The impact of health care financing reform on the

TheimpactofhealthcarefinancingreformontheproductivitychangeinFinnishhospitalsMiikaLinnaNationalResearchandDevelopmentCentreforWelfareandHealth,Helsinki,Finland______________________________________________________AbstractThepurposeofthisstudywastoanalysethedevelopmentofproductivityandefficiencyintheproductionofhospitalcareinFinlandduringtheperiod1988-1994.Aspecialinterestwasdirectedtotheimpactofhealthcarefinancingreformin1993.TheanalysisisbasedontheMalmquistindexapproachusinglinearprogramming.Positiveproductivitychangeswerefoundparticularlytowardstheendoftheobservationperiod,althoughafewyearsdidnotshowsignificantimprovementsinproductivityandefficiency.Therewasasignificantlyhigherrateofproductivity1992-1993and1993-1994,suggestingthatthestatesubsidyreformin1993mayhavestrengthenedhospitals’effortstoimproveperformance.Theuseofoutputbasedreimbursementwasnotmarkedlyassociatedwithincreasedefficiencyorproductivity.JELClassification:C14;C23;D24;I10Keywords:Hospitalfinancing,Hospitals,Productivity,Malmquistindices______________________________________________________Addressforcorrespondence:MiikaLinnaNationalResearchandDevelopmentCentreforWelfareandHealthSiltasaarenkatu18,POBOX22000531Helsinki,FinlandPhone+358939672295,telefax+358939672485,internet:[e-mail]2INTRODUCTIONMostwesterncountrieshaveexperiencedasteadygrowthofhealthcareexpenditureoverthepastfewyears.InFinland,wherehospitalsaccountfor55percentoftotalhealthcareexpenditure,thegovernment’sbudgetdeficithasimposedunrelentingpressureonhospitalstocontaincosts.Theusualexplanationsforincreasingcostshavebeenlackofincentivesinthehospitalreimbursementsystem,theadoptionofcostlynewmedicaltechnologiesandanageingpopulationalongwiththepossibilityoftreatingsickerpatients.However,therehasbeensurprisinglylittlestudyoftheefficiencychangesmadeintheorganisationofhealthcaredelivery.Risingcostsmaybeduetoincreasinginefficiencyratherthantechnologicalimprovements.Onecontroversialissueinhealthpoliticshasbeentheroleofgovernmentintheregulationofhealthcaredelivery.Movestowardsderegulationormanagedmarketsarecommonlyjustifiedbyefficiency(andhence,productivity)gains.TheFinnishstatesubsidyreformof1993wasexpectedtoenhanceproductiveefficiencybyintroducingcompetitiveelementsintohealthcare.Thereformallowedthemunicipalitiestoadoptamoreactiveroleaspurchasersofhospitalservicesastheybecamethebudgetholdersofpublichealthcaremoney.Thischangeinthefinancingsysteminvolvedtheintroductionofperformancebasedreimbursementanditencouragedmunicipalitiestoselectivecontracting.Thepressuretoadoptnewandcostlymedicaltechnologyhasbeenparticularlyacuteintherealmofspecializedcareprovidedbyhospitals.However,althoughtechnologicalchangeshouldgenerallyhaveapositiveimpactonproductivity(a3positivechangeintheproductionpossibilitiesfrontier),thereislittleempiricalevidenceofthis.Theadoptionanddiffusionofnewtechnologiesrequiressubstantialeffortsfromhospitals,includingtheprovisionofeducationformedicalandnursingstudentsandthetestingofnewtechnologiesbyclinicalresearch.Worldwide,theoverallimpactofteachingandresearchonhospitalcostshasbeenestimatedtovarybetweenniland25percent(Linnaetal.,1998),thoughforteachinghospitalsmostestimatesliebetween7and15percent.Theactualreimbursementofteachingandresearchindifferentcountriesvariesfrom8to22percentofhospitals'recurrentcosts.TherehasbeenonlyfewattemptstoevaluatethedevelopmentofhospitalproductivityinFinland,usuallyinvolvingunacceptablemeasuresforproducedoutput(Alanderetal.,1990).Hospitalproductivitystudieselsewherehavehardlysucceededanybetter,clearlyduetoincompletestatistics:bed-daysorthenumberofadmissions(withoutcase-mixstandardisationtheseareinadequatemeasuresofhospitaloutput)havebeenusedinthemajorityofproductivitystudies.Teachingandresearchoutputisusuallyignoredandthereareonlyafewexampleswhereplausiblemethodsforcase-mixstandardisationhavebeenused(Magnussen,1996;BurgessandWilson,1996).Thereisclearlyaneedformorerealisticproductivityestimatesbasedontheoreticallymoresatisfactoryoutputmeasures.Inthisstudyitwaspossibletoobservetheeffectoftechnologicalchangebymeasuringtheshiftintheproductionpossibilitiesfrontier.NonparametricmethodswereusedtocalculateMalmquistindicesofproductivitychange.Themajorpatientoutput-inpatientadmissions-wasderivedfromdischargedatafromallinpatientson4thebasisofcase-mixcomplexity(DiagnosticRelatedGroups,DRGs).Inaddition,teachingandresearchoutputwasmeasuredmoreaccuratelythaninpreviousproductivitystudies.BACKGROUNDProductivityiscommonlydefinedasanindexofoutputsdividedbyanindexofinputs.Themethodsformeasuringproductivitydifferinhowtheweightsfortheseindexesaredetermined.Inthecaseofhospitalsproducingmultipleoutputs,factorpricesarenottypicallyavailableordonotreflectthemarginalcosts.Furthermore,iftheissueoftechnologicalchangeistobeaddressed,themeasurementtechniqueshouldbecapableofseparatingtheeffectsofefficiencychangeandfrontierchangeinapaneldatasetting.Thereareseveralalternativemethodsforhandlingtheserestrictionsandmakingitpossibletodecomposeproductivitychangeintoitssources-ef