This article was downloaded by: [UOV University of Oviedo]On: 15 October 2014, At: 02:41Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Aging & Mental HealthPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/camh20

Barriers to antipsychotic discontinuation in nursinghomes: an exploratory studyMajda Azermaia, Robert R.H. Vander Stichelea, Luc M. Van Bortela & Monique M.Elseviersab

a Heymans Institute of Pharmacology, Division of Clinical Pharmacology, Ghent University,Gent, Belgiumb Department of Nursing Science, University of Antwerp, Antwerpen, BelgiumPublished online: 09 Sep 2013.

To cite this article: Majda Azermai, Robert R.H. Vander Stichele, Luc M. Van Bortel & Monique M. Elseviers (2014) Barriersto antipsychotic discontinuation in nursing homes: an exploratory study, Aging & Mental Health, 18:3, 346-353, DOI:10.1080/13607863.2013.832732

To link to this article: http://dx.doi.org/10.1080/13607863.2013.832732

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shallnot be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and otherliabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Barriers to antipsychotic discontinuation in nursing homes: an exploratory study

Majda Azermaia*, Robert R.H. Vander Stichelea, Luc M. Van Bortela and Monique M. Elseviersa,b

aHeymans Institute of Pharmacology, Division of Clinical Pharmacology, Ghent University, Gent, Belgium; bDepartment of NursingScience, University of Antwerp, Antwerpen, Belgium

(Received 19 April 2013; accepted 31 July 2013)

Objectives: Despite safety warnings on serious adverse effects and guidance advising discontinuation, antipsychotic use innursing homes remains high. Studies documenting the barriers experienced to antipsychotic discontinuation are rare. Thisexploratory study investigates the willingness of nurses and general practitioners (GPs) as well as the barriers to undertakeantipsychotic discontinuation.Design and setting: A mixed-method study involving an expert meeting, followed by a survey using structured question-naires distributed to responsible nurses (primary caregivers) and treating GPs on selected nursing home residents in Belgiannursing homes to generate case-specific information.Results: Antipsychotic users (n ¼ 113) had a mean age of 81 years (range 57–97); 62% were female and 81% hadmoderate to severe cognitive impairment. Nurses and GPs indicated a willingness for antipsychotic discontinuation in asmall proportion of residents, 13.8% and 12.2%, respectively, with a shared willingness in only 4.2%. Residents for whomthere was a higher willingness to try antipsychotic discontinuation were generally older (mean age 84.6 vs. 80.3, p ¼0.07), had high physical dependency (ADL > 14, 93.3% vs. 60.9%, p ¼ 0.01) and resided on a ward with controlled access(80.0% vs. 45.7%, p ¼ 0.02). In contrast, residents for whom there was a significant lower willingness for discontinuationalready had a previously failed discontinuation effort, and may present risk of harm to themselves or to others. Nursesworking longer on the ward, with lower education, presented higher barriers to discontinuation of antipsychotics.Conclusion: Nurses and GPs share a very low willingness and high barriers to antipsychotic discontinuation. To implementdiscontinuation programs, complex multidisciplinary interventions should be offered taking existing barriers into account.

Keywords: psychological and behavioral symptoms; pharmacological treatment; institutional care (nursing homes etc.)

Introduction

Dementia, with Alzheimer’s disease as the most common

cause, has an increasing incidence as our population

grows older (Brookmeyer, Johnson, Ziegler-Graham, &

Arrighi, 2007). Dementia is characterized by progressive

cognitive decline, loss of functional capacity and appear-

ance of behavioral disturbances (Ballard, Creese, Corbett,

& Aarsland, 2011). Noncognitive symptoms currently

labeled as behavioral and psychological symptoms of

dementia (BPSD) are very common (Lyketsos et al.,

2002) and present a risk for increased caregiver burden

and institutionalization.

