multidisciplinary clinical approach The medication discrepancy detection service: Acost-effective Atención Primaria

(1)

www.elsevier.es/ap

Atención Primaria

ORIGINAL ARTICLE

The medication discrepancy detection service: A cost-effective multidisciplinary clinical approach

Ainhoa O˜ natibia-Astibia

^a^,^b^,∗

, Amaia Malet-Larrea

^a

, Amaia Mendizabal

^c

,

Elena Valverde

^c

, Belen Larra˜ naga

^a

, Miguel Ángel Gastelurrutia

^d

, Martín Ezcurra

^e

, Leire Arbillaga

^a

, Bego˜ na Calvo

^b

, Estibaliz Goyenechea

^a

aOfﬁcialPharmacistAssociationofGipuzkoa,Prim2,20006Donostia/SanSebastian,Spain

bPharmaceuticalTechnologyDepartment,FacultyofPharmacy,UniversityoftheBasqueCountry,UPV/EHU,P.Universidad7, 01006Vitoria,Spain

cPrimaryCarePharmacy,BidasoaIntegratedHealthcareOrganisation(Osakidetza),Spain

dPharmaceuticalCareResearchGroup,FacultyofPharmacy,UniversityofGranada,CampusUniversitariodeCartuja,18071 Granada,Spain

eMartinEzcurraFernandezPharmacy,Harmugarrieta2,20305Irun,Spain

Received10October2019;accepted15April2020 Availableonline29September2020

KEYWORDS

Communitypharmacy services;

Medicationerror;

Ambulatorycare

Abstract

Objective: ToestimatetheeffectivenessofaMedicationDiscrepancyDetectionService(MDDS), acollaborativeservicebetweenthecommunitypharmacyandPrimaryCare.

Design:Non-controlledbefore-and-afterstudy.

Setting: BidasoaIntegratedHealthcareOrganisation,Gipuzkoa,Spain.

Participants:Theservicewasprovidedbyamultidisciplinarygroupofcommunitypharmacists (CPs),generalpractitioners(GPs),andprimarycarepharmacists,topatientswithdiscrepancies betweentheiractivemedicalchartsandmedicinesthattheywereactuallytaking.

Outcomes: Theprimaryoutcomeswerethenumberofmedicines,thetypeofdiscrepancy,and GPs’decisions.SecondaryoutcomesweretimespentbyCPs,emergencydepartment(ED)visits, hospitaladmissions,andcosts.

Results:The MDDS was provided to 143 patients, and GPs resolved discrepancies for 126 patients.CPsidentiﬁed259discrepancies,amongwhichthemainonewas patientsnottak- ingmedicineslistedontheiractivemedicalcharts(66.7%,n=152).ThemainGPs’decisionwas towithdrawthetreatment(54.8%,n=125),whichmeantthatthenumberofmedicinesper patientwasreducedby0.92(9.12±3.82vs.8.20±3.81;p<.0001).ThenumberofEDvisits andhospitaladmissionsperpatientwerereducedby0.10(0.61±.13vs0.52±0.91;p=.405 and0.17(0.33±0.66vs.0.16±0.42;p=.007),respectively.Thecostperpatientwasreduced byD444.9(D1003.3±2165.3vs.D558.4±1273.0;p=.018).

∗Correspondingauthor.

E-mailaddress:[email protected](A.O˜natibia-Astibia).

https://doi.org/10.1016/j.aprim.2020.04.008

creativecommons.org/licenses/by-nc-nd/4.0/).

(2)

Conclusion:The MDDSresultedinareductioninthenumber ofmedicinesper patientsand numberofhospitaladmissions,andtheservicewasassociatedwithaffordable,cost-effective ratios.

PALABRASCLAVE Errordemedicación;

Servicios profesionales farmacéuticos comunitarios;

Atenciónprimaria

Serviciodedeteccióndediscrepanciasdelamedicación:unestudiomultidisciplinar ycoste-efectivo

Resumen

Objetivos: Estimarlaefectividaddelserviciodedeteccióndediscrepanciasdelamedicación, unserviciodecolaboraciónentrelafarmaciacomunitariaylaatenciónprimaria.

Dise˜no:Estudiodeintervenciónantes-después,singrupocontrol.

Emplazamiento:OrganizaciónSanitariaIntegradadeBidasoa,Gipuzkoa,Espa˜na.

Participantes: El serviciofue ofrecido por un grupo multidisciplinar que incluía farmacéu- ticos comunitarios (FC), médicos de atención primaria (MAP) y farmacéuticos de atención primariaapacientesquepresentabandiscrepanciasentrelamedicaciónprescritaenlahoja detratamientoactivoyloquerealmenteestabantomando.

