvespostcesareandeliverypainmanagement:arandomized, placebo-controlledtrial.AnesthAnalg2011;112(January(1)):167–73.

[12]ShortJ,DowneyK,BernsteinP,ShahV,CarvalhoJC.Asinglepreoperativedose ofgabapentindoesnotimprovepostcesareandeliverypainmanagement:a randomized, double-blind, placebo-controlled dose-finding trial. Anesth Analg2012;115(December(6)):1336–42.

[13]NofalWH,MahmoudMS,AlAlimAA.Doespreoperativegabapentinaffectsthe characteristicsofpost-duralpunctureheadacheinparturientsundergoing cesareansectionwithspinalanesthesia?SaudiJAnaesth2014;8(July(3)):359– 63.

[14]MemariF,JadidiR,NorooziA,MohammadbeigiA,FalahatiJ.Protectingeffect ofgabapentinfornauseaandvomitinginthesurgeryofcesareanafterspinal anesthesia.AnesthEssaysRes2015;9(September(3)):401–4.

[15]Monks DT, Hoppe DW, Downey K, Shah V, BernsteinP, Carvalho JC. A perioperativecourseofgabapentindoesnotproduceaclinicallymeaningful improvementinanalgesiaaftercesareandeliveryarandomizedcontrolled trial.JAmSocAnesthesiol2015;123(August(2)):320–6.

[16]Hafez MH, Abdelhamid MH, Youssef MM, Abdelrahim IK. Randomized controlledtrialoftwooralregimensofgabapentinversusplaceboinpatients forcesareansectionunderspinalanesthesiaregardingpostoperativepain, sedation,nauseaandvomiting.EgyptJAnaesth2017;33(January(1)):59–65. [17]FabritiusML,GeislerA,PetersenPL,NikolajsenL,HansenMS,KontinenV,etal. Gabapentinforpost-operativepainmanagement–asystematicreviewwith meta-analysesandtrialsequentialanalyses.ActaAnaesthesiolScand2016;60 (October(9)):1188–208.

[18]FabritiusML,WetterslevJ,MathiesenO,DahlJB.Dose-relatedbeneficialand harmfuleffectsofgabapentininpostoperativepainmanagement–posthoc analysesfromasystematicreviewwithmeta-analysesandtrialsequential analyses.JPainRes2017;10:2547.

[19]HoKY,GanTJ,HabibAS.Gabapentinandpostoperativepain–asystematic reviewofrandomizedcontrolledtrials.Pain2006;126(December(1-3)):91– 101.

[20]DolemanB,HeininkTP,ReadDJ,FaleiroRJ,LundJN,WilliamsJP.Asystematic reviewandmeta-regressionanalysisofprophylacticgabapentinfor postop-erativepain.Anaesthesia2015;70(October(10)):1186–204.

[21]TiippanaEM,HamunenK,KontinenVK,KalsoE.Dosurgicalpatientsbenefit fromperioperativegabapentin/pregabalin?Asystematicreviewofefficacy andsafety.AnesthAnalg2007;104(Junuary(6)):1545–56.

[22]MackeenAD,KhalifehA,FleisherJ,HanC,LeibyB,BerghellaV.Painassociated withcesareandeliveryskinclosure:arandomizedcontrolledtrial.Obstetr Gynecol2015;126(October(4)):702–7.

[23]SacconeG,CaissuttiC,CiardulliA,BerghellaV.Uterinemassageforpreventing postpartumhemorrhageatcesareandelivery:whichevidence?EurJObstetr GynecolReprodBiol2018;23(February).

[24]XodoS,SacconeG,CromiA,OzcanP,SpagnoloE,BerghellaV.Cephalad– caudadversus transverseblunt expansion ofthe lowtransverseuterine incisionduringcesareandelivery.EurJObstetrGynecolReprodBiol2016;1 (July(202)):75–80.

[25]HansSL,EdwardsRC,ZhangY.Randomizedcontrolledtrialof doula-home-visitingservices:impactonmaternalandinfanthealth.MaternChildHealthJ 2018;31(May):1–9.

[26]Ozturk Inal Z,GorkemU, AliInal H. Effects ofpreoperative anxiety on postcesareandeliverypainandanalgesicconsumption:generalversusspinal anesthesia.JMaternal-FetalNeonatalMed2018;9(June)1–41just-accepted. [27]MazyA,GadM,BedairyM.Preperitonealpostcesareansectionbupivacaine

analgesia:comparisonbetweendexamethasoneanddexmedetomidineas adjuvants.SaudiJAnaesth2018;12(April(2)):183–9.

[28]Towers CV,Shelton S, van Nes J, Gregory E, Liske E, Smalley A, et al. Preoperativecesarean deliveryintravenous acetaminophen treatment for postoperativepaincontrol:arandomizeddouble-blindedplacebo control trial.AmJObstetGynecol2018;218(March(3))353-e1.

[29]LalmandM,WilwerthM,FilsJF,VanderLindenP.Continuousropivacaine subfascialwoundinfusioncomparedwithintrathecalmorphinefor post-cesareananalgesia:aprospective,randomizedcontrolled,double-blindstudy. AnesthAnalg2017;125(September(3)):907–12.