Evidence-based guidelines state that antipsychotic

treatment for BPSD should only be initiated after a seri-

ous risk–benefit analysis (Azermai et al., 2012). Never-

theless, antipsychotics are often initiated in nursing

homes (Bronskill et al., 2004) as a first-line treatment

for BPSD (Briesacher et al., 2005), resulting in high

and chronic utilization. Over the last years, increasing

concerns have been reported regarding efficacy, tolera-

bility and safety of antipsychotic agents when used by

older people (Burke & Tariot, 2009). Serious adverse

events such as stroke and death have been reported

(Schneider, Dagerman, & Insel, 2005) even in short

studies of 12 weeks use, while antipsychotics are often

chronically used (Azermai, Elseviers, Petrovic, van

Bortel, & Stichele, 2011). Consequently, several safety

warnings on antipsychotic use for people with dementia

followed (European Medicines Agency [EMA], 2004;

Medicines and Healthcare Products Regulatory

Agency [MHRA], 2008, 2009; US Food and Drug

Administration [FDA], 2005, 2008). Adverse events

may outweigh the supporting evidence of efficacy,

weakening the recommendation to prescribe antipsy-

chotics (Azermai et al., 2011). Time-limited use and

discontinuation efforts should be the rule, not the

exception. Despite safety warnings and guidelines

advising temporary use, the use of antipsychotics in

nursing homes remains high and chronic. This indicates

that potential barriers may exist in clinical practice to

performing antipsychotic discontinuation. Therefore,

this study explores the willingness of nurses and gen-

eral practitioners (GPs) to discontinue antipsychotics

and identifies barriers to antipsychotic discontinuation.

Methods

We conducted a mixed-method study involving an expert

meeting followed by a survey using structured question-

naires distributed to the responsible nurse (primary care-

giver) and the treating GP on selected nursing home

residents.

*Corresponding author. Email: majda.azermai@ugent.be

� 2013 Taylor & Francis

Aging & Mental Health, 2014

Vol. 18, No. 3, 346–353, http://dx.doi.org/10.1080/13607863.2013.832732

Dow

nloa

ded

by [

UO

V U

nive

rsity

of

Ovi

edo]

at 0

2:41

15

Oct

ober

201

4

Expert meeting

The initial systematic search (Pubmed, Web of Science,

EBSCO host, Google Scholar) for studies on barriers to

antipsychotic discontinuation in older people did not result

in the retrieval of relevant studies. As we were confronted

with a limited evidence base, we decided to have an expert

meeting – also known as nominal group technique – as a

qualitative research method. In an expert meeting, exchange

and interaction between group members is more guided

than in focus groups and seeks to provide information on

the extent to which experts agree about a given issue. The

result is data that comprise the core set of information

(Powell & Single, 1996). This critical information allowed

us to gain insight in this little researched area, and was the

framework for the development of a questionnaire with a

focus on barriers to antipsychotic discontinuation. After

approval of the Ethics Committee of the Antwerp

University Hospital (approval number 10/45/316), we orga-

nized a multidisciplinary expert meeting. The expert panel

was composed of participants who shared the key character-

istic of working in a nursing home setting as care professio-

nals. For the recruitment of participants, we contacted the

management of several nursing homes (convenience sam-

ple) in the area of Antwerp for collaboration, both public

and private (not-for-profit) and of all sizes. A letter was sent

to the management of selected nursing homes informing

them about the subject of the expert meeting, asking for

participation and to delegate nurses or head nurses, and

other care professionals to represent their staff.

A topic list for the expert meeting was added to the

invitational letters. The items on the topic list were

arranged according to the process of antipsychotic

prescribing and included general questions on initiation,

re-assessment, and discontinuation of antipsychotics. The

nursing home management was asked to distribute the

topic list on the floor and to use it to prepare some ideas,

suggestions, and experiences from their nursing home on

this issue. The researchers received confirmation of the

people who were willing to participate. At the expert

meeting, all participants were informed again on the

process of the expert meeting with taped discussions, and

given the opportunity to refuse participation. People who

wanted to participate gave their consent by signing a list

and by participating in the expert panel.