Medicionesprincipales: Lasvariablesprincipalesdelestudiofueronelnúmerodemedicamen- tos,tipodediscrepanciayladecisióndelMAP.Lasvariablessecundariasfuerontiempoinvertido porelfarmacéutico,visitasalserviciodeurgencias,ingresoshospitalariosyloscostes.

Resultados: Elservicioseofrecióa143pacientes,yelMAPresolviólasdiscrepanciasdeuntotal de126pacientes.ElFCidentiﬁcó259discrepanciasdelascualeslamayoríafuequeelpaciente noestabatomandounmedicamentoprescrito(66,7%,n=152).Enlamayoríadeloscasos,la decisióndelMAPfuesuspendereltratamiento(54,8%,n=125);elnúmerodemedicamentosque tomabaelpacienteseredujoenun0,92(9,12±3,82vs.8,20±3,81;p<0,0001).Elnúmerode visitasalhospitalylosingresoshospitalariosseredujeronen0,10(0,61±0,13vs.0,52±0,91;

p=0,405)y0,17puntos(0,33±0,66vs.0,16±0,42;p=0,007),respectivamente.Elcostepor pacienteseredujoen444,9D (1.003,3±2.165,3vs.558,4D±1.273,0;p=0,018).

Conclusión:Elservicioredujoelnúmerodemedicamentosquetomabaelpacienteeingresos hospitalariosyestoserelacionóconunosratiosdecoste-efectividadpositivos.

Introduction

Medicationerrors(ME)areamongthetop10causesofdeath worldwide.¹Sucherrorscancausepatientsafetyincidents, whichare associatedwitha higherrate ofhospitalisation andincreasedmorbidityandmortality,accountingformore than1%oftotalglobalhealthexpenditures.²MEisthesin- glemost commonpreventable causeof adverse events in medicationpracticeandamajorpublichealthburden,with anestimatedannualcostinEuropeofD4.5billiontoD21.8 billion.³DuetothehealthandeconomicimpactofME,the WorldHealthOrganisation(WHO)hasincludedthereduction ofMEintheGlobalPatientSafetyChallenge.⁴

MEhasbeendeﬁnedas‘anypreventableeventthatmay causeor lead to inappropriate medication use or patient harmwhile themedication isin thecontrol ofthe health careprofessional,patient,orconsumer’.⁵Contributingfac- tors may be associated with health care professionals, patients, the work environment, medicines, comput- erisedinformationsystems,and/orprimary---secondarycare communication.⁶Reducingthefrequencyandimpactofpre- ventableharmrelatedtomedicinesastheconsequenceof

error,accident,orcommunicationproblemwillcontribute to the achievement of medication safety for patients.⁷ Statistics show that these strategies will lead to 95,000 fewerdeathsperyearinEurope.²

Variousstrategies to reduce MEin the community setting have been proposed in recent years; they include medication review and reconciliation services, the use of automated informationsystems, education,and multi- component interventions.^8---10 The effectivenessof clinical pharmacistsin identifyingMEhasbeen demonstrated,but datafromprimarycarearerelativelyscarceandfewstudies have includedcommunitypharmacists(CPs).^11---13 Thislack of research amongCPs and the previous experience that these professionals have in other services¹⁴ have led the WHOtoconsidertheinvolvementofCPsintheprioritisation ofstrategiestoreduceMEinprimarycare.⁶

Inthis context, tomeet the need for high-quality and cost-effective identiﬁcation of medication discrepancies, a medication discrepancy detection service (MDDS) was designed. To ensure patient-centred care, collaboration amongdifferenthealthprofessionalsisneeded.¹⁵TheMDDS is offered by a multidisciplinary team including CPs and

(3)

general practitioners (GPs) in collaboration with primary carepharmacistsandprimarycarenurses.Theidentiﬁcation ofmedicationdiscrepanciesisawaytodetectME,andCPsin Spainareideallypositionedtodoso,astheyhaveaccessto electronicmedicalrecordsandareresponsiblefordispens- ingmedicines.Therefore,theaimofthepresentstudywas toevaluatetheimpactonthenumberof medicineintake andthecosteffectivenessoftheMDDSasimplementedcol- laborativelyinthecommunitypharmacyandprimarycare servicessettings.

Methods

Studydesignandethicalapproval

Thisnon-controlledbefore-and-afterstudy.wasundertaken betweenOctober2015andSeptember2016intheBidasoa Integrated Healthcare Organisation, Spain, which is com- prisedofoneregionalhospitalandthreeprimarycareunits.

