医疗管理HealthcareManagementSci期刊摘要

Thu-Ba,T,Nguyen,Appa,Iyer,Sivakumar,Stephen,C,Graves.Schedulingrulestoachievelead-timetargetsinoutpatientappointmentsystems[J].HealthCareManagementScience,2017,20(4):578-599Thispaperconsidershowtoscheduleappointmentsforoutpatients,foraclinicthatissubjecttoappointmentlead-timetargetsforbothnewandreturningpatients.Wedevelopheuristicrules,whicharetheexactandrelaxedappointmentschedulingrules,toscheduleeachnewpatientappointment(only)inlightofuncertaintyaboutfuturearrivals.Theschedulingrulesentailtwodecisions.First,therulesneedtodeterminewhetherornotapatient'srequestcanbeaccepted;then,iftherequestisnotrejected,therulesprescribehowtoassignthepatienttoanavailableslot.Theintentoftheschedulingrulesistomaximizetheutilizationoftheplannedresource.,thephysicianstaff),orequivalentlytomaximizethenumberofpatientsthatareadmitted,whilemaintainingtheservicetargetsonthemedian,the95thpercentile,andthemaximumappointmentlead-times.WetesttheproposedschedulingruleswithnumericalexperimentsusingrealdatafromthechosenclinicofTanTockSenghospitalinSingapore.Theresultsshowtheefficiencyandtheefficacyoftheschedulingrules,intermsoftheservice-targetsatisfactionandtheresourceutilization.Fromthesensitivityanalysis,wefindthattheperformanceoftheproposedschedulingrulesisfairlyrobusttothespecificationoftheestablishedlead-timetargets.Michael,Samudra,Erik,Demeulemeester,Brecht,Cardoen.Duetimedrivensurgeryscheduling[J].HealthCareManagementScience,2017,20(3):326-352Inmanyhospitalstherearepatientswhoreceivesurgerylaterthanwhatismedicallyindicated.InoneofEurope’slargesthospitals,theUniversityHospitalLeuven,thisisthecaseforapproximatelyeverythirdpatient.ServingpatientslatecannotalwaysbeavoidedasahighlyutilizedORdepartmentwillsometimessuffercapacityshortage,occasionallyleadingtounavoidabledelaysinpatientcare.Nevertheless,servingpatientslateisaproblemasitexposesthemtoanincreasedhealthriskandhenceshouldbeavoidedwheneverpossible.Inordertoimprovethecurrentsituation,thedelayinpatientschedulinghadtobequantifiedandtheresponsiblemechanism,theschedulingprocess,hadtobebetterunderstood.Drawingfromthisunderstanding,weimplementedandtestedrealisticpatientschedulingmethodsinadiscreteeventsimulationmodel.Wefoundthatitisimportanttomodelnon-electivearrivalsandtoincludeelectivereschedulingdecisionsmadeonsurgerydayitself.ReschedulingensuresthatORrelatedperformancemeasures,suchasovertime,willonlylooselydependonthechosenpatientschedulingmethod.Wealsofoundthatcapacityconsiderationsshouldguideactionsperformedbeforethesurgerydaysuchaspatientschedulingandpatientreplanning.ThisisthecaseasthoseschedulingstrategiesthatensurethatORcapacityisefficientlyusedwillalsoresultinahighnumberofpatientsservedwithintheirmedicallyindicatedtimelimit.AnefficientuseofORcapacitycanbeachieved,forinstance,byservingpatientsfirstcome,firstserved.Asapplyingfirstcome,firstservedmightnotalwaysbepossibleinarealsetting,wefounditisimportanttoallowforpatientreplanning.Chih-Ching,Yang.Measuringhealthindicatorsandallocatinghealthresources:aDEA-basedapproach[J].HealthCareManagementScience,2017,20(3):365-378ThispapersuggestsnewempiricalDEAmodelsforthemeasurementofhealthindicatorsandtheallocationofhealthresources.Theproposedmodelsweredevelopedbyfirstsuggestingapopulation-basedhealthindicator.ByintroducingthesuggestedindicatorintoDEAmodels,anewapproachthatsolvestheproblemofhealthresourceallocationhasbeendeveloped.TheproposedmodelsareappliedtoanempiricalstudyofTaiwan’shealthsystem.Empiricalfindingsshowthatthesuggestedindicatorcansuccessfullyaccommodatethedifferencesinhealthresourcedemandsbetweenpopulations,providingmorereliableperformanceinformationthantraditionalindicatorssuchasphysiciandensity.Usingourmodelsandacommonlyusedallocationmechanism,capitation,toallocatemedicalexpenditures,itisfoundthattheproposedmodelalwaysobtainshigherperformancethanthosederivedfromcapitation,andthesuperiorityincreasesasallocatedexpendituresrise.Lara,Wiesche,Matthias,Schacht,Brigitte,Werners.Strategiesforinterdayappointmentschedulinginprimarycare[J].HealthCareManagementScience,2017,20(3):403-418Whenfacedwithamedicalproblem,patientscontacttheirprimarycarephysician(PCP)first.Heremainlytwotypesofpatientrequestsoccur:non-scheduledpatientswhoarewalk-inswithoutanappointmentandscheduledpatientswithanappointment.Numberandpositionofthescheduledappointmentsinfluencewaitingtimesforpatients,capacityfortreatmentandtheutilizationofPCPs.Asthenumberofpatientrequestsdifferssignificantlybetweenweekdays,thechallengeistomatchcapacitywithpatientrequestsandprovideasfewappointmentslotsasnecessary.Inthisway,capacityforwalk-insismaximizedwhileoverallcapacityrestrictionsaremet.Decisionsastotheoptimalappointmentcapacityperdayonatacticaldecisionlevelhasgainedlittleattentionintheliterature.Amixedintegerlinearmodelisdeveloped,wheretheminimumnumberofappointmentsscheduledforaweeklyprofileisdetermined.WearethusabletogivetheanswerastohowmanyappointmentstoofferoneachdayinaweekinordertocreateaschedulethattakespatientpreferencesaswellasPCPpreferencesintoaccount.Appointmentschedulesareofteninfluencedbyuncertaindemandsduetothenumberofurgentpatients,interarrivalsandservicetimes.Basedonanexemplarycasestudy,theadvantagesoftheoptimalappointmentscheduleondifferentperformancecriteriaareshownbydetailedstochasticsimulations.Muge,Capan,Julie,S,Ivy,James,R,Wilson.Astochasticmodelofacute-caredecisionsbasedonpatientandproviderheterogeneity[J].HealthCareManagementScience,2017,20(2):187-206Theprimarycauseofpreventabledeathinmanyhospitalsisthefailuretorecognizeand/orrescuepatientsfromacutephysiologicdeterioration(APD).APDaffectsallhospitalizedpatients,potentiallycausingcardiacarrestanddeath.IdentifyingAPDisdifficult,andresponsetimingiscritical-delaysinresponserepresentasignificantandmodifiablepatientsafetyissue.Hospitalshaveinstitutedrapidresponsesystemsorteams(RRT)toprovidetimelycriticalcareforAPD,withthresholdsthattriggertheinvolvementofcriticalcareexpertise.TheNationalEarlyWarningScore(NEWS)wasdevelopedtodefinethesethresholds.However,currenttriggersareinconsistentandignorepatient-specificfactors.Further,acutecareisdeliveredbyproviderswithdifferentclinicalexperience,resultinginquality-of-carevariation.Thisarticledocumentsasemi-MarkovdecisionprocessmodelofAPDthatincorporatespatientandproviderheterogeneity.Themodelallowsforstochasticallychanginghealthstates,whiledeterminingpatientsubpopulation-specificRRT-activationthresholds.Theobjectivefunctionminimizesthetotaltimeassociatedwithpatientdeteriorationandstabilization;andtherelativevaluesofnursingandRRTtimescanbemodified.AcasestudyfromJanuary2011toDecember2012identifiedsixsubpopulations.RRTactivationwasoptimalforpatientsin“slightlyconcerning”healthstates(NEWS>0)forallsubpopulations,exceptsurgicalpatientswithlowriskofdeteriorationforwhomRRTwasactivatedin“concerning”states(NEWS>4).ClusteringmethodsidentifiedproviderclustersconsideringRRT-activationpreferencesandestimationofstabilization-relatedresourceneeds.ProviderswithconservativeresourceestimatespreferredwaitingoveractivatingRRT.Thisstudyprovidessimplepracticalrulesforpersonalizedacutecaredelivery.Sebastian,Hof,Andreas,Fügener,Jan,Schoenfelder.Casemixplanninginhospitals:areviewandfutureagenda[J].HealthCareManagementScience,2017,20(2):207-220Thecasemixplanningproblemdealswithchoosingtheidealcompositionandvolumeofpatientsinahospital.Withmanycountrieshavingrecentlychangedtosystemswherehospitalsarereimbursedforpatientsaccordingtotheirdiagnosis,casemixplanninghasbecomeanimportanttoolinstrategicandtacticalhospitalplanning.Selectingpatientsinsuchapaymentsystemcanhaveasignificantimpactonahospital’srevenue.Thecontributionofthisarticleistoprovidethefirstliteraturereviewfocusingonthecasemixplanningproblem.Wedescribetheproblem,distinguishitfromsimilarplanningproblems,andevaluatetheexistingliteraturewithregardtoproblemstructureandmanagerialimpact.Further,weidentifygapsintheliterature.Wehopetofosterresearchinthefieldofcasemixplanning,whichonlylatelyhasreceivedgrowingattentiondespiteitsfundamentaleconomicimpactonhospitals.Jeong,Hoon,Choi,Imsu,Park,Ilyoung,Jung,Asoke,Dey.Complementaryeffectofpatientvolumeandqualityofcareonhospitalcostefficiency[J].HealthCareManagementScience,2017,20(2):221-231Thisstudyexploresthedirecteffectofanincreaseinpatientvolumeinahospitalandthecomplementaryeffectofqualityofcareonthecostefficiencyof.hospitalsintermsofpatientvolume.Thesimultaneousequationmodelwiththree-stageleastsquaresisusedtomeasurethedirecteffectofpatientvolumeandthecomplementaryeffectofqualityofcareandvolume.Costefficiencyismeasuredwithadataenvelopmentanalysismethod.PatientvolumehasaU-shapedrelationshipwithhospitalcostefficiencyandaninvertedU-shapedrelationshipwithqualityofcare.Qualityofcarefunctionsasamoderatorfortherelationshipbetweenpatientvolumeandefficiency.Thispaperaddressestheeconomicallyimportantquestionoftherelationshipofvolumewithqualityofcareandhospitalcostefficiency.Thethree-stageleastsquaresimultaneousequationmodelcapturesthesimultaneouseffectsofpatientvolumeonhospitalqualityofcareandcostefficiency.Nathaniel,D,Bastian,Tahir,Ekin,Hyojung,Kang.Stochasticmulti-objectiveauto-optimizationforresourceallocationdecision-makinginfixed-inputhealthsystems[J].HealthCareManagementScience,2017,20(2):246-264Themanagementofhospitalswithinfixed-inputhealthsystemssuchasthe.MilitaryHealthSystem(MHS)canbechallengingduetothelargenumberofhospitals,aswellastheuncertaintyininputresourcesandachievableoutputs.Thispaperintroducesastochasticmulti-objectiveauto-optimizationmodel(SMAOM)forresourceallocationdecision-makinginfixed-inputhealthsystems.Themodelcanautomaticallyidentifywheretore-allocatesysteminputresourcesatthehospitallevelinordertooptimizeoverallsystemperformance,whileconsideringuncertaintyinthemodelparameters.Themodelisappliedto128hospitalsinthethreeservices(AirForce,Army,andNavy)intheMHSusinghospital-leveldatafrom2009–2013.Theresultsarecomparedtothetraditionalinput-orientedvariablereturns-to-scaleDataEnvelopmentAnalysis(DEA)model.TheapplicationofSMAOMtotheMHSincreasestheexpectedsystem-widetechnicalefficiencyby18%overtheDEAmodelwhilealsoaccountingforuncertaintyofhealthsysteminputsandoutputs.Thedevelopedmethodisusefulfordecision-makersintheDefenseHealthAgency(DHA),whohaveastrategiclevelobjectiveofintegratingclinicalandbusinessprocessesthroughbettersharingofresourcesacrosstheMHSandthroughsystem-widestandardizationacrosstheservices.ItisalsolesssensitivetodataoutliersorsamplingerrorsthantraditionalDEAmethods.Ridvan,Gedik,Shengfan,Zhang,Chase,Rainwater.Strategiclevelprotontherapypatientadmissionplanning:aMarkovdecisionprocessmodelingapproach[J].HealthCareManagementScience,2017,20(2):286-302Arelativelynewconsiderationinprotontherapyplanningistherequirementthatthemixofpatientstreatedfromdifferentcategoriessatisfydesiredmixpercentages.Deviationsfromthesepercentagesandtheirimpactsonoperationalcapabilitiesareofparticularinteresttohealthcareplanners.Inthisstudy,weinvestigateintelligentwaysofadmittingpatientstoaprotontherapyfacilitythatmaximizethetotalexpectednumberoftreatmentsessions(fractions)deliveredtopatientsinaplanningperiodwithstochasticpatientarrivalsandpenalizethedeviationfromthepatientmixrestrictions.WeproposeaMarkovDecisionProcess(MDP)modelthatprovidesveryusefulinsightsindeterminingthebestpatientadmissionpoliciesinthecaseofanunexpectedopeninginthefacility.,no-shows,appointmentcancellations,etc.).Inordertoovercomethecurseofdimensionalityforlargerandmorerealisticinstances,weproposeanaggregateMDPmodelthatisabletoapproximateoptimalpatientadmissionpoliciesusingthewordedweightaggregationtechnique.OurmodelsareapplicabletohealthcaretreatmentfacilitiesthroughouttheUnitedStates,butaremotivatedbycollaborationwiththeUniversityofFloridaProtonTherapyInstitute(UFPTI).Xiangyong,Li,N,Rafaliya,M,Fazle,Baki,Ben,A,Chaouch.Schedulingelectivesurgeries:thetradeoffamongbedcapacity,waitingpatientsandoperatingroomutilizationusinggoalprogramming[J].HealthCareManagementScience,2017,20(1):33-54Schedulingofsurgeriesintheoperatingroomsunderlimitedcompetingresourcessuchassurgicalandnursingstaff,anesthesiologist,medicalequipment,andrecoverybedsinsurgicalwardsisacomplicatedprocess.Awell-designedscheduleshouldbeconcernedwiththewelfareoftheentiresystembyallocatingtheavailableresourcesinanefficientandeffectivemanner.Inthispaper,wedevelopanintegerlinearprogrammingmodelinamannerusefulformultiplegoalsforoptimallyschedulingelectivesurgeriesbasedontheavailabilityofsurgeonsandoperatingroomsoveratimehorizon.Inparticular,themodelisconcernedwiththeminimizationofthefollowingimportantgoals:(1)theanticipatednumberofpatientswaitingforservice;(2)theunderutilizationofoperatingroomtime;(3)themaximumexpectednumberofpatientsintherecoveryunit;and(4)theexpectedrange(thedifferencebetweenmaximumandminimumexpectednumber)ofpatientsintherecoveryunit.Wedeveloptwogoalprogramming(GP)models:lexicographicGPmodelandweightedGPmodel.ThelexicographicGPmodelschedulesoperatingroomswhenvariouspreemptiveprioritylevelsaregiventothesefourgoals.Anumericalstudyisconductedtoillustratetheoptimalmaster-surgeryscheduleobtainedfromthemodels.Thenumericalresultsdemonstratethatwhentheavailablenumberofsurgeonsandoperatingroomsisknownwithouterrorovertheplanninghorizon,theproposedmodelscanproducegoodschedulesandprioritylevelsandpreferenceweightsoffourgoalsaffecttheresultingschedules.Theresultsquantifythetradeoffsthatmusttakeplaceasthepreemptive-weightsofthefourgoalsarechanged.Yen-Yi,Feng,I-Chin,Wu,Tzu-Li,Chen.Stochasticresourceallocationinemergencydepartmentswithamulti-objectivesimulationoptimizationalgorithm[J].HealthCareManagementScience,2017,20(1):55-75Thenumberofemergencycasesoremergencyroomvisitsrapidlyincreasesannually,thusleadingtoanimbalanceinsupplyanddemandandtothelong-termovercrowdingofhospitalemergencydepartments(EDs).However,currentsolutionstoincreasemedicalresourcesandimprovethehandlingofpatientneedsareeitherimpracticalorinfeasibleintheTaiwaneseenvironment.Therefore,EDsmustoptimizeresourceallocationgivenlimitedmedicalresourcestominimizetheaveragelengthofstayofpatientsandmedicalresourcewastecosts.Thisstudyconstructsamulti-objectivemathematicalmodelformedicalresourceallocationinEDsinaccordancewithemergencyfloworprocedure.Theproposedmathematicalmodeliscomplexanddifficulttosolvebecauseitsperformancevalueisstochastic;furthermore,themodelconsidersbothobjectivessimultaneously.Thus,thisstudydevelopsamulti-objectivesimulationoptimizationalgorithmbyintegratinganon-dominatedsortinggeneticalgorithmII(NSGAII)withmulti-objectivecomputingbudgetallocation(MOCBA)toaddressthechallengesofmulti-objectivemedicalresourceallocation.NSGAIIisusedtoinvestigateplausiblesolutionsformedicalresourceallocation,andMOCBAidentifieseffectivesetsoffeasiblePareto(non-dominated)medicalresourceallocationsolutionsinadditiontoeffectivelyallocatingsimulationorcomputationbudgets.ThediscreteeventsimulationmodelofEDflowisinspiredbyaTaiwanhospitalcaseandisconstructedtoestimatetheexpectedperformancevaluesofeachmedicalallocationsolutionasobtainedthroughNSGAII.Finally,computationalexperimentsareperformedtoverifytheeffectivenessandperformanceoftheintegratedNSGAIIandMOCBAmethod,aswellastoderivenon-dominatedmedicalresourceallocationsolutionsfromthealgorithms.Andreas,Fügener,Sebastian,Schiffels,Rainer,Kolisch.Overutilizationandunderutilizationofoperatingrooms-insightsfrombehavioralhealthcareoperationsmanagement[J].HealthCareManagementScience,2017,20(1):115-128Theplanningofsurgerydurationsiscrucialforefficientusageofoperatingtheaters.Bothplanningtoolongandtooshortdurationsforsurgeriesleadtoundesirableconsequences,.idletime,overtime,orreschedulingofsurgeries.Wedefinetheseconsequencesasoperatingroominefficiency.Theoverallobjectiveofplanningsurgerydurationsistominimizeexpectedoperatingroominefficiency,sincesurgerydurationsarestochastic.Whilemosthealthcarestudiesassumeeconomicallyrationalbehaviorofdecisionmakers,experimentalstudieshaveshownthatdecisionmakersoftendonotactaccordingtoeconomicincentives.Basedoninsightsfromhealthcareoperationsmanagement,medicaldecisionmaking,behavioraloperationsmanagement,aswellasempiricalobservations,wederivehypothesesthatsurgeons’behaviordeviatesfromeconomicallyrationalbehavior.Toinvestigatethis,weundertakeanexperimentalstudywhereexperiencedsurgeonsareaskedtoplansurgerieswithuncertaindurations.Wediscoversystematicdeviationsfromoptimaldecisionmakingandofferbehavioralexplanationsfortheobservedbiases.Ourresearchprovidesnewinsightstotackleamajorprobleminhospitals,.lowoperatingroomutilizationgoingalongwithstaffovertime.Kayse,Lee,Maass,Boying,Liu,Mark,S,Daskin,Mary,Duck.Incorporatingnurseabsenteeismintostaffingwithdemanduncertainty[J].HealthCareManagementScience,2017,20(1):141-155Increasednurse-to-patientratiosareassociatednegativelywithincreasedcostsandpositivelywithimprovedpatientcareandreducednurseburnoutrates.Thus,itiscriticalfromacost,patientsafety,andnursesatisfactionperspectivethatnursesbeutilizedefficientlyandeffectively.Toaddressthis,weproposeastochasticprogrammingformulationfornursestaffingthataccountsforvariabilityinthepatientcensusandnurseabsenteeism,day-to-daycorrelationsamongthepatientcensuslevels,andcostsassociatedwiththreedifferentclassesofnursingpersonnel:unit,pool,andtemporarynurses.Thedecisionstobemadeinclude:howmanyunitnursestoemploy,howlargeapoolofcross-trainednursestomaintain,howtoallocatethepoolnursesonadailybasis,andhowmanytemporarynursestoutilizedaily.Ageneticalgorithmisdevelopedtosolvetheresultingmodel.Preliminaryresultsusingdatafromalargeuniversityhospitalsuggestthattheproposedmodelcansaveafour-unitpoolhundredsofthousandsofdollarsannuallyasopposedtothecrudeheuristicsthehospitalcurrentlyemploys.Margaret,L,Brandeau.Creatingimpactwithoperationsresearchinhealth:makingroomforpracticeinacademia[J].HealthCareManagementScience,2016,19(4):305-312Operationsresearch(OR)-basedanalyseshavethepotentialtoimprovedecisionmakingformanyimportant,real-worldhealthcareproblems.However,juniorscholarsoftenavoidworkingonpracticalapplicationsinhealthbecausepromotionandtenureprocessestendtovaluetheoreticalstudiesmorehighlythanappliedstudies.Thispaperdiscussestheauthor’sexperiencesinusingORtoinformandinfluencedecisionsinhealthandprovidesablueprintforjuniorresearcherswhowishtofindsuccessbytakingasimilarpath.Thisinvolvesselectinggoodproblemstostudy,formingproductivecollaborationswithdomainexperts,developingappropriatemodels,identifyingthemostsalientresultsfromananalysis,andeffectivelydisseminatingfindingstodecisionmakers.Thepaperthensuggestshowjournals,fundingagencies,andsenioracademicscanencouragesuchworkbytakingabroaderandmoreinformedviewofthepotentialroleandcontributionsofORtosolvinghealthcareproblems.MakingroominacademiafortheapplicationofORinhealthfollowsinthetraditionbegunbythefoundersofoperationsresearch:toworkonimportantreal-worldproblemswhereoperationsresearchcancontributetobetterdecisionmaking.Adam,Pilny,Magdalena,A,Stroka.Determinantsofreceivedlong-termcare–Individualresponsestoregionalnursinghomeprovisions[J].HealthCareManagementScience,2016,19(4):326-337Existingliteratureanalyzingthechoiceofreceivedlong-termcarebyfrailelderly(65+years)predominantlyfocusesonphysicalandpsychologicalconditionsofelderlypeopleasfactorsthatinfluencethedecisionforaparticulartypeofcare.Untilnow,however,theregionalin-patientlong-termcaresupplyhasbeenneglectedasinfluentialfactorinthisdecision-makingprocess.Inthisstudy,weanalyzethedeterminantsofreceivedlong-termcareinGermanybyexplicitlytakingtheregionalsupplyofnursinghomesintoaccount.Weestimateamultinomialprobitmodeltoillustratethisdecision-makingprocess.Therefore,withinthisdiscretechoicesettingwedistinguishbetweenallavailabletypesoflong-termcareinGermany,.fourdifferenttypesofformalandinformalcareprovision.Wefindthatthedecisionforlong-termin-patientcareissignificantlycorrelatedwiththeregionalsupplyofnursinghomebeds,whilecontrollingforphysicalandpsychologicalconditionsoftheindividual.Cynthia,Williams,Yara,Asi,Amanda,Raffenaud.Theeffectofinformationtechnologyonhospitalperformance[J].HealthCareManagementScience,2016,19(4):338-346Whilehealthcareentitieshaveintegratedvariousformsofhealthinformationtechnology(