BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 ContentslistsavailableatScienceDirect Best Practice & Research Clinical Anaesthesiology journal homepage: www.elsevier.com/locate/bean 1 Practical pain management in the neonate Jonathan De Lima, MBBS, PhD, FANZCA, Paediatric Anaesthetista,*, Kathryn Browning Carmo, BMED, FRACP, Neonatologistb aDepartmentofAnaesthesia,Children’sHospitalatWestmead,CnrHawkesburyRoadandHainesworthAve,Westmead2145,Sydney, NewSouthWales,Australia bGraceCentreforNewbornCareChildren’sHospitalatWestmead,Sydney,Australia Neonatalcareisadvancingtolevelswheremoreneonatesarenow Keywords: pain offeredmoreinvasiveinterventions,exposingthemtomorepro- neonate longedhospitalcare.Consequently,theprovisionofeffectiveand analgesia consistent management of pain in these neonates has become sucrose a pressing challenge. Advances in neonatal care have not only paracetamol increased the number of neonates, who are exposed to noxious morphine stimuli, but, over recent decades, also altered the patterns of localanaesthesia exposure.Bothproceduralandpostoperativepainremaindistinct innature,prevalenceandmanagement,andneedtobeaddressed separately.Recentadvancesinthemanagementofneonatalpain havebeenfacilitatedbyimprovedmethodsofpainassessmentand an increased understanding of the developmental aspects of nociception.Overthepastdecade,therehavebeensomeadvances intheavailablepharmacologicalarmamentarium,modestclarifi- cationoftherisksofbothuntreatedpainandaggressiveanalgesic practice and a greater recognition of non-pharmacological anal- gesictechniques. However,evenadvancedhealthsystemsfailtoconsistentlyartic- ulatepainmanagementpolicyforneonates,instituteregularpain assessments and bridge the gaps between research and clinical practice. (cid:1)2010ElsevierLtd.Allrightsreserved. Theuniquephysiologyofthehumanneonatecontinuestoprovidechallengestoboththepractical andacademicaspectsofthehospitalcareofasicknewborn.Effectiveandconsistentmanagementof neonatalpain anddistressremainsoneof themostobviousandpersistentchallenges.Thenoxious stimulithatanacutelysickneonateisexposedtoduringclinicalcarearevastlydifferentinrangeand * Correspondingauthor.Tel.:þ61298450000;Fax:þ61298453959. E-mailaddress:[email protected](J.DeLima). 1521-6896/$–seefrontmatter(cid:1)2010ElsevierLtd.Allrightsreserved. doi:10.1016/j.bpa.2010.04.001 292 J.DeLima,K.B.Carmo/BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 durationtothenormal(limited)painexperiencesofanotherwisehealthyneonate.Thiscontrastswith the experiences of hospitalised infants and children whose development has brought a natural interactionwiththeenvironmentatlargeand,withthis,thetypicalpainexperiencesofchildhood.1 Inthisolderpaediatricpopulation,thepainfulexperiencesofhospitalisationoccurwithinacontext of prior normal pain experiences and associated analgesic and comfort measures. The acutely sick neonate,ontheotherhand,isexposedtoclinicalcareandstimulithatbearnorelationtoanythingthat couldconceivablybetermeda‘normalpainexperience’. The pain and distress imposed by procedures and surgical interventions in neonates were first brought to scientific and public attention bya small but seminal study conducted in the 1980s on neonatesundergoingthoracotomy.2Sincethen,theepidemiologyofneonatalpainwithintheacute careenvironmenthasbeenstudiedandpatternsoverthepastdecadehavebeenwelldocumented.3–5 Thesepatternsconsistentlyrevealthatneonatesinacutecaresettingsareregularlysubjecttorepeated painfulorstressfulstimuliand,often,littleornoanalgesicisadministered.Procedureratesof10–16 perdayhavebeenrepeatedlydocumentedinwell-conductedepidemiologicalstudieswithinmodern neonatalintensivecaresettings.6Itshouldbeemphasisedthatthefocusofthesestudieshasprinci- pally been on brief procedures during intensive care rather than on surgical interventions, post- operativepainorongoingnoxiousstimuli(chronicpain).Themostcommonproceduresincludenasal andtrachealaspiration,heellanceandadhesiveremoval.Althoughslightlylessfrequentlyobserved, gastrictubeinsertion,venepuncture,arterialpunctureandvenouscannulationalsofigurehighlyin thesestudies.Therearefarfewerstudiesaddressingpost-surgicalpainandongoingorchronicpainin neonates.Approachesandmanagementstrategiesforthesetypesofpainareoftensimplemodifica- tionsofthoseusedininfantsandolderchildren. Importantly,advancesinneonatalcarehavenotonlyincreasedthecaseloadbutmayalsoalterthe