The multidisciplinaryprofessionalgroupthat providedthe MDDSconsistedofCPs,primarycarepharmacists,GPs,and hospitalspecialists.AlltheCPsof thepharmacieslocated inthemunicipalitiesattendedbytheIntegratedHealthcare Organisation,wereinvitedtoparticipateintheproject.CPs and GPs attended a 2-hour workshop that presented and described the study protocol. The protocol for this study wasapprovedbytheEthicsCommitteeforClinicalResearch of theBasque Country(PI2015080EPA-SP)and wasinline withtheHelsinkiDeclaration.Allparticipantsprovidedwrit- teninformedconsentatthe timeoftheirenrolment, and CPs delivered information sheets explaining the study to patientswhometthestudycriteria.

Patients

Patients were recruited according the following criteria:

patientsthathadadiscrepancybetweentheiractivemedi- calchartsandthemedicinestheywereactuallytaking.CP identiﬁedthispatients withdiscrepancieslike(i)patients nottakingmedicationsthatappearedintheircharts,(ii)tak- ingmedicationsthatdidnotappearintheircharts,(iii)not followingthe prescribeddosage regime,(iv)notfollowing theprescribedposologyand(v)duplicatedtreatment.

Studyprocedureandhealthoutcomes

CPs offeredthe MDDS service topatients in whom at the time of dispensing they identifieda discrepancy between theiractivemedicalchartandthemedicinestheyweretak- ing.CPsregisteredeachparticipatingpatient’sname,health identificationnumber,willingnesstoparticipateinthestudy, and date of first appointment (record 1). Patients were asked to bring all current medications, including dietary and other products, tothe pharmacy.The CPs performed aclinicalinterviewsandcheckedbrownbagsFortheinter- view,thepharmacistusedaguideconsistingofstructured questions that allowed tocollect asmuch informationas possibleabout takingprescribedmedications,other medications,supplements,creamsorotherproducts.Thebrown bag checking consisted of checking an inventory of the

Service offer

Reject Accept

Clinical interview

Evaluation of the patients’ situation

Report

Clinical interview

Medical chart update Community pharmacistGeneral practitioner

Figure1 Flowchartofthestudyprocedure.

medicationstaken by each patientbased on the medica- tionpackages.At thetime ofthe clinical interview, each patientprovidedwritteninformedconsent.Ifapatientdid notreturnforthescheduledappointment,itwasrecorded as‘‘rejected’’.Aftertheclinicalinterview,theinformation wascomparedwiththepatient’smedicalchartandtheCP preparedareportinwhichalldetecteddiscrepancieswere registered.Oncethe CPevaluatedthepatient’ssituation, the CP completed the report and sent it to the primary carepharmacist.Timeinvestedintheclinicalinterviewand reportpreparationwasalsoregistered.

Uponreceivingthereport,theprimarycarepharmacist contactedthecorrespondingprimarycarenurse,whocited thepatientwiththeGP.TheGPconductedaclinicalinter- viewandwasresponsibleformakinganynecessarychanges tothemedicalchartintheelectronicprescribingsystem.If amedicalspecialistwasresponsiblefortheprescription,the primarycare pharmacist contacted directly by telephone tosolvethe problem.Pharmacotherapeutic changeswere madeinagreementwiththepatient,andtheGPmadesure thatthepatientunderstoodthenewtreatment(Fig.1).Dis- crepant medications were classiﬁed using the Anatomical TherapeuticChemicalsystem.

Primarycarepharmacistscompiledandrecordedalldata, andwereresponsiblefor registeringdiscrepancies andfor ensuringthattheﬂowchart wasfollowedcorrectly.Emer- gency department (ED) visits and hospital admissions 6 monthsbeforeandafter theintervention wereregistered attheendofthestudyperiodusinghospitalrecords.

Economicoutcomes

Aneconomic evaluation wasconductedfromthe National HealthSystem(NHS)perspective.Thecostandeffectiveness oftheservicewasanalysed.Thedirectcostsofmedications (includingdiscrepant medications), ED visits and hospital admissions 6 months before and after the intervention,

(4)

and interventions costs were included. The numbers of medicines, ED visits, and hospital admissions served as effectivenessvariables.Costswere estimatedusingposol- ogyand thepricesofthe medicines.The costsassociated withEDvisits wereestimatedbasedontheBasqueHealth Service(BHS)rates.^16---18 Thediagnosis-relatedgroup(DRG) wasidentiﬁedforeachhospitaladmission.DRGsmakeupan establishedpaymentsystemforgroupsofpatientswithsimi- larclinicalcharacteristicswhoareexpectedtohavesimilar healthresourceconsumption.⁴ The costfor eachDRG was determinedusingBHSrates.^16---19Thetotalcostofeachinter- ventionincludedcostsassociatedwith:(i)thetimespentby theCPontheclinicalinterview,(ii)thetimespentbytheCP tocompletethereport,(iii)thecostofGPconsultation(iv) thecost of hospital telephonespecialist consultation and (v)thecostofthetimespentbyprimarycarepharmacists.

