ORI GI NALRESEARCHARTI CLEConsensusValidationoftheFORTA(FitfORTheAged)List:AClinicalToolforIncreasingtheAppropriatenessofPharmacotherapyintheElderlyAlexandraM.Kuhn-ThielChristelWeiMartinWehlingTheFORTAauthors/expertpanelmembersPublishedonline:19December2013TheAuthor(s)2014.ThisarticleispublishedwithopenaccessatSpringerlink.comAbstractBackground Multimorbidityandpolypharmacyrepresenta major problemfor elderly patients; improvement ofmedication schemes is important and listing approaches(e.g. Beerslist)areconsideredtobepotentiallyuseful.Objectives Theaimofthisstudywastoperformexpertconsensusvalidationof theFORTA(Fit fORTheAged)List,adrugclassicationcombiningpositiveandnegativelabelling of drugs chronically prescribed to elderly patients.Methods Atwo-roundDelphi procedurewas conductedinvolving20experts,17geriatricinternistsand3geriatricpsychiatrists fromGermany and Austria, evaluating thelabels assigned to 190 substances or substance groups.These labels ranged from A (indispensable), B (benecial),C(questionable) toD(avoid), dependingonthestateofevidencefor safety, efcacyandoverall age-appropriate-ness. The experts were also requested to suggest additionalsubstances and indication areas for assessment and possibleinclusion in the FORTA List. Aweighted (corrected)consensuscoefcientwasgeneratedforeachsubstancetoreect (1) agreement withtheoriginal label, and(2) dis-tributionamongraterslabels.Results The overall consensus for all items and raters was92%(corrected). For54/190 items (28.4%), a unanimousresponse was achieved as to the original author-basedFORTA label choice. Twenty-four substances (12.6%) fellshort of theconsensus cutoff andwerere-evaluatedinasecond round. This yielded conrmation of 171/190, or90%, oftheoriginalauthor-basedFORTAlabels. Atotalof35newsubstanceswerealsoacceptedfortheFORTAList. Drugs used for dementia and dementia syndromesprovokedparticularresponseheterogeneity.Conclusion TheFORTAList nowreectsawider con-sensus amongexperts, increasingits validityfor clinicaluse. It represents a tool toimprove the qualityof drugprescriptioninolderpatientsbyidentifyingbothinappro-priate and omitted drugs, and thus overtreatment andundertreatment. The validation of FORTAs impact onclinical endpoints has yielded promising preliminaryresults, tobecorroboratedinongoinglargertrials.1 IntroductionThe increase in life expectancy is triggering dramaticdemographicchanges inindustrializedcountries. Figuresfor 2010providedbytheUSCensusBureau[1] indicatethat thepopulationaged65yearsandolderhasincreasedover the past decade from35.0millionin2000to40.3millionin2010, representing13.0%of thetotal popula-tion. InGermany, thispercentageisat20%for2009[2].Elderlymultimorbidpatientsaremorelikelytoreceivemultiple drugtreatments (polypharmacy) comparedwithyounger patients. It has beendemonstrated that patientsTheFORTAauthors/expertpanelmembersarelistedintheAppendix.Electronicsupplementarymaterial Theonlineversionofthisarticle(doi:10.1007/s40266-013-0146-0)containssupplementarymaterial, whichisavailabletoauthorizedusers.A. M.Kuhn-Thiel M.Wehling(&)InstituteforExperimentalandClinicalPharmacologyandToxicology, DepartmentofClinicalPharmacology, CenterforGeriatric Pharmacology, Medical Faculty Mannheim, UniversityofHeidelberg, Maybachstr. 14, 68169Mannheim, Germanye-mail:martin.wehling@medma.uni-heidelberg.deC. WeiDepartmentofMedicalStatistics, BiomathematicsandInformationProcessing,MedicalFacultyMannheim,UniversityofHeidelberg, Mannheim, GermanyDrugsAging(2014)31:131140DOI10.1007/s40266-013-0146-0aged65yearsandoldertakeveormoredrugsin44%(male)and57%(female)ofcases,andtenormoredrugsin12%ofcases[3].Eachmedicalguidelinerecommendsan average of three medications. According to gures from1998, persons over 80years of age have anaverage ofthree diagnoses; this means 393=9 medications inelderlypatients, whichisalsoreectedinreallife[4].Multimorbidity and polypharmacy often harbourunpredictable dangers due to age-related alterations inpharmacokinetics and pharmacodynamics [5], adverse drugreactions[6], whichmaytriggertheso-calledprescribingcascade [7], drugdrug or drugdisease interactions,problemswithdosages, medicationerrors, andevendeath[8]. This translates, for example, intoapproximately2.1million side effect-related hospital admissions and 100,000deathsperyearintheUS,outofatotalpopulationof265millioninhabitants, andcostsofapproximately1.54bil-liondollarsperyear[9]. Theriskofpotential druginter-actions necessarily increases with the number of drugsprescribed[10].Geriatric medicine is a rapidly growing discipline in theWestern world [11, 12]. The paucity of evidence-basedguidelinesandclinical studiesfor theelderlyisalarmingandcontributestothechallengesofrationalisticdrugpre-scribinginelderlypatients. Manyelderlypersonsdisplaylimitations in their physical and mental capacities [13],nearlyalways precludingtheir inclusioninclinical drugtrials. Thesometimesrigidadherencetoavailableguide-lines presents another signicant problemas there arevirtuallynoevidence-basedguidelinesforthisveryheter-ogeneousgroup;theunderlyingassumptiononeguidelinets all simply does not work in this age bracket [14].These limitations necessitate the development of criteria orconcepts for safer and more efcient drug use in theelderly,ideallyamalgamatedataninternationallevel.Inresponse tothese challenges, manycountries havebegun to develop strategies for the safer prescribing ofmedicationsinelderlypatients[15]. Aprecedent