epidemiology of neonatal pain. The widespread use of antenatal steroids, postnatal surfactant and nasal continuous positive airway pressure has resulted in a marked reduction in the need for and duration of ventilation of premature babies and, consequently, the need for related invasive procedures.7 Despitethis,proceduralpainclearlyremainsoneofthemorepressingissuesfacingneonatologists. Anaesthetists,ontheotherhand,aremoreusuallyconcernedwithpostoperativepain.Thesetwopain types – procedural and postoperative – are distinct in nature, prevalence and management. The distinctionmaybefurtheraccentuatedbydifferencesincaresettings(i.e.,operatingroomandsurgical intensive care unit (ICU) as opposed to neonatal nursery) as these areas are subject to different caseloadsandclinicalinput.Thisreviewwillcoveraspectsofbothtypesofpain. Proceduralpainisanunavoidableaspectofneonatalintensivecareandischaracterisedbyrepeated episodes of short duration and minimal tissue damage. It must be assessed and managed quite differentlytoinflammatory(post-surgical)pain.Proceduralpainactivatespainpathwaysinawaythat canbeseenasa‘warningsystem’.Whenmature,thesepathwaysprovideforresponsesthatarerapid, welllocalisedandcloselytiedtowithdrawalreflexesandattention.Thesepathwaysarenowknownto befunctionalby24weeks’gestationbutinearlydevelopment,tendtofacilitategeneralisedand/or exaggeratedresponsestoacutenoxiousstimuli. Post-surgicalpain,ontheotherhand,ischaracterisedbyongoingactivityandmodulationofthese samenociceptorsandpathwaysastheyinnervatedamagedtissues.Thisactivityanditsmodulationare linkedtoinflammatoryandhealingprocesseswithinthesetissues.Thepainexperiencesassociated withsurgeryandtissuedamagemaybetermeda‘painstate’,implyinganalteredneurophysiology. Thisformofpain,expressedashyperalgesia,spontaneouspainandallodynia,maybebetterseenas a‘protectiveremindersystem’–causingthesubjecttoshield,protectandimmobilisetheaffectedbody part.Theneurobiologicalmechanismsunderpinningthisstatehavebeenwellstudiedandincludeboth central and peripheral sensitisation of primaryafferent nociceptive fibres. In studying the develop- mentalaspectsofthesemechanisms,importantquestionsaboutthedurationandseverityofhyper- algesiafollowingsurgeryinneonatesarebeinganswered.8 In both neonates and adults, the distinction between procedural pain and post-surgical pain providesthebasisforunderstandingwhytypicalanalgesictechniquesanddrugs(e.g.,paracetamoland morphine)areoftenineffectiveinrelievingproceduralpain9–11andwhyproceduralpainmustoftenbe managedusinganaestheticandsedativeagentsinstead.12,13 J.DeLima,K.B.Carmo/BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 293 Recentadvancesinneonatalpainmanagementhavebeenfacilitatedbyimprovedmethodsofpain assessment, an increased understanding of the developmental aspects of nociception and some evidenceofmeasurableeffectsofearlynoxiousstimulionlaterneurologicaldevelopment.14 Painassessment Withouttheadvantageofself-report,clinicalassessmentsofpainintheneonatearedependenton indirect measures. These measures include physiological, behavioural and contextual parameters. Contextualfactorsincludetheneonates’behaviouralstate(asleep/awake),gestationalage,severityof illnessandthecarer’sperception.Eachofthesemayinfluencetheassessmentorresponseofaneonate toanoxiousstimulus. Taken independently, behavioural, physiological and contextual measures lack specificity and sensitivityandmostwillchangewithclinicalcontext.Correlationbetweenphysiologicalandbehav- iouralmeasuresisoftenpoor.15Novelapproachessuchasheartratevariability,cryanalysis16,17and skin conductancemeasures haveall proved to be no morespecific or sensitive thansimplerobser- vationalscales.18Newertechnologiessuchasnear-infraredspectroscopy(NIRS),electroencephalogram (EEG)andfunctionalneuroimaging,whichattempttoassessdegreesofcorticalactivationinresponse tonoxiousstimuli,19,20haveasyetundefinedutility,feasibilityandclinicalsignificance.21Thesemore technicallydemandingmeasuresremainresearchtoolsthatrequirefurthervalidation.Theymayyield fascinatinginsightsintotherelationshipsbetweenbehaviouralresponses,physiologicalchangesand cortical activation in response to noxious stimuli. Despite the documented dissociation between