Costs (i) and (ii)were estimated using collective CP bar- gainingdata.Costs (iii)and(iv)wereestimated usingBHS rates.¹⁶ Allcosts wereexpressedin eurosand updatedto 2017usingtheSpanishRetailPriceIndex.Theincremental cost-effectivenessratio(ICER)wascalculated tocompare costsbeforeandaftertheintervention.

Statisticalanalysis

Changesinthenumbersofmedicines,EDvisits,andhospital admissionswereevaluatedandcomparedbeforeandafter MDDS implementation withthe paired ttest or Student’s ttest for parametric variables. The chi-squared test and Fisher’sexacttestwereusedtoanalysethefrequencydistri- butionsofthestudyvariables.Aone-waysensitivityanalysis wasconductedtoexaminetheimpactsofthestudyvariables ontheresultsoftheeconomicevaluation.Generaldataare expressedasmeans±standarddeviations.Statisticalanal- yseswereperformedusingtheSPSSsoftware(version18.0 forWindowsXP;Microsoft Corporation,Armonk,NY,USA).

Two-tailedpvalues<0.05wereconsideredtobestatistically signiﬁcant.

Results

Tenofthe30communitypharmacieslocatedinthemunici- palitiesattendedbytheIntegratedHealthcareOrganisation

participatedintheprojectandofferedtheMDDStoatotal of 240 patients. CPs identiﬁed 259 discrepancies in 143 patients,leadingto228medicationreconciliationsfor126 patients byGPsandother medicalspecialists.Themajor- ity(72.3%)ofparticipants werewomenandthemeanage was 72.3±13.1 years. The mean number of prescribed medicinestakewas9.1±3.8perpatientandthemeannum- berofmedicationinterventionswas1.8±1.3perpatient.

ThemaintypeofdiscrepancyregisteredbyCPswasthat patients were not taking medicines listed on their active medicalcharts(58.7%,n=152).

Inmorethanhalf(54.8%,n=125)ofdiscrepancycases, GPsdecidedtowithdrawthetreatment.Inothercases,the treatment wasnot modified (24.6%, n=56), it was modified(13.6%,n=31),ornewtreatmentwasinitiated (7.0%, n=16).The groups of medicineswiththemost discrepan- ciesweredrugsforobstructiveairwaydiseases(R03;8.3%, n=19),psycholeptics(N05;8.3%,n=18),andnon-steroidal anti-inflammatoryandantirheumaticproducts(M01A;7.5%, n=17).

After the intervention, a signiﬁcant reduction in the number of medicines in patients’ active medical charts (−0.92±1.09,p<0.0001)wasseen.CPsinvestedanaverage of 11.8±4.1minperformingeach initialpatientinterview and 13.8±5.0min drafting the report. They thus spent a meantotalof25.5±7.4minperpatientprovidingtheser- vice.Thirteencasesweretransferredtomedicalspecialists whohadprescribeddiscrepantmedicines.

The number of hospital admissions decreased (−0.17±0.68, p=0.007) after MDDS implementation comparedwithbaseline(Table1).The numberofEDvisits alsodecreased,butthisdifferencewasnotsigniﬁcant.

Economicoutcomes

The mean cost of the intervention wasD71.5±15.8. GP consultationswerethecostliestcomponents(D55each)fol- lowedbythetelephonespecialistconsultation(50D each);

the average costs of CPand specialist consultationswere D11.3±3.3 and D5.2±15.3, respectively. The costs of medication, ED visits, andhospitaladmissions werelower after theintervention (Table2).Even taking intoaccount

Table1 Numbersofmedicines,emergencydepartmentvisits,andhospitaladmissions6monthsbeforeandaftertheresolution ofmedicationdiscrepancies(n=126).

Variable n ¯x (SD) Difference: ¯x (SD) pvalue

Numberofmedicines

Before 1149 9.12(3.82) −0.92(1.09) <0.0001

After 1033 8.20(3.81)

NumberofEDvisits

Before 77 0.61(1.13) −0.10(1.28) 0.405

After 65 0.52(0.91)

Numberofhospitaladmissions

Before 41 0.33(0.66) −0.17(0.68) 0.007

After 20 0.16(0.42)

¯x,mean;SD,standarddeviation;ED,emergencydepartment.