wassetbytheBeersCriteria[16]whichclassifypotentiallyinap-propriate medications (PIMs) according to threecategoriesdescribingthedegreeofinappropriateness;thislistinghasundergone several updates, recently winning the support oftheAmericanGeriatricSociety[17]. TheGermancoun-terpart, thePRISCUS(PIM) List, waspublishedin2010[18]. Suchnegativelistshaveproventobequitepracti-cable, butstilllackconrmationastoeffectivenessattheclinicalendpointlevel[19].Gallagheretal.introducedtheSTOPP(ScreeningToolof Older Persons Prescriptions)/START(ScreeningTooltoAlert DoctorstotheRight Treatment)Criteriain2008[20]. TheSTOPPcriteriaallowthedetectionofpotentialovertreatment and place a special emphasis on drugdiseaseinteractions;theSTARTcriteriaservetoassistthephysicianintargetingpotentialerrorsofomissionbypin-pointing treatments that may be indicated but not pre-scribed[21]. It has been suggested in the literaturethattheSTOPP criteria may have a higher sensitivity than theBeerscriteriafordetectingPIMs[22].TheFORTA(Fit fORTheAged) classicationsystemwasproposedin2008[23, 24]asatool foraidingphysi-cians fromparticipating countries (initially Germany) inscreeningforunnecessary, inappropriateorharmful medi-cations and drug omissions in older patients in an everydayclinicalsetting.Itistherstclassicationsysteminwhichbothnegative andpositivelabellingarecombinedat thelevelofindividualdrugsordruggroups. Asitaimsattheindividual indications (implicit listing requiring patientcharacteristics/diagnoses) it is thus clearlydifferent fromnegativelistswhichfocusonmajorproblemsindrugpre-scribing(errorsofcommissionoromission,orriskymedi-cationswithinallfrequentlyuseddrugclasses)thatshouldrather be avoided when prescribing to geriatric patientsbecauseofage-relatedchanges(explicit listslargelyinde-pendent of individual patient characteristics). FORTAisevidence-based and real-life oriented. Factors such asadherence issues, age-dependent tolerance and frequency ofrelativecontraindicationsaregivendueconsiderationsincestrict andcitableevidence, astypicallyderivedfromran-domized clinical trials, is still rare for this population,althoughimportant ifavailable. Amedicationcanreceivedifferent FORTAlabels for different indications (indica-tion-dependent). FORTAdoes not taketheplaceof indi-vidual therapeutic considerations or decisions.Contraindications always take precedence over the FORTA-classication.Thesystemdoesallowforexceptions.TheFORTAclassesaredenedasfollows: Class A(A-bsolutely)=indispensabledrug, clear-cutbenet in terms of efcacy/safety ratio proven inelderlypatientsforagivenindication ClassB(B-enecial)=drugswithprovenorobviousefcacyintheelderly, but limitedextent of effect orsafetyconcerns ClassC(C-areful)=drugswithquestionableefcacy/safety proles in the elderly, to be avoided or omitted inthe presence of too many drugs, lack of benets oremergingsideeffects;review/ndalternatives Class D(D-ont)=avoid in the elderly, omit rst,review/ndalternativesTheFORTAList isacompilationof 190medications(primarilylong-termtreatment, exceptionsarenoted)[25]mostfrequentlyprescribedinolderpatients, alignedto20main indication groups. Each substance or group is assigneda FORTA class A, B, C or D. In cases in which homogeneity132 A.M.Kuhn-Thieletal.wasconsideredtobehighandoflessersignicancewithrespect tootheraspectsofthegroupofdrugsconcerningage-related issues, similar drugs were grouped and assessedassuch(angiotensin-convertingenzyme[ACE] inhibitorsas an example). If individual compounds were considered tobe heterogenous, rating was performed for individual drugs(acetylsalicylicacid, clopidogrelalthoughbotharecon-sideredasplatelet inhibitors). ThisoriginalversionoftheFORTA list is part of reference [25] and was created by thebookauthors, includingtheoriginalauthorofthemethod.Theratingwasopinion-basedbyanintegrativeapproachwhichcomprisedavailablestudyevidenceascitedinthebook. It thus was also evidence-based where such evidencewasavailable.Althoughexact-usedataareasyetmissing,theoriginalFORTAlistisincreasinglyrecognizedinGer-many, assuggestedbythefact that the3rdeditionoftheseminal book(2013) hadtobepublishedwithin3yearsafter the rst (2010); for this survey, the 2nd edition (2011)wasrelevant[25].Our aims, within the context of a two-round DelphiConsensus procedure, includedthe rater-basedconrma-tion/determinationof labelsfor 190items intheoriginalauthor-basedFORTAList, andidenticationandlabelingof new indications/substances. As a consensus-basedapproachthisprocessreectstheratingofmanyexperts,drawingonbothavailableevidenceandpersonal experi-ence/opinion.TheconsensusvalidationoftheFORTAListrepresentsthe rst phase of a two-part development programmefunded by the German Research Foundation; a clinicalstudy is running to test the impact of the FORTA system onthequalityof pharmacotherapyandclinical endpoints in400patientsfromtwoGermangeriatricclinicsbyimple-mentingtheFORTAList.2 MethodsTheconsensusvalidationprocedureincluded1. Reviewof the available literature and examples ofpracticalapplicationsoftheDelphimethod.2. RecruitmentofexpertsintheGerman-speakingcoun-tries(Germany,AustriaandSwitzerland)representinggeriatricinternists/geriatricians andgeriatricpsychia-trists withextensive clinical experience inthe phar-macotherapy of (multimorbid) elderly patients; highacademic status; prominent standing in the leadinggeriatric/psychiatric medical associations; and number,quality and relevance of experts publications. Theselectionwas