physiologicalandbehaviouralindicatorsofpain,compositescoringtoolsthatincludebothhavebeen reasonablyvalidatedandarewidelyaccepted.22 Manypainanddistressscoringtoolshavebeendescribedinthepastthreedecades,followingthe acceptance that pain in neonates cannot be ignored. The study and validation of these tools have tended tofocus onprocedural pain and less work has been donewith these tools in the settingof postoperative or chronic pain. Although no single tool can reasonablyexpect to cover the range of clinicalsettingsinwhichneonatesfindcare,therestillremainsaneedforgreaterconsistencyintheuse andchoiceofthescoringtoolacrossneonatalcarecentres. Inrecentyears,sevenpainscoringtoolshavefoundwidespreaduse:theNeonatalInfantPainScale (NIPS),theChildren’sRevisedImpactofEventScale(CRIES),thePrematureInfantPainProfile(PIPP), theLiverpoolInfantDistressScale(LIDS),theDistressScaleforVentilatedNewbornInfants(DVSNI), thePainAssessmentTool(PAT)andtheCOMFORTscale.Eachofthesehasbeenvalidatedandshow reasonableinternalconsistencyandreliabilities(seeTable1).Twoscales(NIPSandLIDS)arebehav- iouralscales,whiletheothersarecompositetools.Morerecently,thebehaviouralIndicatorsofInfant Pain(BIIP) scalehasbeenshowntobevalidandreliablein measuringproceduralpain inpre-term neonates.23 Rather uniquely, this tool includes two specific hand actions (finger splay and fist clenching). Twoparticularareasofweaknessaresharedbyallneonatalpainassessmenttools.First,upto20%of extremelyprematureinfantsmayshownomeasurableresponsetonoxiousstimuli.24Whetherthis representsatruelackofnociception,adistinct(energysaving)painstate,afailureofpainmeasuring toolsorexhaustionisunknown.Second,thereareveryfewtoolsformeasuringresponsestocontin- uousnoxiousstimulation(chronicpain)inneonatesandthesearenotinwidespreadclinicaluse.One suchscale–theEDIN(EchelleDouleurInconfortNouveau-Ne)scaleusesfivebehaviouralindicators includingthequalityofsleep,qualityofcontactwithnursesandconsolability.Importantly,theseare scored after several hours of observation and clinical interaction.25 Chronic or ongoing pain can reasonably be assumed to arise from inflammatory conditions, for example, inflammatory and desquamating skin conditions, necrotising enterocolitis, indwelling/ percutaneous catheters and intercostal drainsand surgical wounds.26 Prolongedintubation and nasal mask application are also sourcesofcontinuousnoxiousstimulation.25 It would appear self-evident that regular documented objective pain scoring of all neonates admitted for acute care is essential for optimising the provision of analgesia. Unfortunately, even advancedhealthsystemscanfailtoinstituteregularpainassessmentsanddevelopmanagementpolicy basedonobjectivescoresasshownbyarecentAustraliansurvey.27 294 J.DeLima,K.B.Carmo/BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 Table1 Painscoringtoolsforneonates. Description Validity/internal Reliability Suggested Reference consistency (Cronbach’s clinicalutility alpha) NIPS Variablesinclude:facial Face,constructand Interrater>0.92 Proceduralpain LawrenceNeonatal expression,crying, concurrent intermand Network1993;12 breathingpattern,posture (r¼0.53–0.84) pre-term (6):59–66 andarousal. neonates28–38 weeks CRIES Variablesinclude:crying, Face,content, Interrater>0.72 Postoperative KrechelPaed changesinvitalsigns, discriminationand inneonates32– Anaesth1995; oxygensaturation,facial concurrent 60weeks 5:53–61 actionandwakefulness. (r¼0.53–0.84) PIPP Variablesinclude:changes Face,content, 0.89orgreater Proceduralpain Ballantyne.23 inheartrate,oxygen construct–inpre- inpremature Stevensetal.Clin saturation,facialactionand term andterm JPain1996; neurobehaviouralstate. andtermneonates. neonates 12:13–22 Weightedforgestational Internal age consistency:0.59– 0.76 LIDS Variablesinclude:facial Establishedcontent 0.84orgreater Postoperative Horgan1996 expression,sleeppattern, andconcurrent interm PaediatricNursing cry,movement,postureand validity neonates 8(10):24–7 tone. DSVNI Variablesinclude:Facial Ventilated, Sparshott1996 expression,movementand neonates JournNeonatal colour,vitalsigns, Nursing;2:5–11 oxygenationand temperature PAT Variablesinclude:infant Requiresregular Interater Postoperative Spence2005 behaviour,facial calibration.Good reliability0.85 painin ObstetGynecol expression,vitalsignsand