For credibility, we used site triangulation, by including

participants within several nursing homes so as to reduce

the effect of local factors peculiar to one institution (Powell

& Single, 1996).

During the multidisciplinary expert meeting, the par-

ticipants’ disciplines were equally divided into two groups

of eight with a moderator in each group guiding the dis-

cussion. Other than nurses, participants of the expert

meeting included GPs, head nurses, nurse assistants, and

pharmacists. The moderator had the responsibility to fol-

low each item on the topic list (initiation, re-assessment,

discontinuation, and barriers) making sure every partici-

pant shared his/her experience. For each item on the topic

list, the core set of information had to be gathered. This

was done at place by a secretary noting the main

statements of participants. Additionally, the discussions

were audio-taped. In the light of gathering a credible, core

set of information we also used the principle of member

check. After the discussion, the results of each group were

presented in a final session by the moderators, leaving the

opportunity for further discussion, final remarks, or cor-

rection of wrong interpretations. Furthermore, the analysis

of the expert meeting involved a transcript of the final pre-

sentations, the general discussions, and the conclusions

drawn. We used the transcripts for further analysis based

on the scissor-and-sort technique. With this technique we

identified those sections from the transcript that were rele-

vant to the research question. We identified information

from the transcript that matched the information from the

predefined topics categorized as ‘initiation’, ‘re-asses-

sment’, and ‘discontinuation’.

Development of the questionnaire

In this study, we used the expert meeting as a qualitative

strategy to define a core set of information first. This criti-

cal information became a starting point for the develop-

ment of a questionnaire, including statements on barriers.

The quantitative survey was performed to additionally

collect case-specific information. This questionnaire fol-

lowed the framework from the topic list with sections

focusing on antipsychotic initiation, re-assessment, and

barriers to discontinuation.

For initiation, we mainly focused on time, place, and

reasons for initiation using the clustered symptoms of the

validated BEHAVE-AD (Behavioral pathology in

Alzheimer’s Disease) (Reisberg et al., 1987) as a basis.

For re-assessment, we focused on re-evaluation of the

antipsychotic prescription, dose adjustments as well as on

(multidisciplinary) general medication review.

For discontinuation, the willingness to stop antipsy-

chotics was evaluated.

For the barriers of discontinuation, we developed a set

of 13 statements resulting from the transcript analysis of

the expert meeting. These statements were considered to

be the main barriers to antipsychotic discontinuation as

discussed by the experts. The barriers were scored on a

scale from 1 indicating no barrier to antipsychotic discon-

tinuation, to 10 indicating an extreme barrier to antipsy-

chotic discontinuation.

The questionnaire was reviewed, revised, and then

approved by a selected group (n ¼ 6) of nurses, GPs, and

pharmacists.

Setting

The Belgian long-term residential care infrastructure for

older people consists of public or private nursing homes,

offering home replacement with and without intensive nurs-

ing care. Nursing home residents are still supervised by their

own GP. Residents with advanced dementia reside in a ward

with controlled access, while residents with a normal cogni-

tive status reside in an open ward or in a mixed ward

together with residents with cognitive impairment.

Aging & Mental Health 347

Dow

nloa

ded

by [

UO

V U

nive

rsity

of

Ovi

edo]

at 0

2:41

15

Oct

ober

201

4

Selection of residents

In a convenience sample of four nursing homes,

anonymized medication charts of all residents were

screened by the researcher for antipsychotic users. Only

residents using antipsychotics for at least one month were

selected.

Procedure and data collection

The developed questionnaire was presented to the respon-

sible nurse and the responsible GP of the antipsychotic

user/resident, independently from each other. The

questionnaires were not intended as a general survey to

collect general opinions, but focused on individual,

selected antipsychotic users (i.e. case-specific informa-

tion). Personal information about the cases was not known

by the researchers. The interviewed nurses and GPs were

allowed to use the resident’s medical and nursing records.