(5)

Table2 Meancostsperpatient(D,2017;n=126).

Item (SD) Difference:(SD) pvalue

Medication

Before 1.4(3.0) −0.77(2.5) <0.0001

After 0.6(2.1)

EDvisits

Before 92.3(171.9) −14.4(193.3) 0.007

After 77.9(138.7)

Hospitaladmissions

Before 909.7(2079.8) −501.2 (2001.9) <0.0001

After 408.4(1229.6)

Intervention

Before --- 71.5(15.8) ---

After 71.5(15.8)

Total

Before 1003.3(2165.3) −444.9(2089.8) 0.018

After 558.4(1273.0)

SD,standarddeviation;ED,emergencydepartment.

ChospAdB

ChospAdA Nmedic CEDvisitsB CEDvisitsA

Cinterv

-30 € -20 € -10 € 0 € 10 € 20 € 30 €

Sensitivity analysis

Figure 2 Results of one-way sensitivity analysis including variablescriticaltotheeconomicevaluation.CHospAdB,cost ofhospitaladmissionbefore intervention;CHospAAB,cost of hospitaladmissionafterintervention;Nmedic,numberofmedi- cations;CEDvisitsB,costofemergencydepartmentvisitsbefore intervention;CEDvisitsA,costofemergencydepartmentvisits afterintervention;Cinterv,costofintervention.

thecostoftheintervention,allcostswerelowerthereafter (p<0.05).

Forall threecost-economic variables,the intervention wascosteffectivebecausehealthoutcomeswerebetterand costswerelower.The sensitivityanalysisshowedthat the variablewiththegreatestimpactwasthenumberofhospi- taladmissions,asitwastheonlyvariablethatcouldinvert thecost.Allothervariablesanalysed slightlyincreasedor decreasedthebeneﬁtsobtainedwiththeservice(Fig.2).

Discussion

ThisstudyshowedthattheMDDSisaneffectiveandinnova- tivewaytodetectmedicationdiscrepanciesincommunity pharmaciesandtoresolve themwiththecollaborationof diversehealthprofessionals,suchasCPs,GPs,othermedical specialistsandprimarycarepharmacists.Thehighpercent- age(88%)ofresolveddiscrepanciesandthereductioninthe

numberofdrugstaken(byalmostoneperpatient)suggest asigniﬁcantimprovementinpatientsafety.

CPs identiﬁed 240 patients with medication discrepancies, of whom 143 accepted study participation. The majority of these 143 patients had single discrepancies, and the rest had discrepancies in more than one medication. Medication discrepancies can be detected at different levels. Several systematic reviews have shown that pharmacist-based interventions are effective in the community setting.^20,21 The MDDS identiﬁes and reduces discrepancies being the particularity of this study the involvement of all health agents, especially community pharmacists,inthecontrolofmedicationerrors.Ourdata suggest thatCPs areideally positionedto detect medica- tiondiscrepancies,in agreementwiththe WHO’sstrategy toincludeCPinplanstodetectME.⁶

Removingamedicationfromthemedicalchartwasthe most common intervention performed by the GP. It has beendemonstratedthataftertheMDDSintervention,each patientin this study used, onaverage, almost one fewer medication than at baseline. Polypharmacy is related to poor adherence, interactions and ME,²² and reducing this condition is included in the WHO’s third Global Patient SafetyChallenge.⁷ Thus, the MDDS could provide a strat- egyfor the reduction of polypharmacy-related problems.

Furthermore, this service represents that it could be an efﬁcient way of improving patients’ medication-related safety and a strategy to prevent and manage patients’

frailty.²³

One problem associated with medication reconciliation interventions for CPs is the difﬁculty of contacting physicians.²⁴ Several authors have stated that future ini- tiativesshouldfocusoncollaborationbetweenhealthcare professionals,andsuchcollaborationisalsoessentialwhen designing services.^25,26 Therefore, CPs and primary care pharmacistsparticipatedinthedesignoftheMDDS.Primary carepharmacistsservedasintermediariesbetweenCPsand GPs,andthisstrategywaseffective.