correlationwith neonates NeonatalNurs nursingimpression CRIESscore(0.76). COMFORT Variablesinclude: Established Goodinter-rater Ventilated DijkvanMetal movement,behavioural congruentvalidity reliability neonates Pain2000; state,facialtension,vital (Kappa:0.63– Postoperative 84:367–77 signsandmuscletone 0.93) painandPICU distress. Although clear evidence is still awaited, linking objective pain scores to care interventions and analgesiashouldfosterbetteroutcomesandprovidethenecessaryincentivetomaintaintheeffortof repeatedlymeasuringanddocumentingpainscores. Minorsurgery Ingeneral,theadvancesinanalgesicpharmacologythatcharacteriseadultacutepostoperativepain managementhavegraduallybeentranslatedintoclinicalpracticechangesinmanypaediatriccentres. Theethicaldriveforthishasbeenhelpedbyeconomicpressuresonacutecaresettings.Thereisstrong incentivetoreducehospitallengthofstayandgoodperioperativeanalgesiafacilitatesearlydischarge. With some natural delay, advances inpaediatric pain management havedriven change in neonatal perioperative care. Neonates undergoing surgical interventions form a distinct subgroup within neonatal nurseries and modern care of these infants regularly includes general anaesthesia, local anaesthesia,paracetamoland,formoremajorsurgery,systemicopioids. Manylessersurgicalprocedures(e.g.,circumcision,pyloromyotomyandinguinalherniarepair)in neonatescannowbemanagedona‘short-stayadmission’basis.Thismodelofcareisnowlimitedonly by concerns about postoperative apnoea, hydration and feeding, and parental anxiety rather than postoperative pain. Successful perioperative care and postoperative analgesia is based on local anaesthesiaandregularparacetamol.Rescueanalgesiaisrarelyrequiredandanearlyreturntofeeding J.DeLima,K.B.Carmo/BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 295 andnormalsleeppatternscanbeexpected.Therelativesuccessoftheseregimensreducestheneedto expose neonates to the risk of opioid-induced respiratory depression and more invasive analgesic techniquesinthiscaresetting. Localanaestheticagentsincludethe amideslignocaine,bupivacaine, ropivacaineandlevobupiva- caine. All are absorbed rapidly and are hepatically metabolised by cytochrome P450 enzymes. The metabolismofthelatterthreelonger-actingagentsissubjecttoalowerhepaticextractionratio(0.4) thanthatforlignocaine(0.7).RopivacaineismainlymetabolisedbytheCYP1A2isoform,whichmay mature later than the CYP3A4 form that is responsible for the metabolism of bupivacaine.28 Phar- macokineticparametersforthesedrugshavebeenestablishedininfantsandshowthatyoungerage predicts a larger volume of distribution, lower clearance and longer half-life.29 Being alkaline, the amide anaesthetic agents are bound in plasma by alpha-1-acid glycoprotein (a low capacity, high affinitysite).Thisphasereactantproteinispresentinonlylowconcentrationsinneonates,risingto adultvaluesbetween6and12monthsofage.Althoughthisisanimportantconsideration,theriskof toxicitywithlargedosesorprolongedcontinuousinfusionisrelatedprincipallytothereducedintrinsic clearanceofthedrugs.29 Pharmacodynamicdifferencesbetweenagentshavebeendebatedandasyethavenotbeenclearly demonstratedinapaediatricpopulation.Bothropivacaineandlevobupivacainehavebeenadvertised assaferagents,basedonasmallnumberofcasereportsandlaboratorystudies.Apartfromdifferences in potency, there appear to be no fundamental differences between the effects of these drugs on sodiumchannelelectrophysiology.30Thepharmacologyofthesedrugsisdescribedingreaterdetailin asubsequentsection.Recommendedmaximumdosesforsingle-doseinfiltrationsaregiveninTable2. Paracetamol.Perioperativeanalgesiausuallyincludestheuseoforalorrectalparacetamolandis rarely requiredfor more than48hrs following minorsurgery. The effectivenessof paracetamol has beenwelldocumentedininfantsandchildrenundergoingavarietyofsurgicalprocedures.31Although similarevidenceisyettobepublishedforneonates,itisreasonabletoextrapolateanalgesicefficacy dataforparacetamoltotheyoungeragegroups.32Inneonates,itcanbedeliveredorallyorrectallyfor mostminorsurgicalprocedures.Dosingschedulesinwideclinicalusageandbasedontheavailable pharmacokineticdatatendtobemoreliberalthanlicensedrecommendations(seeTable3).33Para- cetamolisconjugatedintheliverwithglucuronideorsulphateandthenrenallyexcreted.Thesephase