basedonavailableinformationontheInternet inaniterative, semi-quantitative process bytwooftheauthors(AKTandMW).3. Round 1: The FORTAList was adapted fromitsoriginal publicationform[25] toaquestionnaireandsent totheexpertsviae-mail (seeoriginal surveyasElectronicSupplementaryMaterial [ESM] 2). Partic-ipants were requested to reviewthe instructions onhowtoapplytheFORTAprinciple;studytheauthor-basedlabels (AD) for eachitemandprovide theirown FORTA labels (or abstention) whenever indisagreement; and make comments. In a separatesection, the experts were requested to suggest newsubstances/indications with labels, to augment theFORTAList. All experts had been exposed to therelatedbook[25] asthecommon, althoughcertainlynot exclusive, base of evidence compiled fromtheliterature.4. Statistical analysis based on Round 1 input wasperformed as follows: the Likert scale is oftenfavouredforconsensusprocedures, aswell asmeans,medianandmode(central tendencyindicators)[26,27]. For evaluatingtheFORTAlabellingsystem, weadaptedtheLikert scalewiththeaimof devisinganalgorithm combining collective/central tendencyregardingtheoriginal labelswithimpact of distribu-tion/dispersionof raters labels. Toachievethis, thepercentage of raters labels (excluding abstentions)agreeing with the original author-based labels wascalculated, bothoverall andforeachitemseparately.The resulting percentages were then weighted togenerateacorrectedconsensuscoefcient(cons_corr,denition see ESM 1, FORTA list, p. 41) for each itemreecting the degrees of deviation between the expertsindividual FORTAratings. Although at rst glanceseemingly arbitrary, those weighing factors appear, forour purposes, plausible [2830]. This ultimately allowstheactual assignment of FORTAclass values tothesubstancesinquestion.Theweightingsystem,reect-ingdegrees of deviation, was expressedinterms ofrangeclass(Table1), denedas: Range0: unanimity among all experts giving aFORTArating(nodeviation); Range1:greatestrangeonlyfromAtoB, BtoCorCtoD(neighbouringclasses), halfweight;Table1 FrequencyofsubstancesindenedrangegroupsaccordingtodegreeofconsensusRange Frequency(n=190) Percent0 54 28.421 86 45.262 43 22.633 7 3.68ConsensusValidationoftheFORTAList 133 Range2:greatestdistancefromAtoCorBtoD,two-thirdsweight; Range3: greatest distance from A to D, full weight.Inordertoconrmtheoriginal/determinenew, rater-based labels, we converted the experts FORTAratings intonumericalvaluesrepresentingthemedian:A ? 1,B ? 2,C ? 3andD ? 4,respectively.Themean and mode were calculated for each item,reconvertedtoFORTAlabelsandcomparedwiththeoriginalauthor-basedlabels. Therangeforeachlabelwasdenedas:If1Bm\1.5 ? FORTAClassAIf1.5Bm\2.5 ? FORTAClassBIf2.5Bm\3.5 ? FORTAClassCIfmC3.5 ? FORTAClassADwhere m=arithmetic mean based on the ratersgrades14.Thescalehasnot beendenedinorder toallowforcomplexstatistical calculations. The purpose of thisscale was to pool the judgements by taking intoaccount eachrater judgement. ThiswasnecessarytoenableacomparisontobemadebetweentheratersopinionsandtheFORTAclassications. Theassign-mentofA ? 1,B ? 2,C ? 3andD ? 4seemstobeplausiblewhenassumingthat anydifference(12,23, 34)isequallyimportant[31].5. Round 2: Substances falling short of the presetcorrectedconsensus cutoff of 0.800 were re-sent totheexperts. Newsubstances suggestedbyC2ratersandall newindications weresent totheexperts forevaluationintheformofaquestionnaire.6. Analysis basedonRound2input was performedasfollows: conrmationofFORTAlabels/determinationof newrater-based FORTAlabels for re-evaluateditemsderivedfromthearithmeticmean(asdescribedabove); reviewof all comments. Asimple agreedisagree approachor 5-point scale was not chosen;quantication of disagreement seemed necessary astherearefourcategoriesofanswers(FORTAclasses),and a full match is unlikely to occur; thus, this approachmay possibly have led to an equally large second round.In this way, the actual FORTA classes could bepreservedassuchandeitherconrmedorchallenged.7. Compilationof anannotatedFORTAList basedonexpertsinputovertworounds.The substances and indications suggested by the expertswereselectedasfollows:1. Acceptanceof all substances/indicationsreceivinganafrmative response by[50%of experts duringRound 2 and receiving a FORTArating (excludingabstentions)byC8raters.2. Calculationof akappaindexreectinglabel disper-sion: here, kappa is dened as the (proportion ofmatching labels-0.25)/0.75. This gives due con-sideration to the fact that a gure of 25%cantheoretically be attained by chance alone, with thechoiceoffourdistinctlabels.3. Conversion to median and calculation of mean andmode, asintherst procedure. ThearithmeticmeanprovidedthebasisforconversiontoFORTAlabels.4. Compilationofallsubstancesinaseparate,annotatedlist.3 ResultsTwentyexperts, 17geriatricinternistsandthreegeriatricpsychiatrists representing Germany (13) and Austria (7)agreed to participate in the survey. The return rate for bothroundswas100%.The overall consensus for all items andexperts afterRound1was foundtobe92%, corrected(mean0.922,median 0.950, range 0.5001.000). Overall, 54/190(28.4%) of the evaluated items elicited unanimousagreement among the raters (Table1) and 24/190 (12.6%)items fell short of the cutoff of 0.800 and were re-evaluatedin a second round (Table2). Of these 24 items, 19 (79.2%)representedsubstancescommonlyusedforthepreventionor treatment of dementia anddementia syndromes; 3/24substances (12.5%) represented