1 23
European Journal of ClinicalPharmacology ISSN 0031-6970Volume 68Number 8 Eur J Clin Pharmacol (2012)68:1123-1138DOI 10.1007/s00228-012-1238-1
Potentially inappropriate medications inthe elderly: a comprehensive protocol
Suzana Mimica Matanovi & VeraVlahovic-Palcevski
1 23
Your article is protected by copyright andall rights are held exclusively by Springer-Verlag. This e-offprint is for personal use onlyand shall not be self-archived in electronicrepositories. If you wish to self-archive yourwork, please use the accepted authorsversion for posting to your own website oryour institutions repository. You may furtherdeposit the accepted authors version on afunders repository at a funders request,provided it is not made publicly available until12 months after publication.
REVIEWARTICLE
Potentially inappropriate medications in the elderly:a comprehensive protocol
Suzana Mimica Matanovi & Vera Vlahovic-Palcevski
Received: 19 November 2011 /Accepted: 31 January 2012 /Published online: 24 February 2012# Springer-Verlag 2012
Abstract Elderly patients are at increased risk of drug-relatedmorbidity and mortality. Avoiding the use of potentially inap-propriate medications (PIMs) is one of the strategies that hasbeen widely adopted to reduce the harmful consequences ofdrug use. There are several PIM screening tools available. Inthis review, we provide an overview of existing screening toolsto detect PIMs in the elderly, emphasizing the advantages anddisadvantages of each. Combining previously published andadopted tools (adjusted Beers list, French consensus panel,McLeods list, and Lindblads list of clinically importantdrugdisease interactions), we develop a new comprehensivetool that also includes the adjusted Hanlons and Malones listsof potentially serious drugdrug interactions in the elderly. Inaddition to listed PIMs and clinically important drugdruginteractions, alternative therapeutic solutions are suggested.The new protocol differentiates: drugs with an unfavorablebenefit/risk ratio (to be avoided regardless of the underlyingdisease/condition), drugs with a questionable efficacy, anddrugs to be avoided with certain diseases/conditions, and pro-vides a list of potentially serious drugdrug interactions. A toolconsisting of PIMs and potential drugdrug interactions withinthe same protocol provides more comprehensive quality as-sessment of drug-prescribing behavior to the elderly, which inturn may lead to better prescribing practices.
Keywords Potentially inappropriate medications . Elderlypatients . PIM screening tools . Drug-prescribing behavior .
Drugdrug interactions
Introduction
The percentage of the total elderly population is increasingin most countries, and it is estimated that by 2050 almost30% of the population in developed countries will be over65 years of age [1]. Elderly patients consume approximately30% of all healthcare resources and, therefore, the growth ofthis population group will have significant implications onfuture healthcare budgets [2]. Elderly individuals often havemany chronic diseases and are consequently taking multiplemedications. They also have increased risk for adverse drugreactions (ADRs) due to age-related changes in the pharma-codynamics and pharmacokinetics of drugs, co-morbidities,and polypharmacy [3]. It may thus be anticipated that thisincrease in the numbers of elderly people will also lead tohigher drug-related morbidity and mortality [2, 4].
Suboptimal or inappropriate prescribing in elderly patientspose the risk of drug-related morbidity and mortality. Inappro-priate prescribing includes the prescribing of medications withpotentially serious drugdrug interactions or the underuse,overuse, and misuse of drugs. Misuse encompasses the use ofpotentially inappropriate medications (PIMs), inappropriatedose, or inappropriate duration of treatment. PIMs are definedas drugs with a potential risk that is higher than their potentialbenefit to the patient, particularly when safer alternative thera-pies exist for the same condition [4, 5].
Several screening tools for detecting PIMs in the elderlyhave been developed in the USA, Canada, and Europeancountries [616]. Their role is to optimize the appropriatenessof prescribing behavior and to reduce negative outcomes,
S. Mimica Matanovi (*)Clinical Pharmacology Unit, University Hospital Center Osijek,Osijek, Croatiae-mail: [email protected]
S. Mimica MatanoviDepartment of Pharmacology, Medical School Osijek,University J.J. Strossmayer Osijek,Osijek, Croatia
V. Vlahovic-PalcevskiClinical Pharmacology Unit, University Hospital Center Rijeka,Rijeka, Croatia
Eur J Clin Pharmacol (2012) 68:11231138DOI 10.1007/s00228-012-1238-1
Author's personal copy
including preventable adverse drug effects (ADEs). Thesescreening tools encompass lists of drugs that should generallybe avoided by the elderly and drugs that should be avoidedwithcertain diagnoses or conditions.While most of the protocols areexplicit (criterion-based), there is only one implicit (judgement-based) protocol, the Medication Appropriateness Index (MAI),which was developed by Hanlon and colleagues [17] .