IIreactionscontinuetomatureupto3monthsofage.Neonatalimmaturityinthesesystemsresultsin alowerclearanceratecomparedtoinfantsandisthebasisofthereduceddose(andincreaseddosing interval)recommendationsgivenbelow.Theontogenyoftheenzymesysteminvolved(glucuronosyl transferase isoenzymes) has been recently studied. These studies confirm major variability in the metabolic capacity of neonates and that this variability is only partlyexplained by gestational and postnatalage.Asignificanteffectofrepeateddosingwasdetected–implyingthattheenzymesystem undergoes a degree of ‘induction’.34 Hepatotoxicity is related to metabolic products of oxidative reactions.ThesephaseIpathwaysarelesswelldevelopedinneonatesprovidingsomerationaleforthe relativesafetyofthedruginneonatalpractice.35,36 Circumcision Circumcisionintheneonateforsocialreasonsisnolongersupportedinmostpubliclyfundedhealth jurisdictions.Itremainsacommonpracticeoutsideofthismilieuandraises,amongotherissues,pain Table2 Localanaesthesticsforsingle-doseinfiltrations. Typicalconcentration Maxdose Maxdose Withadrenaline Withoutadrenaline (mg/kg) (mg/kg) Lignocaine 0.5% 5 4 Bupivacaine 0.25% 2 2 Levobupivacaine 0.25% 2 2 Ropivacaine 0.2% 2 2 296 J.DeLima,K.B.Carmo/BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 Table3 Paracetamoldosing Agegroup Oral Rectal Loadingdose Maintenance Dosing Loadingdose Maintenance Dosing Maximum Durationat (mg/kg) dose(mg/kg) interval(h) (mg/kg) dose(mg/kg) interval dailydose maximum (h) (mg/kg/d) dose(h) Neonate 28–32weeks 20 10–15 8–12 20 15 12 40 48 PCA >32weeks 20 10–15 6–8 30 20 8 60 48 PCA Infant 1–3 20 15 6 30 20 8 60 48 months 3–12 20 15 4–6 40 20 6–8 90(for48h) 72 months then60 AdaptedfromAranaandMortonetal.31(PCA:Post-conceptionalage). managementinnewbornsasprovidedbygeneralpractitionersandobstetricians.Americantraining programmeshaverespondedtotheseissuesbyteachingtopicalorlocalanaesthetictechniquesbut thesemaynotbeuniversallyapplied.37Inmanyjurisdictions,circumcisionsininfantsareperformed under general anaesthesia and circumcision in the awake neonate has become less frequent. Circumcisionwithoutanalgesiaisnolongeracceptableinanymedicalsetting. Recommendedanalgesicregimenstypicallyincludeoralorrectalparacetamolandsomeformof localanalgesia(topicalanaesthesia,dorsalpenilenerveblock(DPNB),ringblockorcaudalinjection). Ring blocks and DPNBs should not be performed with adrenaline-containing solutions. In awake neonates, topical anaesthesia with lignocaine–prilocaine is more effective than placebo but is less effective than DPNB.38 Liposomal lignocaine may provide some advantages over lignociane– prilocaine.39 Whencombinedwithgeneralanaesthesia,thereisnodatatosuggestthatoneorotheroftheselocal techniquesissuperior.Bupivacaineconfersbetteranalgesiathanlignocaineasjudgedbytheneedfor rescue analgesia.40 Penile nerve block and caudal injection studied in young children are equally effectiveandsafewhenperformedbyexperiencedpractitioners.41 Herniotomy/pyloromyotomy Painmanagementfortheseneonatalsurgeriesisnowwellestablishedandisbasedontheuseof local analgesic drugs and simple analgesics. Local anaesthetic agents are most commonly injected subcutaneouslyjustpriortoskinincisionafterinductionofanaesthesia. Forherniotomysurgery,deeperinjectionintotheplanebetweentransversusabdominisandthe internalobliquemuscleblockstheilio-hypogastricandilio-inguinalnerves.Furtherinjectionbythe surgeon around the genitalbranch of the genito-femoral nerve allows complete regional analgesia. Transversus abdominis plane (TAP) blocks, while technically feasible in neonates,42 have not been shown to be any more efficacious than simple wound infiltration. Typically, bupivacaine, levobupi- vacaineorropivacaineareusedinvolumesof1mlkg(cid:2)1toamaximumdoseof2mgkg(cid:2)1.Thereareno datacomparingthesafetyorefficacyoftheseagentsinneonatalsubjects.Clinicalchoicesarebasedon thelimiteddatafrominfants,olderchildrenandadults.Pre-clinical(animaloradultvolunteer)studies suggest that levobupivacaine may have less adverse effects on cardiovascular and central nervous systemphysiology43but,likebupivacaineandropivacaine,italsohassignificantlyreducedclearancein neonates.Thepharmacologyoftheselocalanaestheticdrugsinneonateswillbecoveredinsubsequent articles. If herniotomy is performed under spinal or caudal epidural anaesthesia, postoperative