drugs used for treatingcardiovascular diseases; 1/24 (4.2%) was a drug pre-scribedfordepressionand1/24(4.2%)forosteoporosis.Backed by experts largely convergent comments on theindividual substances, twoconsistent trendscouldthusbedetected. Theseindicatedashift intheFORTAlabelsfordrugs used to prevent or treat dementia and dementiasyndromes:1. Agreement with the original Clabel for substancesadministeredfordementiawasobservedinmostcasesfortheparticipatinggeriatricpsychiatrists,asopposedto geriatric internists, who tended to favour the D labelinthesecases. InfutureFORTAdevelopments, thisarea should be rated by a larger group of geriatricpsychiatriststoemphasizetheir particular experienceinthisarea.2. Theoriginal Dlabel was challengedspecicallyforneuroleptic drugs bybothgeriatric psychiatrists andgeriatric internists; many experts tended towards C andexpressedthewishfor further differentiation/qualify-ing statements pertaining to the therapy of behaviouralandpsychologicalsymptomsofdementia(BPSD).134 A.M.Kuhn-Thieletal.Table2 Analysisofthe24re-evaluatedsubstancesRe-evaluatedsubstance/group(originalFORTArating)FORTAindicationareaNo. ofraters(n=20)ConsensuscoefcientExpertratingonanumericalscaleaProposedFORTArating, basedonmeanvaluefromRound2Round1 Round 1 (cutoff0.800)Round1Round2 Round2mean;modeDrugstopreventortreatdementiaordementia-relatedsyndromesNimodipine(C) Dementia 20 0.750 3.5;3 D19 3.7;4Antioxidants(C) Dementia 19 0.710 3.6;4 D20 3.9;4Phytotherapeuticagents, e.g.Ginseng(C)Dementia 20 0.725 3.6;4 D20 3.8;4Hormonepreparations, e.g. DHEA(C)Dementia 20 0.700 3.6;4 D20 3.9;4Gingkobiloba(C) Dementia 20 0.775 3.5;3 D20 3.6;4Ergolinederivatives(C) Dementia 19 0.763 3.5;3 D20 3.8;4Piracetam(C) Dementia 20 0.800 3.4;3 D20 3.6;4Pyritinol(C) Dementia 18 0.778 3.4;3 D19 3.7;4Selegiline(C) Dementia 19 0.763 3.5;3 D20 3.7;4Haloperidol(D) BPSDparanoia,hallucinations19 0.632 3.3;4 C20 3.0;3Risperidone(D) BPSDparanoia,hallucinations20 0.500 3.0;2 C20 2.7;2Quetiapine(D) BPSDparanoia,hallucinations20 0.575 3.2;4 C20 2.9;3Aripiprazole(D)[215mg/day] BPSDparanoia,hallucinations19 0.789 3.6;4 C17 3.4;4Clozapine(D)[1050mg/day] BPSDparanoia,hallucinations20 0.800 3.6;4 D19 3.7;4Risperidone(D) BPSDrestlessness 20 0.625 3.3;4 C20 2.7;2Melperone(D) BPSDrestlessness 20 0.675 3.4;4 C20 3.4;4Quetiapine(D)[25200mg/day] BPSDrestlessness 19 0.763 3.5;4 C19 3.3;3Pipamperone(D)[20150mg/day] BPSDrestlessness 19 0.790 3.6;4 D17 3.6;4Mirtazapine(C)[1530mg] BPSDsleepdisorders20 0.775 3.0;3 C20 3.0;3Mirtazapine(D) Insomnia 20 0.700 3.4;4 C20 3.45;4OtherdrugsDigitoxin(D) Atrialbrillation 20 0.525 3.1;4 C19 2.5;2Digoxin(B) Atrialbrillation 20 0.800 2.4;2 B19 2.4;2ConsensusValidationoftheFORTAList 135Duetoabstentions,thenumberofratersvariedforeachitem(maximum20, range520). Accordingtotheraterscomments, the most commonreason for abstaining wasinsufcient experience with a particular substance orindicationgroupand,inindividualcases, potentialconictofinterest.The indicationarea of oncological/haematological ill-nesses (27items) receivedthemost abstentionsandthusthe lowest number of raters (mean 7.41, median 7.22, range512), yetthecalculatedcorrectedconsensusvalueswereconsistentlyhigh(mean0.964, range0.8571.000)forallitems tested. Most experts gave the reason for abstention asinsufcientexperienceorlackoffamiliaritywiththecur-rent state of evidence for oncological treatments. One raterdocumented a consultation with other experts in the eld ofoncology. Re-evaluationwasforegone, but thisindicationgroup will be kept under close scrutiny during furtherclinicaldevelopment.According to the calculations based on a numerical scalefor purposes of comparing the rater-based labels to theoriginal author-based labels, it was found that, after Round1, 12of the24re-evaluateditems(6.3%of theoriginal190 substances) hadreceived a FORTAlabel divergingfromtheoriginallabel. These12itemscorrelateddirectlywith the substances receiving the lowest corrected con-sensus coefcients. After Round 2, this number hadincreasedto19ofthe24retestedsubstances(5/24labelsconrmed), elicitingnalconrmationof90%(171/190)of the original labels. This increase in label deviationappears toindicatethat other factors mayhaveplayedaprominent role in the decision-making process for labellingduringRound2.A total of 35 new substances were accepted for potentialincorporationintotheFORTAList. Nineteensubstanceswereincludedinfournewindicationareas:epilepsy(12),anaemia (4), gastrointestinal illnesses/concomitant appli-cationofnon-steroidalanti-inammatorydrugs[NSAIDs](2) andbipolar disorder (1). Thus, the original opinion-basedproposal of indicationareas bytheauthors of ref-erence [25] was largely conrmed by the rater panel.Sixteen substances associated with pre-existing FORTAindications were included: drugs for the therapy ofdepression (6), chronic pain (3), atrial brillation (2),arterial hypertension(2), coronaryheart disease(1), oste-oporosis(1)andinsomnia(1).These results are summarized in the FORTA List,availableonlineasESM1(fullstatisticaldetailsincludingallresultsfromtherstroundareavailableuponrequest).4 DiscussionTheDelphimethodoftenpresentsachallengetocarryoutinpractice, not least due tothelackof evidence intheliteratureastooptimal standardoperationproceduresandforms of interpretation. Nevertheless, it has become anacceptablemode, andsometimestheonlyfeasibleoption,of obtaining experts opinions on particularly complextopics[27, 32,33].Medication