GPs received the questionnaires by mail from the medical

coordinator as an intermediary. They filled out the

questionnaires and returned them in pre-stamped, anony-

mized envelopes. Questionnaires filled out by responsible

nurses were collected at the nursing home from the

management.

Data collection included informants’ (i.e. head nurses

and GPs) characteristics, but also clinical information

such as residents’ physical dependency and mental status.

Physical dependency was assessed using activities of daily

living (ADL) (Katz & Akpom, 1976), which consists out

of six categories scored from 1 (independent) to 4 (fully

dependent). Cognitive impairment was assessed using dis-

orientation in time and place from the Katz-scale (Katz &

Akpom, 1976) as a proxy. This instrument is mandatory

in Belgian nursing homes, and assesses the mental status

by disorientation in time and place ranging in this study

from 1 (normal) to 4 (severe disorientation considered as

a proxy for severe cognitive impairment).

Possible indications for antipsychotic treatment were

based on the BEHAVE-AD and included paranoid and

delusional ideation, hallucinations, activity disturbances

(e.g. wandering, purposeless activity), aggression, diurnal

rhythm disturbances (i.e. day/night disturbance), and

affective disturbance (e.g. depressed mood).

The barriers were scored on a scale from 1 to 10. For

each of the selected residents the treating GP and

responsible nurse (primary caregiver) had to express the

strength of barriers to discontinuation of antipsychotics by

scoring each of the 13 statements.

Statistical analysis

The data were analyzed using SPSS 15.0 (SPSS, Inc.,

Chicago, IL, USA). Differences, comparison of means

between groups, and associations were analyzed through

cross-tabulation, correlations, and logistic regression

models. The level of significance was set at p < 0.05.

ADL ranks adequacy of performance in the six functions

of bathing, dressing, toileting, transferring, continence,

and feeding (total ADL score ranging from minimum 6 to

maximum 24). We defined high physical dependency as

an ADL score of >14. The 13 potential barriers were

scored on a scale from 1 to 10, and summed into a barrier

sum score ranging from 13 as a minimum score to 130 as

a maximum score.

Results

For this exploratory study, we included four nursing

homes – three private (not-for-profit) nursing homes and

one public nursing home. The number of beds were 51,

96, 110 and 190, respectively, per nursing home. The

average prevalence of antipsychotic use (>1 month) was

found to be 25% (range 19%–31%). There were a total of

226 nurses and nurse assistants, ranging from 20 to 98 per

nursing home, with a resident-to-nursing-staff ratio of 1.8,

1.9, 2.4, and 2.6, respectively, per nursing home. There

were a total of 181 visiting GPs ranging from 28 to 66 per

nursing home, with a resident-to-GP ratio of 1.7, 1.8, 2.8,

and 3.6, respectively, per nursing home.

From the 4 included nursing homes, we identified 113

antipsychotic users (>1 month use). The mean age of

included antipsychotic users was 81 years (range 57–97)

of whom 62% were female (Table 1). Antipsychotic users

had a significant need of care with a mean ADL score of

16.4 (range 6–24) and had in 81.7% moderate to severe

cognitive impairment. Half of the antipsychotic users

resided in a ward with controlled access. Time since initi-

ation of antipsychotic treatment was long, with a median

use of 27 weeks (range 5–828). Atypical antipsychotics

were used by 76.8% of the residents, while typical agents

Table 1. Description of antipsychotic users (n ¼ 113).

Characteristics Mean (range)

Age 80.9 (57–97)Female 61.9%Activities of daily living 16.4 (6–24)Disorientation in time and placeStage 1 (normal) 9.2%Stage 2 (mild) 9.2%Stage 3 (moderate) 56.9%Stage 4 (severe) 24.8%

Length of nursing home stay (median, range) 18 (1–190 months)Duration of antipsychotic treatmenta (median, range) 27 (5–828 weeks)

aExact initiation date was only known for 56 residents.