pain is usuallymanagedwithsimpleoralorrectalparacetamolandpostoperativedistressusuallyrespondsto immediatefeeding. J.DeLima,K.B.Carmo/BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 297 Majorneonatalsurgery Opioidsremainthemainstayofpainmanagementfollowingmajorneonatalsurgery.Morphine,in particular, has proved effective and has widespread use despite its well-recognised limitation of aprolongeddurationofactioninneonates.Fentanylisfindingincreasinguseduringneonatalsurgery andpost-operative care.Greaterhaemodynamicstabilityassociatedwith thesyntheticopioidmust be balanced againstrecognised disadvantages suchas theearlyonsetof tolerance44 and the devel- opmentof thoracicmusclerigidity(‘frozenchest’).45Bothopioidsdisplaynarrowmarginsbetween analgesic doses and those that cause respiratory depression and apnoea. This generallymeans that neonates,whoundergomajorsurgery,willrequirepostoperativeventilation.Severalsmallstudiesof morphineformthebasisofourunderstandingofopioidpharmacologyinneonates.46Thesestudies demonstratedistinctdifferences inmetabolismof thedrugin neonates andinfants comparedwith olderchildrenandadults.Theabilitytoconjugatemorphineintoeithertheglucuronideorsulphate derivativeispresentinthenewborn(includingpre-termneonates)butthecapacitymaybelimited byimmaturityoftheenzymesystems.Whilesulphationhasbeendocumentedinpre-termandterm neonates, it remains a minor metabolic pathway and is soon surpassed by glucuronidation during neonatal life.47 Possibly due to the reduced metabolic capacity, a larger proportion (16–80%) of administeredmorphineisexcretedunchangedintheurineintermneonatesthaninolderchildren and adults (3–15%).48,49 Debate continues regarding the relative ratios of morphine-6-glucuronide andmorphine-3-glucuronidetomorphine,butpotentialdifferencesinthelevelsofthesemetabolites areunlikelytocompletelyexplainthewidelyappreciatedsensitivityofneonatestoopioids. While volumes of distribution (mean value 2.8(cid:3)2.6lkg(cid:2)1) appear to be stable across early developmentandconsistentwithadultvalues,bothhalf-lifeandclearancechangerapidlyduringthe firstfewweeksoflife.47Thepooledestimateformorphine’shalf-lifeinpre-termneonatesis9.0(cid:3)3.4h andthisdecreasesto6.5(cid:3)2.8hintermneonates.Thisvaluedecreasesfurtherstillto2.0(cid:3)1.8hin childrenandinfantsover11days.47Theeliminationphaseinneonatesisextendedbyalongterminal half-life (24.8(cid:3)4.6h), which may be particularly important when managing the sick ventilated neonate.50 Clearance of morphine also shows strong developmental patterns going from pre-term values of 2.2(cid:3)0.7mlmin(cid:2)1kg(cid:2)1 to 8.1(cid:3)3.2mlmin(cid:2)1kg(cid:2)1 at term. Values close to those found in adults have been measured in neonates older than 11 days and are consistently more than 20mlmin(cid:2)1kg(cid:2)1inbabiesolderthan2months.47Thepreciseeffectofeithergestationalorpostnatal ageonthesedevelopmentalpatternsremainselusive.Keepinginmindalargeinter-andintra-indi- vidual variability, it is reasonable to assume that babies achieve adult-like values for half-life and clearanceby2monthsofage.47 Morphine is typically delivered by continuous infusion or intermittent bolus doses with or without a continuous background infusion (NCA). Differences in analgesia between continuous infusion regimens and intermittentinjections are difficult todemonstrate.51 Infusionratestypically range between 10 and 30mgkg(cid:2)1h(cid:2)1. Slightly lower rates (5–10mgkg(cid:2)1h(cid:2)1) have been recom- mended for pre-term neonates52 and for neonates under 7 days of age.50 Safedeliveryofopioidsispredicatedoncloseobservation.Regularlymeasuredanddocumented painscoresaswellasrespiratoryobservationandcontinuouspulseoximetryaremandatory. Thepharmacodynamiceffectsofopioidsareparticularlydifficulttostudyinneonatesandthefew availablestudiesallowonlylimitedconclusions.Thereisclearlylargeindividualvariationandseem- inglylittlecorrelationbetweenplasmaconcentrationanddrugeffect.Susceptibilitytotherespiratory depressanteffectsofmorphinehasbeenstudiedinventilatedneonates,infantsandchildrenfollowing cardiac surgery. While a plasma concentration threshold for respiratory depression was detected, aclearchangeinsensitivityacrosstheagegroupswasnotdetected.53 