(6)

The numbers of hospitaladmissions andED visits were 45%and16%lower,respectively,aftertheinterventionthan at baseline. Similar reductions have been observed after clinicalpharmacists-basedinterventions.^27,28Duetotheuse of a wide range of methods to calculate the costof ME, calculationoftheworldwidehealthcareexpenditureassoci- atedwithhospitaladmissionsandEDvisitsduetosucherror isdifﬁcult.²⁹However,authorsagreethatthiscostishigh.³⁰ A study conducted in the Netherlands showed that the costof hospitaladmissiondue topreventable medication- relatedeventsincreasedtoD3171perpatient,andEDvisits accountedfor D30,896,or 5.3%of thetotal healthcosts, duringthestudyperiod.³¹OneobjectiveoftheOrganisation forEconomicCo-operationandDevelopmentin2014wasto identifygoodpracticesinmanaginghealthcarebudgets.³² Therefore,reducinghospitaladmissions andEDvisits with theMDDScouldcontributetoimprovingthesustainabilityof thehealthsystem.

Our analysis supports the hypothesis that the MDDS is adominant intervention,asit improves clinical outcomes with lower costs than usual care, regardless of the cost of the intervention itself.The sensitivity analysisshowed that only the cost related to hospital admissions could inverttheICER.The variabilityinthe costofsuchadmis- sionsisgreaterthanvariabilitiesforotherhealthoutcomes.

Some authors have stated that use of the DRG system may lead to inequities in associated costs.²⁶ To reduce this variability, the identiﬁcation of hospital admissions relatedtomedicinesandexclusionofunrelatedadmission fromanalysiscould beuseful.³³ Previous economic evalu- ationshave focused ontransitional care programmesthat included interventionstopreventMEamongsettings,and theyhaveproducedvariableresults.^34---37Recentevaluations haveshownthattheservicesprovidedbyCPstendtobecost effective.^38,39Theimplementationofprofessionalpharmacy servicesliketheMDDSmaybean efﬁcientwaytoimprove patientsafety.

The groups of medicaments with the most discrepancies in this study were drugs for obstructive airway diseases (R03), psycholeptics (N05), and non-steroidal anti-inﬂammatoryandantirheumaticproducts(M01A).Con- sideringthatmostdiscrepanciesdetectedinthisstudywere duetopatientsnottakingmedicinesincludedintheirmedi- cal charts we could state that patients’ more frequently haveadherentproblems.Patientswithmedicinesprescribed for obstructive airway diseases, psycholeptics and non- steroidalanti-inﬂammatoryandantirheumaticproductsare one of the most prevalent groups of patients to have adherentproblems.⁴⁰AlthoughCPsshouldbeawareofdis- crepanciesinalltypesofmedication,specialattentionmust be given to these medication groups when providing the MDDS.

The present study has several limitations. Firstly, it was conducted within the Bidasoa Integrated Healthcare Organisation,andarelativelysmallnumberofpatientspar- ticipated.Toincreasetheexternalvalidityofourﬁndings, thestudyshouldbereplicatedinotherregions.Sencondly, onlypatientsintheNHSareeligiblefortheMDDS,asthey are the only ones for whom CPs receive electronic pres- criptioninformation. However,theauthorsdonotbelieve that the inclusion of the entire target population would alter the results. Thirdly, the present study included no

random assignment or control group, and the modiﬁca- tions observed could be attributed to factors other than the intervention. To increase the reliability of the MDDS and our ﬁnding that it is cost effective compared with usual care, the results of this study should be compared instudiesconductedwithcontrolgroups.Finally,allhospi- taladmissionsandEDvisitswereincludedinanalysis,with no evaluation of cause. To minimise possible bias, future analyses should include only hospital admissions and ED visits associatedwithME.Futurehealthpoliciesmustpro- vide support for the development and implementation of evidence-basedservicestopreventMEandimprovepatient safety.

What was known?

• The reduction of preventable harms related to medicineswillreducepatientsafetyincidents.

• Communityandprimarycarepharmacistsareuseful healthprofessionalsineffortstoreducemedication errorsinprimarycare.

What is new?

• TheMDDSmightreducethenumbersofmedications prescribed,emergency departmentvisits, andhospitaladmissions.

• The MDDS might be a cost-effective service that couldcontribute toimproving the sustainability of thehealthsystem.

• The MDDS might be effective as a collaborative approach between the community pharmacy and ambulatorysettings.

Funding

Theauthorsreceivednospeciﬁcfundingforthiswork.

Conﬂict of interest

Theauthorsdeclarenoconﬂictofinterest.

Acknowledgment

Authorswouldliketothankthepatients,communitypharmacists, in special to Ainhoa Ayerza and Heillen Mora pharmacists in Martin Ezcurras’pharmacy, generalpracti- tionersandprimarycarenurseswhosharedtheirtimefor thepurposesofthisproject.