348 M. Azermai et al.

Dow

nloa

ded

by [

UO

V U

nive

rsity

of

Ovi

edo]

at 0

2:41

15

Oct

ober

201

4

were used by 34.2%. Duplicate antipsychotic use was

found in 11%.

For each antipsychotic user, the responsible nurse and

GP received a questionnaire, independently from each

other. These questionnaires focused on individual,

selected antipsychotic users to retrieve case-specific infor-

mation. Thirteen nurses filled the questionnaires of all of

their residences under their responsibility (an average of

nine questionnaires per nurse).

Respondent nurses’ mean age was 41.2 years (range

27–56), working on average 7 years (range 0–23) on the

ward. For GPs, 28 (of a total of 60) returned 51 question-

naires (an average of two questionnaires per GP). Respon-

dent GPs’ mean age was 50.6 years (range 32–63)

working on average 24 years (range 3–39).

Comparison between nurses and GPs was therefore

limited, only possible for 51 of the 113 identified residents

on chronic antipsychotics, leading to a response rate of

45%. There were no significant differences in resident

characteristics between the population of 113 residents

and the response sample of 51 residents.

Initiation, re-assessment, and discontinuation

In 38.1% of cases, antipsychotic initiation occurred in the

nursing home, 26.5% was initiated at home, 16.8% during

hospitalization, and in 18.6% it was not known where

exactly the antipsychotic prescription was initiated.

In 82.3% of long-term antipsychotic users, behavioral

disturbance was the main reason for antipsychotic treat-

ment with activity disturbance (47.1%), agitation/aggres-

sion (39.4%), and hallucination (27.9%) as main

indications (Table 2).

A general medication chart review had occurred in

74.1% of cases in the last 3 months. Specific review of

antipsychotic treatment was performed in 57.5% of users.

Dose reduction of antipsychotics had been considered in

38.1% of the users, but an actual reduction effort was only

performed in 30.4%. In 20.5% of users, antipsychotic dis-

continuation was considered, but an actual discontinuation

attempt was only performed in 9.8%. Nurses (n ¼ 13) and

GPs (n ¼ 28) indicated that they were willing to try anti-

psychotic discontinuation in a small proportion of resi-

dents, 13.8% and 12.2%, respectively. Shared willingness

was even lower, only 4.2%. The percentages of overlap

(Table 2) between nurses and GPs were generally low,

indicating that nurses and GPs evaluate the same resident

differently.

The residents for whom there was a higher willingness

to try antipsychotic discontinuation by the nurse were

generally older (mean age 84.6 vs. 80.3, p ¼ 0.07), had

highly impaired ADL (ADL > 14, 93.3% vs. 60.9%, p ¼0.01), and were residing in a ward with controlled access

(80.0% vs. 45.7%, p ¼ 0.02). In contrast, residents for

whom there was a significant lower willingness to discon-

tinue antipsychotics already had a previously failed dis-

continuation effort, and presented a risk of harm to

themselves or to others (Table 3). Nurses who were will-

ing to discontinue antipsychotics generally had less work

experience on the ward (nonsignificant trend, p ¼ 0.09).

Barriers

In Figure 1, all barriers are presented as scored by nurses

and GPs. The mean scores on general barriers were com-

parable for nurses and GPs. In contrast, the level of agree-

ment between nurses and GPs when considering the same

resident was often low.

Nurses as well as GPs indicated that antipsychotic dis-

continuation would potentially negatively affect the qual-

ity of life of the resident. This was the barrier with the

highest score (Figure 1). Also, the recurrence of the initial

Table 2. Initiation, re-assessment, and discontinuation of antipsychotic treatment as indicated by nurses and GPs.