Paracetamolisrecommendedasananalgesicadjuncttoopioidsinneonatesfollowingmajorsurgery, althoughevidenceofitsefficacyisextrapolatedfromstudiesinolderinfantsandchildren.32,54Dosage recommendationssummarisedbyAranaandMortonetal.havewidespreadacceptance.31SeeTable3. Intravenousparacetamolhasbeenavailable(atleastasapro-drug)forthepasttwodecadesandhas beenasignificantadvanceintheanalgesicformulary.Itisgenerallyonlyindicatedwhenotherroutes ofadministrationarenotpossible.Itisbecomingbetteracceptedforuseinneonatesaskineticdataand studies of safety have been published.55,56 These studies confirm that both postnatal and post- 298 J.DeLima,K.B.Carmo/BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 conceptionalage(PCA)effectthemetabolicconjugationofparacetamoltoglucuronide.Importantly, repeatedadministrationisalsocorrelatedwithincreasingglucuronidationcapacity.34Despitesignif- icant gaps in our knowledge of paracetamol pharmacology in neonates, the regimen based on that originallyproposedbyAllegaerthasthemostcurrencyandcanberecommendeduntilfurtherdata becomeavailable(seeTable4). Table4 Intravenousparacetamoldosing. Agegroup Maintenancedose(mg/kg) Dosinginterval(h) Maximumdailydose(mg/kg/d) Neonate 28–32weeksPCA 7.5–10 6–8 40 32–36weeksPCA 7.5–10 6 50 >36weeks 15 6 60 Infant 1–3months 15 6 60 3–12months 15 4–6 60 FromAllegaert2004. Non-steroidalanti-inflammatorydrugs(NSAIDs)nowhaveanestablishedplaceintheprovisionof analgesiaininfantsandchildren.Goodevidencefortheirsafetyandefficacyinavarietyofsurgical settings has been published.57–59 Their use in neonates has been relatively limited to the medical closure of persistent ductus arteriosus.60 In this setting, the risk of cerebral, gastrointestinal or reversiblerenalinjuryisbalancedbythesignificantadvantageofdecreasingpulmonaryhaemorrhage and intraventricular haemorrhage.61 Medical closure has the added advantage of avoiding general anaesthesia and thoracotomy.62 A significant negative effect on glomerular filtration rate has been demonstratedinneonatesincludingprematurebabies.63–65 Despite thesewell-documented risks, ketorolachas been administered to small cohorts of well- selectedneonatesforpainrelieffollowingsurgeryanddatashowingefficacyhavebeenpublished.66 Unfortunately,evidenceofsafetyandanadequaterisk:benefitratiofortheuseofNSAIDsforanal- gesia in neonates is not available. Selective NSAIDs (cyclooxygenase (COX)-2 specific drugs) find occasionaluseinchildrendespiteaseriouslackofpharmacokineticandsafetydata.67Theevidencefor anyadvantageovernon-selectivedrugsinchildrenislimitedandtherearenodataontheuseofthese drugsinneonates.68RoutineuseofNSAIDdrugsforanalgesiainneonatescannotberecommended untilfurtherpharmacokineticandsafetydataareavailable.69–71 Neuraxialblockadeandinfusions.Asininfantsandchildren,localanaestheticagentshaveasignificant opioid-sparingeffect.Followingneonatalsurgery,thiscancriticallyinfluencethedecisiontoextubate ababyandtherebyavoidpostoperativeventilation.72Theadvantagesofearlyextubationandareturnto ‘kangaroocare’(i.e.,nursingtheneonateonthechestofaparent)arenotalwaysimmediatelyobviousas careoftheventilatedneonateinmodernnurseriesisbotheffectiveandrelativelysafe.Earlyextubation ofaneonatefollowingmajorsurgerydemandsahighernursingacuityintheimmediatepost-extuba- tion/operative hours. This is later followed by decreased nursing acuity due to earlier parental re- involvementintheneonate’scareandanearlierreturntofeeding. Local infiltration of surgical wounds provides effective immediate perioperative analgesia but currentlyavailableagentscannotprovidereliefbeyondafewhours. Caudalinjectionoflocalanaestheticsprovidesgoodanalgesiaforlowerabdominalwallandperi- nealsurgerythatmaybeexpectedtoextendwellintotheearlyrecoveryphasefollowingsurgery. Newertechniquessuchasthetransverseabdominisplaneblockhaverecentlybeendescribedfor upperandmid-abdominalsurgeryinneonates.42Whiletechnicallyfeasible,advantagesoversimple woundinfiltrationareyettobedemonstrated. Continuousepiduralanalgesiainpaediatricshasnotbeensubjecttothemeta-analysisthathasbeen applied to adult practice.73 Many conclusions from the latter are extrapolated to varying degrees perhaps explaining the variety of approaches taken by different institutions. Earlyexperience from majorpaediatriccentresconfirmedthatcontinuousepiduralanalgesiawasabletofacilitateearlyor