The aim of this review is to critically evaluate availableprotocols for detecting PIMs in the elderly and summarize theseinto a new comprehensive and widely applicable protocol.
Overview of the existing screening tools for detectionof PIMs in the elderly
A literature search was performed within the MEDLINE,PubMed, OVID, and Google Scholar databases using MeSHterms aged, inappropriate prescribing, clinical proto-cols, medication errors, and polypharmacy. Articlespublished between January 1991 and December 2010 wereselected if they contained explicit criteria adressing poten-tially inappropriate prescribing in the elderly.
Table 1 summarizes the protocols and screening tools fordetection of PIMs prescribed to the elderly, published inchronological order. For each tool in the table, data on theauthors, year, country of origin, scope, main content, andmain advantages/disadvantages are presented. One of thetools was developed by Barry et al. [13] to evaluate under-prescribing, i.e., for detecting omission of evidence-basedmedications.
We considered the potential clinical applicability of thecriteria and PIMs resulting in common clinical consequen-ces as well as the wide applicability of a protocol to differenthealthcare settings and different geographical regions to beadvantageous. Limitations or disadvantages of a protocolwere considered to be a lack of clinical assessment andevaluation or validation of the protocol, incomplete druglistings, and/or obsolete drugs listed.
According to different authors, newer criteria offer cer-tain innovations and improvements compared to those onthe original Beers list and address drugdrug interactions,underuse of drugs, drugs with questionable efficacy, amongothers [1820]. However, while several new tools need to beevaluated and assessed in clinical studies, Beers criteriahave been the most widely used and an association of thelisted drugs with adverse outcomes has been shown bynumber of authors [2134].
The new comprehensive protocol
Taking into account the reported advantages and disadvan-tages of existing screening tools, it may be assumed that by
combining their clinically most useful parts together, itshould be possible to develop a new comprehensive tool.In evaluating the advantages of previous tools, we focusedon the potential clinical applicability of the criteria and onPIMs having common clinical consequences in the elderly[e.g., focusing on non-steroid anti-inflammatory drugs(NSAIDs), associated with multiple possible adverseoutcomes].
The new protocol is developed by combining the adjust-ed Beers list, the French consensus panel, McLeods list,and Lindblads list of clinically important drugdiseaseinteractions, with the addition of several new drugs [8,1012]. As part of the protocol, a list of clinically importantdrugdrug interactions in the elderly is developed by mod-ifying Malones and Hanlons lists and adding four newdrugdrug interactions [35, 36]. A clinically based approachwas used to build the protocol. The new criteria have beenshown by many researchers to be associated with adverseclinical and healthcare outcomes, thus confirming their rel-evance [2734]. By combining parts of the North Americanand European tools together, we assume that new combinedtool will be widely applicable in the elderly population(defined as 65 years or older).
The protocol groups PIMs into: (1) those with an unfa-vorable benefit/risk ratio (drugs to be avoided regardless ofthe underlying disease/condition); (2) drugs with a question-able efficacy, and (3) drugs to be avoided with certaindiseases/conditions. Those three major subgroups are de-fined similarly in the French consensus panel and on Beerslist. For each criterion listed, a possible alternative solutionis given [8, 12, 16]. Part (4) of the protocol presents the listof potentially serious drugdrug interactions in the elderlypopulation.
Drugs with unfavorable benefit/risk ratio(part 1 of new protocol)
The list of drugs with an unfavorable benefit/risk ratio isbased on a combination of the adjusted Beers list and theFrench consensus panel, as shown in Table 2, and consistsof 33 criteria or individual drugs [10, 12]. Drugs definedsolely by the French panel include antipsychotic drugs withanticholinergic properties and concomitant use of 2NSAIDs, clonidine, and moxonidine among the centrallyacting antihypertensives and dipyridamole (according toBeers list, dipyridamole is inappropriate medication onlyin patients receiving anticoagulant therapy or in those withblood clotting disorders). These drugs are included in thenew protocol because their use can lead to severe ADRs[3742] .