lists and classication systems developedduringthe past fewdecades represent a variationonanestablishedtheme. IntheComprehensiveDrugAbuseandControlAct of1970, forexample, harmfulorhabit-form-ingdrugswererankedaccordingtotheirdangerousness;selected drugs were assigned to specic categories withappropriaterestrictions [34]. Our proposedFORTAsys-teminvolvestheevidence-basedclassicationofmedica-tions according to age-appropriateness. Through the expertTable2 continuedRe-evaluatedsubstance/group(originalFORTArating)FORTAindicationareaNo. ofraters(n=20)ConsensuscoefcientExpertratingonanumericalscaleaProposedFORTArating, basedonmeanvaluefromRound2Round1 Round 1 (cutoff0.800)Round1Round2 Round2mean;modeDronedarone(B) Atrialbrillation 18 0.556 2.9;3 C18 3.0;3Strontiumranelate(B) Osteoporosis 17 0.794 2.1;2 B18 2.1;2FORTAFitfORTheAged, DHEAdehydroepiandrosterone, BPSDbehaviouralandpsychologicalsymptomsofdementiaaA !1; B !2C !3; D !4If 1 m\1:5 !AIf 1:5 m\2:5 !BIf 2:5 m\3:5 !CIf m3:5 !D136 A.M.Kuhn-Thieletal.validation procedure, the FORTAList, a drug-appropri-atenessratingsystem,hasbeenendorsedandimprovedbythe input of 20experts, therebyenhancingits value forimplementationinaclinical setting, whileareasrequiringfurtherattentionanddevelopmentclearlycameintoview.The panel was chosen fromclinical specialties only asclinical experiences in the elderly appeared to be mostvaluable inthis patient population, whichaffecteddeci-sions made as to the choice of inclusion of other specialties(e.g. pharmacoepidemiology, pharmacists).Thestill relativelytenuousorinconsistent stateofevi-dence associated with medications for dementia is alsoreected in examples from the literature [3537]. This areawouldappear torequirefurther observationanddevelop-ment duringclinical studies. Futureclinical projects willalsospecicallyhavetotargettheproblemofhowbesttoclassifydementiaintheFORTAList. Manyexperts dis-cussedpossiblebenetsofclassicationaccordingtoeti-ology(i.e. Alzheimersvs. vascular origin). TheFORTAList divides dementia intosubclasses accordingtoaddi-tional syndromes (BPSD); drugtherapyis either of pre-ventive nature or symptom-oriented and has beensimpliedheretothegreatest extent possible. Morespe-cic differentiation during further clinical studies mayimprove the overall quality and practicability of thesystem.The rst positive indications of FORTAs potentialusefulness in everyday clinical routine are apparent inresults obtainedfromapilot studyapplyingtheFORTAprinciple[38]tothedrugtherapyof46patientsinageri-atricclinicinEssen, Germany. It couldbedemonstratedthat the number of Class A and B medications signicantlyincreased, andthenumberofClassCandDmedicationswerereduced. Preliminarydataobtainedfromaprospec-tive, single blinded, randomized trial involving 97 patients,also conducted in Essen, further revealed that use ofFORTAmaybeassociatedwithareductioninin-hospitalfalls[39].Further-reaching applications of the FORTA system mayinclude rening the process of dening and assigningFORTA labels (classes AD) to newly selected and already-establisheddrugsassessedbyHealthTechnologyAssess-ment institutions [for example, theNational InstituteforHealth and Care Excellence (NICE) in England, or theInstitutfurQualitatundWirtschaftlichkeitimGesundheit-swesen (IQWiG, Institute for Quality and EconomicEffectiveness in Health Care in Germany)]. Although as yetlackinginmostmajorsituations, futureinputprovidedbycontrolled and, ultimately, real-life studies representequally essential components of the development procedurefor this drug labelling system, since adherence factors,availability and application issues also play important rolesin determining the ultimate effectiveness and safety of thesesubstances[40]. Another futuretaskisthefurther differ-entiationandseparationof distinct compounds that nowresideinamixedpot, e.g. thegroupoffrequency-low-ering b-blockers(FORTAlist,p.11),whichformallystillcontainssotalol but whichisnowconsideredmainlyasaclass III antiarrhythmic (Ddrug), or propranolol, whichshouldnot beused(withexceptions)forpharmacokineticreasons. For treatment of heart failure with b-blockers,positively labelledcompoundsare listed (p.10).Similarly,not all dihydropyridines are well studied in the elderly, andamlodipine is given as a lead example in the FORTA list (p.9), reectingthe results of the AvoidingCardiovascularEvents inCombination TherapyinPatients LivingwithSystolicHypertension(ACCOMPLISH) andAnglo-Scan-dinavianCardiacOutcomesTrial(ASCOT)trials[41,42].Norare diureticsfurtherdifferentiated,althoughlong-termdataaremainlyavailablefor thiazides, e.g. inACCOM-PLISH. Thechoicebetweenloopandthiazidediureticsisguided by renal function and/or severity of heart failure, butdoes not lead to different assessments as yet, mainly due tothesimilarityinsideeffects(e.g.electrolytedisorders).If compared with START/STOPPcriteria, drugs rec-ommendedbySTARTseemtobelongpredominantlytoFORTAclassesAorB. Examplesincludestatinsorace-tylsalicylicacidinthetreatmentofcoronaryheartdisease,ACEinhibitorsforheartfailure, levodopaforParkinsonsdisease, or corticosteroids and inhaled b2-agonists inchronic obstructive pulmonarydisease. Substances men-tionedintheSTOPPlist [20] appear tocorrespondwithcategories C and D as assigned by the FORTA system (seealsorevisedFORTAList). Here, examples include ben-zodiazepinesunanimouslyvotedacrossallsystemstobepotentially inappropriate or negatively rated, either indrugdiseaseconnectionorinandof