Results for n ¼ 51a

Indication for initiation of antipsychoticsIndicated by nurses

for n ¼ 113Indicatedby GPs

Indicatedby nurses

% positiveagreementb

Activity disturbance 47.1% 52.0% 43.5% 32.0%Aggression 39.4% 13.7% 23.5% 12.0%Hallucination 27.9% 7.8% 25.5% 8.0%Anxiety/phobia 22.1% 13.7% 27.5% 16.0%Delusion 17.3% 19.6% 17.6% 8.0%Diurnal rhythm disturbances 7.7% 8.2% 10.9% –Re-assessment in the last 3 monthsGeneral medication chart review 74.1% 65.3% 82.0% 60.4%Specific review of antipsychotics 57.5% 46.5% 58.8% 12.0%Change of antipsychotic prescription 41.6% 27.5% 41.2% 8.0%DiscontinuationWillingness to discontinue antipsychotics in this

resident13.8% 12.2% 12.0% 4.2%

Antipsychotic dose reduction has been considered 38.1% 70.6% 49.0% 37.3%Antipsychotic dose reduction has been performed 30.4% 62.7% 42.0% 14.0%Antipsychotic discontinuation has been considered 20.5% 35.3% 26.0% 13.0%An actual discontinuation effort has been performed 9.8% 13.7% 16.0% 6.0%

aThe level of agreement between nurses and GPs could only be assessed for residents of which nurses and GPs filled in the questionnaires (n ¼ 51).bPercentage of residents for which both nurses and GPs agreed on a positive answer.

Aging & Mental Health 349

Dow

nloa

ded

by [

UO

V U

nive

rsity

of

Ovi

edo]

at 0

2:41

15

Oct

ober

201

4

behavioral problems after antipsychotic discontinuation

was a major concern, as well as hindrance to others and

risk of harm to the resident. The potential of a higher

workload and a more intensive observation of the resident

when discontinuing antipsychotics was an important bar-

rier as well, but not a major one.

Nurses and GPs agreed on the majority of barriers for-

mulated for particular residents, with the exception of

nonpharmacological alternatives.

The internal validity of the set of 13 barriers was good,

enabling to make a sum score of barriers with Cronbach’s

a coefficients of 0.82 for nurses and 0.88 for GPs.

The mean barrier sum score of nurses was 62/130.

Nurses formulated significantly higher barriers in resi-

dents where there was no willingness to try discontinua-

tion, with a mean barrier sum score of 67.6 vs. 41.0 in

residents where a willingness existed (p< 0.001)

(Table 3). Nurses indicated having higher barriers for res-

idents with at least one disruptive behavior, especially

when residents used antipsychotics for aggression (52.8%

vs. 25.0%, p ¼ 0.005). Also, lower educated nurses and,

nurses working longer on the ward, had higher barriers.

GPs had a slightly higher mean barrier sum score (65/

130), experiencing in general slightly more barriers to

discontinuation.

Discussion

To our best knowledge, this study is the first exploratory

study to investigate barriers associated with the lack of

antipsychotic discontinuation in selected nursing home

residents. To explore this not-well-researched area, we

initiated an expert meeting, followed by a survey of nurses

and GPs on selected antipsychotic users (case-specific

information). Our exploratory study revealed that nurses

and GPs share high barriers and a low willingness to

undertake antipsychotic discontinuation, with a shared

willingness of only 4.2%. These low rates indicate that

the current antipsychotic users are likely to become

chronic users of antipsychotics, as was already indicated

in a previous study (Azermai et al., 2011). Long-term use

of antipsychotics and the associated lack of re-assessment

of the need for continuing use are important quality prob-

lems that need to be addressed in the future.