References

1.MakaryM,DanielM.Medicalerror-thethirdleadingcauseof deathintheUS.BMJ.2016;353:2139.

(7)

2.World Health Organization (WHO). WHO launches global effort to halve medication-related errors in 5 years.

www.who.int/mediacentre.

3.EuropeanMedicinesAgency.Tacklingmedicationerrors:Euro- pean Medicines Agency workshop calls for coordinated EU approach Proposals to improve reporting and prevention of medicationerrorsaremade.2013;44:8---9.

4.WorldHealthOrganization(WHO).Addressingtheglobalchal- lengeofmedicationsafetytoimprovepatientsafetyandquality ofcare.Sixty-ninthWorldHealthAssemblySideEvent.2016.

5.National Coordinating Council for Medication Error Reporting and Prevention. What is a medication error? New York, NY: National Coordinating Council for Medication Error Reporting and Prevention; 2015.

http://www.nccmerp.org/about-medication-errors,acc.

6.Medication Errors: Technical series on safer primary care.

Geneva: World Health Organization; 2016. Available from:

http://www.nccmerp.org/about-medication-errors.

7.Patient safety. WHO global patient safety challenge:

medication without harm. Geneva: World Health Orga- nization; 2017. Available from: http://www.who.int/

patientsafety/medication-safety/en/.

8.SarfatiL,Ranchon F,VantardN,SchwiertzV,LarbreV,Parat S, et al. Human-simulation-based learning to prevent medication error: a systematic review. J Eval Clin Pr. 2018, http://dx.doi.org/10.1111/jep.12883.

9.NiY,LingrenT,Hall ES,Leonard M,MeltonK,KirkendallES.

Designingandevaluating anautomatedsystem for real-time medicationadministrationerrordetectioninaneonatalinten- sivecareunit.JAmMedInfAssoc.2018:1---9.Availablefrom:

http://academic.oup.com/jamia/advance-article/doi/10.1093 /jamia/ocx156/4797402

10.DigiantonioN,LundJ,BastowS.Impactofapharmacy-ledmed- icationreconciliationprogram.PT.2018;43:105---10.

11.Smith S, Mango M. Pharmacy-based medication reconcilia- tionprogramutilizingpharmacistsandtechnicians:aprocess improvementinitiative.HospPharm.2013;48:112---9.

12.Kraus SK, Sen S, Murphy M, Pontiggia L. Impact of a phar- macytechnician-centered medication reconciliationprogram onmedicationdiscrepanciesandimplementationofrecommen- dations.PharmPract(Granada).2017;15:2---5.

13.Salameh L, Abu Farha R, Basheti I. Identiﬁcation of medication discrepancies during hospital admission in Jordan: prevalence and risk factors. Saudi Pharm J.

2017;26:125---32. King Saud University. Available from:

https://doi.org/10.1016/j.jsps.2017.10.002

14.Rotta I, Salgado TM, Silva ML, Correr CJ, Fernandez- Llimos F. Effectiveness of clinical pharmacy services:

an overview of systematic reviews (2000---2010).

Int J Clin Pharm. 2015;37:687---97. Available from:

http://www.ncbi.nlm.nih.gov/pubmed/26001356 [cited 9.05.16].

15.KellerM,KellingS,CorneliusD,OniH,BrightD.Enhancingprac- ticeefﬁciencyand patientcare bysharing electronichealth records.PerspectHealInfManag.2015;12:1---6.

16.Osakidetza.Tarifaspara facturaciónde serviciossanitarios y docentesdeOsakidetzaparaela˜no2017.2017.

19.LeisterJ,StausbergJ. Comparisonof costaccounting meth- odsfromdifferentDRGsystemsandtheireffectonhealthcare quality.HealthPolicy(NewYork).2005;74:46---55.

20.KwanJ,LoL,Sampson M,ShojaniaK.Medication reconcilia- tionduringtransitionsofcareasapatientsafetystrategy:a systematicreview.AnnInternMed.2013;158:397---403.

21.Mueller S, Sponsler K, Kripalani S, Schnippe r J. Hospital- based medication reconciliation practices: a systematic review. Arch Intern Med. 2012;172:1057---69. Available from:

http://archpsyc.jamanetwork.com/article.aspx?articleid=

1203516

22.FarrellB,ShamjiS,MonahanA,MerkleyVF.Reducingpolyphar- macyintheelderly:casestohelpyou‘‘rocktheboat.’’.Can PharmJ.2013;146:243---4.