themselves[17, 18,20, 25, 43]neurolepticdrugs, rst-generationantihista-minesortheophyllineasmonotherapyinchronicobstruc-tivepulmonarydisease.Not unexpectedlyfor consensus processes, afewdis-crepancies with the updated Beers list are also present.Althoughdifcult tocomparedirectly, asindividual sub-stancesaretakenintoconsiderationinrelationtospecicillnessesorconditions(drugdiseaseaspect)[17, 43], onenotable example here includes the classication or use/non-use of digoxin, which is listed to be avoided in higherdosages. Concordance between Beers and FORTA ishowever high; specic examples include the mentionofNSAIDstobeavoidedinchronicuse, doxazosintoavoidas anantihypertensive, benzodiazepines andzolpidemtoavoid in most instances, and carbamazepine to be used withcaution,correspondingtotheFORTAclassesCandDfortheseselectedsubstances.ComparedtothePRISCUSlist, most PIMshavebeenlabelled C or D, with few exceptions, most notably digoxin,ConsensusValidationoftheFORTAList 137as already mentioned above (FORTA B, PIM in PRISCUS,Beers and STOPP) [17, 18, 20, 43]. This compound and itscongener, digitoxin, areonthelist withmajor discrepan-ciesinratings(Table2); digoxinwasratedfavourablyinthetreatment of atrial brillationasrenal dosingisame-nable andintoxicationeffects are muchshorter thanfordigitoxin, still prescribed in Germany and rated FORTA C.AmajoradvancemadebydevelopmentoftheFORTAsystemhasledtothequantitativeassessmentallowingforcross-therapeutic prioritizationandreectionof multiplediseasesleadingtoreducedmedicationschemes, whereasSTART/STOPPcriteriacouldstill leadtoadditivepoly-pharmacy if multiple conditions are met. The user isintroduced to a standard, reproducible system, the repeatedemployment of whichmayencourageanoverall learningeffect(geriatricpharmacologyinanutshell).Theinternationalization of theFORTAList maybeviewedasoneofthenextimportantstepsinthedevelop-ment of FORTA. In this context it is important toacknowledge that most PIMlists and clinical tools doremaincountry-specic, bothinEurope (most Europeancountries suchas Germany, France, NorwayandAustriahave their own negative listings) and in the US(BeersList). Thisreectsthediversityof national druguseandregulatory status. It may however be noted that the originalauthorsdidnot encounteranymajorproblemsconvertingtotheUSsystem, asdocumentedintherst English-lan-guage edition of the original FORTA source containing theauthor-based, US version of the FORTA List [44].AlthoughnotdevelopedforEuropeancountries,theBeersandMcLeodlistsfromtheUSandCanada, respectively,have been successfully used to detect and compare PIMs ineight Europeancountries[45]. Thus, druglistingapproa-chesseemtobeprincipallyapplicableeventogeographi-cally removed industrialized societies. Yet, the well-knownpotential obstacles of divergenceanddifferences indrugavailability, aswellascountry-specicprescribingtrends,demography and disease epidemiology must not beignored.Thus,ensuing gaps orinconsistencies, whilenotactively presenting a hindrance in our estimation, stillrepresent anarearequiringintensiedcooperativeefforts,ideallyonaninternationallevel.5 LimitationsoftheSurveyImportant limitations of the Delphi process whicharoseand should be mentioned here include the following issues:1. Thechoiceof ratersdidnot includeawidearrayofexperts,e.g.generalpractitionersforambulatorycare,pharmacists or higher numbers of geriatric psychiatrists.2. TheFORTAList mayhavelimitedapplicabilityforinternational use and still awaits adaptations in aninternationalizationprocess.3. There is a relative lack of evidence-driven ratingscompared with consensus-driven ratings. Future modi-cationsshouldemphasizeevidence-drivenratings, par-ticularlyinemergingareas of age-relatedtherapeuticknowledge or innovations, e.g. novel oral anticoagulants.4. Duetotheemphasisonimplicitcriteria(theindividualpatient hastobeconsideredfor theapplicationof theFORTAList) theutilityof theFORTAtool maybe limitedregarding its use in pharmacoepidemiological research.5. FORTA does not specically address drugdruginteractions or contraindications which still need tobe checked individually, as well as drugdoses andmedication scheduling; it does not aimat detectingprescribingcascades,althoughanincreasedqualityofprescriptionswillcertainlyhelptoreducethem.6 ConclusionWhenappliedaccordingtospecic, well-denedcriteriawithin the context of individualized patient care andmanagement, theFORTAList shouldhelpphysicians tooptimizedrugtreatmentintheirolderpatients.Theexpertconsensus validation process for the FORTAList wasessential in its development, and it is our hope that this willultimatelyfacilitateitsuseinclinicalpractice.Acknowledgments Thisstudywassupportedbyagrant fromtheDeutsche Forschungsgemeinschaft (WE 1184/15-1), which sponsoredand approved the planning, executionand analysisas proposed in thegrant application. We are grateful toChristine Schummer for hersecretarialsupport.MartinWehlingwasresponsibleforthecreationanddevelopmentof the original FORTA concept, as well as the overall supervision andapproval of all contents. Alexandra Kuhn-Thiel coordinated andoversaw the Delphi consensus procedure and contributed to the designand organization of the revised FORTAList. All aspects of thisprocesswereperformedundercloseteammonitoring.ChristelWeiperformedall statistical analysisfor theDelphi procedure. All par-ticipatingexperts reviewedandapprovedthe manuscript