immediatepostoperativeextubation.74Infact,continuouspostoperativeepiduralinfusionsarepred- icatedonthedistinctassumptionthattheneonatewillbeextubatedintheimmediatepostoperative J.DeLima,K.B.Carmo/BestPractice&ResearchClinicalAnaesthesiology24(2010)291–307 299 period so that spinal cord function can be assessed. Although bupivacaine infusions have been successfully used in several large neonatal series,74 concern regarding the risk of rising plasma concentrations and reduced clearance suggests that continuous infusion should always be closely monitored.75 Ropivacaine for epidural infusion in neonates has been found to provide acceptable analgesiawithoutevidenceofdrugaccumulationdespiteaclearancecalculatedtobelowerthanthat ininfants>30daysold.76 Epiduralanalgesiacanofferbenefitstowell-selectedneonates,parentsandneonatalnurseries.It requires dedicated follow-up carefor ittowork as neonatalepidural infusions oftenneed regular adjustment. Also, caring for the awake postoperative neonate is quite different from the care requiredfortheventilatedneonate.Whileweawaitastrongevidencebaseforneonatalepidurals, current practice depends on individual enthusiasts and close communication between the anaes- thetists and neonatologists involved in postoperative care. Epidural analgesia in neonates will be enhanced by ultra-sonographic imaging (placement and monitoring) and further advances in pharmacology. Fromanecdotal reports, continuousregionaltechniques suchasrectus sheathcatheterinfusions have proved effective when used in well-selected cases and can avert the need for postoperative ventilation. Intrathecal opioids have occasionally been used with success in managing neonates undergoing upper abdominal or thoracic surgery where avoidance of postoperative ventilation has significant advantages.77 Morphine (10mgkg(cid:2)1) can be combined with a postoperative intravenous naloxone infusion (5mgkg(cid:2)1h(cid:2)1). This practice is suited to major paediatric centres with a large neonatal experiencethatregularlyreportoutcomestothewiderpaediatriccommunity. Proceduralpaininneonates In the first days of life, almost all neonates are subject to two or three painful percutaneous proceduresincludingintramuscularinjection(e.g.,vitaminKinjectionandimmunisation)andheel lanceformetabolicscreening.Thepastdecadehasseenasignificantamountofresearchconfirming theefficacyofseveralstrategiesincludingbreast-feeding,sucrose,78maternalholdingandpacifier use(non-nutritivesucking)79forthesepainfulprocedures.80Clearly,theseinterventionsreducethe size of behavioural and physiological responses to single acute noxious stimuli. It must be emphasised though, that none of these interventions have been shown to have a specific anti- hyperalgesiceffect.Moreover,whilethisremainsso,itisnotpossibletodistinguishasedativeeffect fromatrueanalgesiceffect.12 Cuddling or swaddling, skin-to-skin contact, verbal reassurance and feeding are part of the mothercraftrepertoiretosoothaneonate.Withinacutecareenvironments,separationoftheneonate from its mother oftenprecludes these simple, effective and psychologically important interactions. Mechanical ventilation of the critically ill neonate prolongs this separation and often commits the neonatetorepeatedproceduresduringacuteillness,surgicalinterventionand/orthemanagementof prematurity. Ensuringthatnon-pharmacologicalinterventionsareapartofclinicalpracticewillprovidenotonly effective management of neonatal distress but also much greater satisfaction to both parents and clinicalstaff. Theevidencefortheeffectivenessofsucroseandbreastmilkinreducingpainscoresinnewborn infantsisnowreasonablystrong.Theeffectivenessofsucrosehasbeenconfirmedforheellanceand eyeexaminations.81Postulatedmechanismsremainvaguebutregularlymentionreleaseofendoge- nous neuropeptides including endorphins82 endocannabinoids83 and cholecystokinin.84 An under- standing of the mechanism could provide an explanation for the observed lack of efficacy of these measuresin inflammatoryandpostoperative pain.ACochrane review85confirmed equivalentanal- gesicefficacyofbreastmilkandsucroseandthatthiswasgreaterthanthatofswaddlingandpacifiers. Kangaroocarehasbeenshowntosignificantlyreducebothphysiological86andbehaviouralindices ofpain87,88duringheellance.Kangaroocarewasfoundtobemoreeffectivethanoralglucoseinatleast onerandomisedtrial.87