23.Gutierrez-ValenciaM,IzquierdoM,CesariM,Casas-HerreroA, InzitariM,Martínez-VelillaN.TherelationshipbetweenFrailty andPolypharmacyinolderpeople:aSystematicReview.BrJ ClinPharmacol.2018.

24.PevnickJM,ShaneR,SchnipperJL.Theproblemwithmedica- tionreconciliation.BMJQualSaf.2016;25:726---30.

25.WeissenbornM,Haefeli W,Peters-Klimm F,SeidlingH.Inter- professionalcommunicationbetween communitypharmacists andgeneralpractitioners:aqualitativestudy.IntJClinPharm.

2017;39:495---506.

26.LöfﬂerC,KoudmaniC,BöhmerF,PaschkaSD,HöckJ,Drewelow E,etal.Perceptionsofinterprofessionalcollaborationofgen- eralpractitionersandcommunitypharmacists---aqualitative study.BMCHealthServRes.2017;17:1---7.

27.KoehlerBE,RichterKM,YoungbloodL,CohenBA,PrenglerID, ChengD,etal.Reductionof30-daypostdischargehospitalread- missionoremergencydepartment(ED)visitratesinhigh-risk elderly medicalpatientsthrough deliveryof atargeted care bundle.JHospMed.2009;4:211---8.

28.Gillespie U, AlassaadA, Henrohn D,Garmo H, Hammarlund- Udenaes M, Toss H, et al. A comprehensive pharmacist intervention to reduce morbidity in patients 80 years or older: a randomized controlled trial. Arch Intern Med.

2009;169:894---900.

29.PatelK,JayR,ShahzadM,GreenW,PatelR.Asystematicreview ofapproachesforcalculatingthecostofmedicationerrors.Eur JHospPharm.2016;23.

30.Frontier Economics. Exploring the costs of unsafe care in the NHS [Internet]. 2014. Available from:

http://www.frontier-economics.com/documents/2014/10/

exploring-the-costs-of-unsafe-care-in-the-nhs-frontier-report-2 -2-2-2.pdf.

31.MagdelijnsF,Stassen P,StehouwerC,PijpersE.Directhealth carecostsofhospitaladmissionsduetoadverseeventsinThe Netherlands.EurJPublicHeal.2014;24:1028---33.

32.LiaropoulosL, GoranitisI.Healthcareﬁnancingand thesus- tainability of health systems. Int J Equity Health. 2015;14:

5---8.

33.Malet-Larrea A, Goyenechea E, García-Cárdenas V, Calvo B, Arteche JM, Aranegui P, et al. The impact of a medication reviewwith follow-upservice onhospital admissions in aged polypharmacy patients. Br J Clin Pharmacol. 2016;94:

831---8.

34.Karapinar-C¸arkıt F, van der Knaap R, Bouhannouch F, Borg- steedeSD,JanssenMJA,SiegertCEH,etal.Cost-effectiveness of a transitional pharmaceutical care program for patients discharged from thehospital. PLOS ONE. 2017;12:e0174513.

Available from: http://dx.plos.org/10.1371/journal.pone.

0174513

35.Karnon J, Campbell F, Czoski-Murray C. Model-based cost- effectiveness analysis of interventions aimed at preventing medicationerrorathospitaladmission(medicinesreconcilia- tion).JEvalClinPr.2009;15:299---306.

36.ChinthammitC,ArmstrongE,WarholakT.Acost-effectiveness evaluationofhospitaldischargecounselingbypharmacists.J PharmPr.2012;25:201---8.

37.Simoens S, Spinewine A, FoulonV, Paulus D. Reviewof the cost-effectiveness ofinterventionsto improve seamless care focusingonmedication.IntJClinPharm.2011;33:909---17.

(8)

38.Perraudin C,BugnonO,Pelletier-Fleury N.Expandingprofes- sionalpharmacyservicesinEuropeancommunitysetting:Isit cost-effective?Asystematicreviewforhealthpolicyconsider- ations.HealthPolicy(NewYork).2016;120:1350---62.

39.Malet-Larrea A, García-Cárdenas V, Sáez-Benito L, Ben- rimoj S, Calvo B, Goyenechea E. Cost-effectiveness of professional pharmacy services in community pharmacy:

a systematic review. Expert Rev Pharmacoecon Outcomes Res. 2016;16:747---58. Taylor & Francis. Available from:

https://doi.org/10.1080/14737167.2016.1259071

40.PottegårdA,ChristensenRD,HoujiA,ChristiansenCB,Paulsen MS, Thomsen JL, et al. Primary non-adherence in general practice: a Danish register study. Eur J Clin Pharmacol.

2014;70:757---63.