andcon-tributedactivelytothecontentandpresentationoftheFORTAList.MartinWehlingwasemployedbyAstraZenecaR&D,Molndal,asDirectorofdiscoverymedicine(=translationalmedicine)from2003to 2006, while on sabbatical leave fromhis professorship at theUniversityofHeidelberg. SincereturningtothispositioninJanuary2007 he has received lecturing and consulting fees fromSano-Aventis, Novartis, Takeda, Roche, Pzer, Bristol-Myers, Daichii-Sankyo,Lilly,Novo-Nordisk,ShireandLEOPharma.ChristelWeiandAlexandraKuhn-Thielhavenoconictsofinteresttodeclare.Open Access This article is distributed under the terms of theCreativeCommonsAttributionNoncommercial Licensewhichper-mits anynoncommercial use, distribution, andreproductioninanymedium, providedtheoriginalauthor(s)andthesourcearecredited.138 A.M.Kuhn-Thieletal.AppendixOriginal FORTA List Authors[25]and Their AfliationsMartin Wehling, MD: Institute for Experimental andClinical Pharmacology and Toxicology, Department ofClinical Pharmacology, Center for Geriatric Pharmacology,Medical Faculty Mannheim, University of Heidelberg,Maybachstr. 14, 68169Mannheim, GermanyHeinrich Burkhardt, MD: University Medical CentreMannheim, 4thDepartment of Medicine, GeriatricMedi-cine, Theodor-Kutzer-Ufer 1-3, 68167 Mannheim, GermanyLutz Frolich, MD: Central Institute of Mental Health, J5,68159Mannheim, GermanyStefan Schwarz, MD: Central Institute of Mental Health,J5, 68159Mannheim, GermanyUlrich Wedding, MD: University Hospital, Clinic forInternal Medicine II, Erlanger Allee 101, 07747 Jena, GermanyExpertPanelMembersandTheirAfliationsJurgenBauer, MD: GeriatricsCentreOldenburg, Uni-versityofOldenburg, Rahel-Straus-Strae10, 26133Old-enburg,GermanyHeiner K. Berthold, MD: Charite UniversityMedicineBerlin, Evangelical GeriatricsCenterBerlin(EGZB), Re-inickendorferStr. 61,13347Berlin, GermanyPeterDovjak, MD: GmundenHospital, Department ofAcuteGeriatricMedicine, Miller-von-Aichholz-Strae49,A-4810Gmunden, AustriaHelmut Frohnhofen, MD: Essen-MitteHospital, Knap-pschaftsHospital, TeachingHospital, UniversityofDuis-burg-Essen, Am Deimelsberg34a, 45276 Essen, Germany,and Faculty of Health, University of Witten-Herdecke,Witten, GermanyThomasFruhwald, MD: HietzingHospital andNeuro-logical Center Rosenhugel, Wolkersbergenstrae1, 1130Vienna,AustriaChristoph Gisinger, MD: Haus der Barmherzigkeit,Danube University Krems, Seebockgasse 30a, 1160Vienna,AustriaManfredGogol, MD: LindenbrunnHospital, GeriatricDepartment, Lindenbrunn 1, 31863 Coppenbruegge,GermanyMarkus Gosch,MD:RegionalHospital Hochzirl,Anna-DengelHouse,6170Zirl, AustriaHans Gutzmann, MD: Hedwigshohe Hospital, Clinic forPsychiatry, Psychotherapy and Psychosomatic Medicine,Hohensteig1, 12526Berlin, GermanyIsabellaHeuser, MD: Charite UniversityHospital Ber-lin, DepartmentofPsychiatryandPsychotherapy, Univer-sity Medicine Berlin, Campus Benjamin Franklin,Eschenallee3, 14050Berlin, GermanyWernerHofmann, MD:FriedrichEbertHospital,Clinicfor GeriatricMedicine, Friesenstrasse11, 24534Neumu-enster, GermanyMichael Hull, MD: Center for GeriatricMedicineandGerontology Freiburg, University Clinic Freiburg, LehenerStrae88,79106Freiburg, GermanyBernhardIglseder, MD: Department ofGeriatricMed-icine, Christian-Doppler-Klinik, Paracelsus Medical Uni-versity,Ignaz-Harrer-Str.79, 5020Salzburg, AustriaAnjaKwetkat, MD: Jena UniversityHospital, Depart-ment of Geriatric Medicine, Bachstrae 18, 07743Jena,GermanyMichael Meisel, MD: Deaconess Hospital Dessau,ClinicforInternalandGeriatricMedicine,Gropiusallee3,06846Dessau, GermanyWolfgang Muhlberg, MD: Clinic for Internal Medicine 4 Geriatric Medicine, Frankfurt Hochst Hospital, Gote-nstrae6-8, 65929FrankfurtamMain, GermanyWolfgangvonRenteln-Kruse,MD:AlbertinenHospital/Albertinen House, non-prot company (GmbH), Center forGeriatricMedicineandGerontology, ScienticInstitutionat theUniversityofHamburg, Sellhopsweg18-22, 22459Hamburg,GermanyRegina Roller, MD: Medical University of Graz,Department of Internal Medicine, Auenbruggerplatz 15,8036Graz, AustriaRalf-JoachimSchulz, MD: Geriatric Clinic at the St.-Marien Hospital, Kunibertkloster 11-13, 50668Koln,GermanyUlrikeSommeregger, MD: HietzingHospital andNeu-rological Center Rosenhugel, Wolkersbergenstrae 1, 1130Vienna,Austria.References1. USDepartment of Commerce, USCensus Bureau. The olderpopulation; 2011 (online). http://www.census.gov/population/age/.Accessed2Jul2013.2. BohmK, Tesch-Romer C, ZieseT(eds) Gesundheitsberichter-stattungdesBundes2009: Gesundheit undKrankheit imAlter.Berlin:Robert Koch Institute;2009 [in German].http://www.rki.de/DE/Content/Gesundheitsmonitoring/Gesundheitsberichterstattung/GBEDownloadsB/alter_gesundheit.pdf?__blob=publicationFile.Accessed2Jul2013.3. KaufmanDW, KellyJP, RosenbergL, et al. Recent patternsofmedicationuseintheambulatoryadultpopulationoftheUnitedStates:theSloneSurvey. JAMA.2002;287:33744.4. Van denAkker M, BuntinxF, Metsemakers JF, et al. Multi-morbidityingeneral practice: prevalence, incidence, anddeter-minants of co-occurringchronicandrecurrent diseases. J ClinEpidemiol. 1998;51:36775.5. Mangoni AA, JacksonSHD. Age-relatedchangesinpharmaco-kinetics andpharmacodynamics: basic principles andpracticalapplications. BrJClinPharmacol. 2003;57:614.ConsensusValidationoftheFORTAList 1396. Edwards IR, AronsonJK. Adverse drugreactions: denitions,diagnosis, andmanagement. Lancet. 2000;356:12559.7. RochonPA, GurwitzJH. Optimisingdrugtreatment forelderlypeople:theprescribingcascade. BrMedJ. 