Drugs defined solely by the Beers list include: doxazosin,amiodarone, fluoxetine, thioridazine, ferrous sulfate >325mg/day, estrogens only (oral), methyltestosterone, long-term use
1124 Eur J Clin Pharmacol (2012) 68:11231138
Author's personal copy
Table1
Characteristicsof
existin
gscreeningtoolsforpotentially
inappropriatemedications
Authors
Year
Country
Scope
Maincontent
Prosandcons
Beersetal.[6]
1991
USA
Nursing
homeresidents6
5years
Delphiconsensusbased.
Thirtycriteriafordrugsto
beavoidedin
theelderly
Pros:thefirsttool
developedforPIM
screening
intheelderly
Cons:manydrugsfrom
thelistareunavailablein
other
countries.Developed
fornursinghomeresidentsbut
also
used
instudieswith
otherpatient
populatio
ns
Beers[7]
1997
USA
Allpatients6
5years
Delphiconsensusbased;
updatedandexpanded
version.
Fifteen
drugsom
itted
from
theoriginal
version.
Contains28
drugsor
drug
classesto
beavoidedin
ambulatory
elderlyindependentof
diagnosisand35
drugsor
drug
classesto
beavoidedin
patientswith
certaindiseaseor
condition
Pros:moregenerally
applicable(for
ambulatory
patients)
Cons:manydrugsfrom
thelistareunavailablein
other
countries.Drugdrug
interactions
andduplicationof
treatm
entsarenotevaluated
McL
eodetal.[8]
1997
Canada
Allpatients6
5years
Delphiconsensusbased.
Thirty-eightinappropriate
practices
(grouped
into
cardiovascular,psychotropic,
analgesicandmiscellaneousdrugs)
drug
generally
contraindicated(18)
drug
diseaseinteractions
(16)drug
drugs
interactions
(4)
Pros:Nineinappropriatepractices
addressprescribingof
NSAID
s,includinglong-term
prescriptio
nin
patients
with
ahistoryof
pepticulcer,hypertension,chronic
renal,or
congestiveheartfailure.D
rugdrug
interactions
addressed.Alternativetherapyforeach
criterion
suggested
Cons:someof
thecriteriaobsolete(e.g.,betablockers
inpatientswith
asthmaor
COPDor
betablockersin
patientswith
congestiv
eheartfailu
re)
Naughleretal.[9]
2000
Canada
Allpatients7
0years
Derived
from
McL
eodslist.Fourteeninappropriate
combinatio
nsof
drugsanddiseases
Pros:simpleandeasily
applicabletool
Cons:someof
thecriteriaobsolete(e.g.,betablockersin
patientswith
asthmaor
COPDor
betablockersin
patientswith
congestiv
eheartfailu
re).Three
ofthe
criteriainvolvetodayuncommonly
used
tricyclic
antid
epressants
Ficketal.[10]
2003
USA
Allpatients6
5years
Delphiconsensusbasedupdatedversionof
Beerslist.
Sixty-eight
criteria:48
drugsor
drug
classesgenerally
tobe
avoided;
20diseases
orconditionswith
drugs
tobe
avoided
Pros:themostwidelycitedexplicitcriteria.Associatio
nwith
adversehealthcare
outcom
esshow
n
Cons:manydrugsfrom
thelistareunavailablein
other
countries.Drugdrug
interactions
andduplicationof
treatm
entsarenotevaluated.
Appropriateness
ofsome
drugsfrom
theliststill
subjectof
debate(e.g.,
amiodarone
oram
itriptylin
e).Onlyfour
inappropriate
practices
associated
with
theuseof
NSAID
saddressed
Lindbladetal.[11]
2006
USA
Allpatients6
5years
Delphiconsensusbased.
Twenty-eight
clinically
importantdrug
diseaseinteractions,involving14
diseases
orconditions.Elevendrug-disease
interactions
included
onBeerslistand5included
inMcL
eodslist
Pros:simpleandeasily
applicabletool.Introduces
newcriterianotdefinedby
Beersor
McL
eodslist
Cons:drugsto
beavoidedregardless
ofadiseaseor
condition
arenotincluded
Laroche
etal.[12]
2007
France
Allpatients7
5years
Delphiconsensusbased,
firstEuropeanscreeningtool.
Thirty-four
criteriaforinappropriateness.PIM
sgrouped
into
thosewith
unfavorablebenefit/riskratio
(25criteria),
questio
nableefficacy
(one
criterion),andboth
unfavorablebenefit/riskratio
andquestio
nableefficacy
Pros:Alternativedrugsor
therapeutic
abstentio
nfor
each
criterion
suggested.
The
firsttool
toaddress
drugswith
questio
nableefficacy
aspotentially
Eur J Clin Pharmacol (2012) 68:11231138 1125
Author's personal copy
Table1
(contin
ued)
Authors
Year
Country
Scope
Maincontent
Prosandcons
(7criteria).Tw
enty-ninecriteriaareindependentof
adiseaseor
condition
andfive
arelin
kedto
diseaseor
condition
inappropriate(cerebralvasodilators).Duplication
oftreatm
entsaddressedas
potentially
inappropriate
Cons:underuse
ofdrugsisnotaddressed.
Needs
tobe
assessed
andconfirmed
inclinicalstudies
Barry
etal.[13]
2007
Ireland
Allpatients6
5years
Delphiconsensusbased.
STA
RT(Screening
Tool
toAlert
Doctorsto
Right
Treatment).Tw
enty-twomedications
included.Underprescribingor
omission
ofclinically
indicated,
evidence-based
medications
evaluated(e.g.,
ACEinhibitorsin
chronicheartfailu
reor
betablockerin
chronicstableangina,ifno
contraindicatio
ns).
Pros:firsttool
evaluatin
gunderuse
ofdrugs
Gallagher
etal.[14]
2008
Ireland
Allpatients6
5years
Delphiconsensusbased.
STOPP(Screening
Tool
ofOlder
Persons
Prescription).Sixty-fivecriteriaforPIM
sevaluatio
narranged
accordingto
relevant
physiological
system
s.
Pros:innovativ
eapproach
introduced
(firsttool
toaddressinappropriateuseof
PPIforpepticulcerand
useof
aspirinwith
outhistoryof
coronary,cerebral
orperipheralvascular
symptom
sor
occlusive
event).
Seven
criteriaaddressprescribingof
NSAID
s,includingprescriptio
nin
patients
hypertension,chronicrenalor
congestiv
eheartfailu
re.
Drugclassduplicationanddrug
drug
interactions
addressedas
potentially
inappropriate.Sensitiv
efor
identifying
patientswith
potentialto
suffer
PIM
s-relatedADRs
Cons:drugswith
questio
nableefficacy
notaddressed.
Needs
tobe
assessed
andconfirmed
inclinicalstudies.
Rognstadetal.[15]
2009
Norway
Patients7
0yearsin
generalpractice
Delphiconsensusbased.
Twenty-one
explicitcriteriafor
singledrugsand15
criteriafordrug
drug
interactions.
Pros:Listsinappropriatesingledrugs(i.e.,theophyllin
eor
sotalol)anddrug
combinatio
ns[e.g.,combinatio
nof
NSAID
s(orcoxib)
andACEinhibitors(orARBs)
which
may
increase
therisk
ofrenalfailu
re,or
combinatio
nof
NSAID
sanddiuretics,resulting
inreduceddiureticeffect]notaddressedby
previous
tools
Cons:aimed
atpatientsin
generalpractice.Needs
tobe
assessed
andconfirmed
inclinicalstudies
Holtetal.[16]
2010
Germany
Allpatients6
5years
Delphiconsensusbased.
Listsatotalof
83drugsto
beavoidedregardless
oftheunderlying
disease/condition,
containedin
18drug
classes.
Pros:easily
applicabletool.Nam
esmainconcerns,
possibletherapeutic
alternatives
andprecautio
nsto
betaken.