1997;315:10969.8. ChykaPA. HowmanydeathsoccurannuallyfromadversedrugreactionsintheUnitedStates?AmJMed. 2000;109:12230.9. LazarouJ, PomeranzBH, CoreyPN. Incidenceofadversedrugreactionsinhospitalizedpatients:ameta-analysisofprospectivestudies. JAMA. 1998;279:12005.10. Borchelt M. Potential side-effects andinteractions inmultiplemedicationinelderlypatients: methodologyandresults of theBerlinStudyofAging.ZGerontolGeriatr. 1995;28:4208.11. Michel J-P, Pils K, Sieber C. Commentary on The future historyof geriatrics: geriatricsat thecrossroads. JGerontol MedSci.2002;57A:M8123.12. Evans JG. Geriatric medicine: a brief history. Br Med J.1997;315:10757.13. Wehling M. Guideline-driven polypharmacy in elderly, multi-morbid patients is basically awed: there are almost no guidelinesforthesepatients. JAmGeriatrSoc. 2011;59:3767.14. GillickM. Choosingappropriate medical care for the elderly.JAmMedDirAssoc. 2001;2:3059.15. BarenholtzLevyH, MarcusEL, ChristenC. BeyondtheBeersCriteria: acomparativeoverviewof explicit criteria. AnnPhar-macother. 2010;44:196875.16. BeersMH. Explicit criteriafordeterminingpotentiallyinappro-priate medication use by the elderly: an update. Arch Intern Med.1997;157:15316.17. Fick DM, Semla TP. American Geriatrics Society Beers Criteria:newyear, newcriteria, newperspective. J AmGeriatr Soc.2012;60:6145.18. Holt S, Schmiedl S, Thurmann PA. Potentially inappropriatemedications intheelderly: thePriscus List. DtschArztebl Int.2010;107:54351.19. PageRL, RuscinJM. Theriskofadversedrugeventsandhos-pital-related morbidity and mortality among older adults withpotentially inappropriate medication use. AmJ Geriatr Phar-macother. 2006;4:297305.20. GallagherP,RyanC,ByrneS,etal.STOPP(ScreeningToolofOlder Persons Prescriptions) and START(Screening Tool toAlert Doctors toRight Treatment). Consensus validation. Int JClinPharmacolTher. 2008;46:7283.21. Barry PJ, Gallagher P, Ryan C, OMahony D. START (screeningtool toalert doctors totheright treatment): anevidence-basedscreeningtool todetectprescribingomissionsin elderlypatients.AgeAgeing.2007;36:6328.22. Lam MPS, Cheung BMY. The use of STOPP/START criteria as ascreening tool for assessing the appropriateness of medications inthe elderly population. Expert Rev Clin Pharmacol. 2012;5:18797.23. WehlingM. Drugtherapyintheelderly: toomuchortoolittle,what to do? A new assessment system: t for the aged (FORTA).DtschMedWochenschr. 2008;133:228991[inGerman].24. WehlingM. Multimorbidityandpolypharmacy: howtoreducetheharmful drugloadandyet addneededdrugsintheelderly?Proposal of a newdrug classication: t for the aged. J AmGeriatrSoc. 2009;57:5601.25. Wehling M, Burkhardt H. Arzneitherapie fur Altere. 2nd ed.Heidelberg:Springer;2011[inGerman].26. HsuC, SandfordBA. The Delphi technique: makingsense ofconsensus.PractAssessResEval2007;12:17.http://pareonline.net/getvn.asp?v=12&n=10.Accessed2Jul2013.27. Hader M. Delphi-Befragungen: EinArbeitsbuch. 2nded. Wie-sbaden: VSVerlagfur Sozialwissenschaften/GWVFachverlageGmbH;2009[inGerman].28. Cicchetti DV, AllisonT. Anewprocedure for assessingreli-abilityof EEGsleeprecordings. AmJEEGTechnol. 1971;11:1019.29. CohenJ. Weightedkappa: nominal scaleagreement withprovi-sion for scaled disagreement or partial credit. Psychol Bull.1968;70:21320.30. Stevens SS. Onthetheoryof scales of measurement. Science.1946;103(2684):67780. http://www.jstor.org/stable/1671815.Accessed12Oct2013.31. Altman DG. Practical statistics for medical research. London:Chapman&Hall;1990. p.15.32. Fitch K, Bernstein S, Aguilar MD, et al. The RAND/UCLAappropriateness methodusers manual. Santa Monica: RAND;2001.33. Stitt-GohdesWL, CrewsTB. TheDelphi technique: aresearchstrategy for career and technical education. JCTE. 2004;20:5567.34. MustoDF. The Americandisease: origins of narcotic control,expanded edition. New York: Oxford University Press, Inc; 1987.35. Alexopoulos GS, StreimJ, Carpenter D, Docherty JP, ExpertConsensus Panel for Using Antipsychotic Drugs in OlderPatients. Using antipsychotic agents in older patients. J ClinPsychiatry. 2004;65Suppl2:599.36. StevensonDG, DeckerSL, DwyerLL, et al. Antipsychoticandbenzodiazepine use among nursing home residents: ndings fromthe2004National NursingHomeSurvey. AmJGeriatrPsychi-atry. 2010;18:107892.37. Birks J, Grimley Evans J. Ginkgo biloba for cognitive impairmentand dementia. Cochrane Database Syst Rev. 2009;(1):CD003120.38. Frohnhofen H, Michalek C, Wehling M. Assessment of drugtreatment in geriatrics with the new FORTA criteria:preliminaryclinical experience. DtschMedWochenschr. 2011;136:141721[inGerman].39. FrohnhofenH, WehlingM, MichalekC. ErsteklinischeErfahr-ungenmit der neuenFORTA-Klassikation(Abstract). ZGer-ontolGeriatr. 2012;45:21[inGerman].40. Field TS, Mazor KM, Briesacher B. Adverse drug eventsresultingfrompatient errorsinolder adults. JAmGeriatr Soc.2007;55:2716.41. JamersonK,BakrisGL,Dahlof B,etal.Exceptionalearlybloodpressurecontrolrates:theACCOMPLISHTrial.BloodPressure.2007;16:806.42. MoserM. TheAscotTrial. JClinHypertens. 2005;7:74850.43. AmericanGeriatricsSociety2012BeersCriteriaUpdateExpertPanel. AmericanGeriatrics SocietyupdatedBeers Criteria forpotentiallyinappropriate medicationuseinolder adults. J AmGeriatrSoc. 2012;2012(60):61631.44. Wehling M. Drug therapy for the elderly. Vienna: Springer; 2013.45. FialovaD,TopinkovaE,GambassiG,AdHOCProjectResearchGroup, et al. Potentially inappropriate medication use amongelderly home care patients in Europe. JAMA. 2005;293:134858.140 A.M.Kuhn-Thieletal.