Listsmanydrugsnotaddressedby
previous
tools(e.g.,ketoprofen,m
eloxicam
,prasugrel,flecainide,
metildigoxin,haloperidol>2mg,zaleplon
>5mg,
phenobarbital).D
rugs
with
questionableefficacy
are
addressed(circulation-prom
otingagents)
Cons:aimed
prim
arily
atGerman
elderlypopulatio
n.Needs
tobe
assessed
andconfirmed
inclinicalstudies
ACE,A
ngiotensinconvertin
genzyme;ADRs,adversedrug
reactio
ns;A
RBs,angiotensinIIreceptor
blockers;C
OPD,obstructiv
epulm
onarydisease;NSAID
s,non-steroidanti-inflam
matorydrugs;
PIM
S,potentially
inappropriatemedications,PPI,proton
pumpinhibitor
1126 Eur J Clin Pharmacol (2012) 68:11231138
Author's personal copy
Table2
Drugs
with
unfavorablebenefit/riskratio
Drug
Possibleadverseeffects
Possibletherapeutic
solutio
ns
Analgesics
Indomethacin
SevereCNSside
effects
Short-termuseof
aweakNSAID
(e.g.,ibuprofen)
oracetam
inophen
oraweakopioid
(e.g.,tram
adol)
Concomitant
useof
2or
moreNSAID
sNoenhancem
entof
efficacy,increasedrisk
ofADRs
Short-term
useof
only
oneweakNSAID
(e.g.,ibuprofen)
Long-term
useof
full-dosage,longer
half-lifeNSAID
s:naproxen,
piroxicam
Increasedrisk
ofGIbleeding,renalfailu
re,h
ighbloodpressure
andheartfailu
reShort-term
useof
aweakNSAID
(e.g.,ibuprofen)
oruseof
acetam
inophenor
aweakopioid
(tramadol,codeine)
Drugs
with
anticholin
ergicproperties
Antidepressants:am
itriptylin
e,maprotiline
Muscarinic-blocking
side-effects,cardiotoxicity
when
overdosed
SSRIs(exceptflouxetin
e)or
SNRIs
Antipsychoticdrugs:fluphenazine,levomepromazine
Muscarinic-blocking
side-effects
Atypicalantipsychoticdrug
with
lessanticholinergicactivity
(e.g.,
olanzapine,risperidone,quetiapine)
Antihistamines:diphenhydram
ine,dimenhydrinate
Muscarinic-blocking
side-effects,sedatio
n,drow
siness
Antihistamines
with
outanticholin
ergicactiv
ity(e.g.,cetirizine,
levocetirizine,loratadine,desloratadine)
Concomitant
useof
drugswith
anticholin
ergicproperties
Enhancedanticholin
ergicADRs
Avoid
drugswith
anticholin
ergicactiv
ityin
general
Sedativeor
hypnoticdrugs
Long-actin
gbenzodiazepines:diazepam
,bromazepam
,nitrazepam
,flurazepam
Prolonged
sedatio
nanddrow
siness,increasedrisk
offalls
Short-actingbenzodiazepinesgivenin
dose
halfthedose
inyoungeradults
Short-actingbenzodiazepines,dose
>halfthedose
inyounger
adults(lorazepam
>3mg,
oksazepam
>60
mg,
alprazolam
>2mg)
Increasedrisk
ofADRswith
outincreasedefficacy
Short-actingbenzodiazepinesgivenin
dose
halfthedose
inyoungeradults
Meprobamat
Verysedativ
eproperties,addictivewith
prolongeduse
Short-actingbenzodiazepinesgivenin
dose
halfthedose
inyoungeradults
Antihypertensives
Methyldopa
Bradycardia,exacerbatio
nof
depression
Other
antih
ypertensivedrugs,except
theones
listedhere
[i.e.,
diuretics,calcium
channelblockers(exceptshort-actin
gones),
ACEinhibitors,AT1blockers]
Clonidine
Ortostatic
hypotension
Moxonidine
Headache,vertigo,
asthenia
Nifedipine,short-actin
gPosturalhypotension,
myocardialinfarctio
n,stroke
Doxazosine
Hypotension,drymouth,urinaryincontinence
Anti-arrhythm
ics
Amiodarone
Prolonged
QTinterval,risk
oftorsade
depointes,reduced
efficacy
intheelderly
Otherantiarrhythmics,dependingon
thetype
ofarrhythm
ia(e.g.,
propafenone,betablockers,calcium
channelblockers)
Disopiram
ide
Negativeinotropicandanticholin
ergicproperties
Digoxin
>0.125mg
Reduced
renalclearanceandincreasedrisk
ofADRs
Digoxin