Nurses were more willing to discontinue antipsy-

chotics in physically dependent residents, preferably

residing in a ward with controlled access. Lower educated

nurses, and nurses working longer on the ward, generally

had higher barriers to antipsychotic discontinuation. Anti-

psychotic discontinuation is only likely to succeed with

involvement and cooperation of nurses. Therefore, suffi-

cient attention should be given to educational sessions, to

motivate nurses, especially those who are working long

on the ward. A decrease in use of antipsychotic drugs is

more likely to occur in homes with a ‘resident-centered

culture’ characterized by nurses favoring psychosocial

interventions, more regular review of drug use, and feed-

back from pharmacists to the staff nurses (Cody, Beck, &

Svarstad, 2002). This reinforces the need for multifaceted

interventions and multidisciplinary collaboration.

The main barrier opposing nurses and GPs to discon-

tinue antipsychotics was the potential adverse influence

on quality of life. Apparently, nurses and GPs consider

antipsychotic discontinuation to induce more suffering for

Table 3. Factors associated with the willingness of nurses to discontinue antipsychotics.

Resident characteristicsWilling to discontinue

(n ¼ 15)Not willing to

discontinue (n ¼ 94) p OR (95% CI)

Length of nursing home stay (mean in months) 31.2 28.2 0.762 1.01 (0.98–1.02)High physical dependency (ADL >14) 93.3% 60.9% 0.018 9.0 (1.13–71.43)Age (mean in years) 84.6 80.3 0.075 1.08 (0.99–1.19)Female gender 60.0% 62.8% 0.833 0.89 (0.29–2.71)Moderate to severe cognitive impairment 80.0% 82.6% 0.812 0.84 (0.21–3.33)Indications and initiationResidents with at least one behavioral problem 73.3% 84.0% 0.311 0.52 (0.15–1.86)Initiation at home 26.7% 26.6% 0.991 1.01 (0.29–3.44)Initiation at the hospital 13.3% 18.1% 0.658 0.69 (0.14–3.37)Initiation at the nursing home 20.0% 41.5% 0.112 0.35 (0.09–1.33)Barriers (mean score on 10)Intensive observation 4.3 5.7 0.147 0.88 (0.74–1.04)Failure of a previous effort 1.0 2.7 0.045 0.95 (0.71–1.27)Family is opposed 2.3 3.1 0.325 0.89 (0.72–1.11)Risk of harm to other residents 2.0 4.8 0.002 0.68 (0.52–0.89)Risk of harm to the staff 2.5 4.6 0.022 0.78 (0.62–0.98)Risk of harm to himself 3.3 6.5 < 0.001 0.73 (0.60–0.89)Recurrence of behavioral problems 5.5 7.9 < 0.001 0.73 (0.60–0.88)Hindrance to other residents 3.1 6.8 < 0.001 0.68 (0.65–0.84)Higher workload 3.2 5.5 0.012 0.80 (0.67–0.96)Affecting quality of life (negative) 5.3 8.4 < 0.001 0.62 (0.50–0.78)Insufficient nonpharmacological alternatives 1.3 2.8 0.061 0.67 (0.41–1.10)Physical restraint is the only alternative 2.0 4.5 0.02 0.78 (0.62–0.98)Type of ward 1.1 1.9 0.242 0.68 (0.30–1.53)Total barrier sum score (mean score on 130) 41.0 67.6 < 0.001 0.89 (0.84–0.95)

Note: Univariate logistic regression model.

350 M. Azermai et al.

Dow

nloa

ded

by [

UO

V U

nive

rsity

of

Ovi

edo]

at 0

2:41

15

Oct

ober

201

4

the nursing home resident and adversely affect the resi-

dent’s quality of life. Especially in this later phase of life,

it is important to focus on quality rather than on quantity

of life. In the Netherlands, a recent study of Cornege-

Blokland, Kleijer, Hertogh, and van Marum (2012) con-

firmed that in the majority of nursing home residents,

physicians and nurses felt that antipsychotic treatment

was the right decision and therefore justified. However,

the idea that antipsychotics improve quality of life is

rather in conflict with the evidence on limited efficacy,

and risk of adverse effects such as falls (Sterke et al.,

2012) and consecutive fractures, cognitive impairment

(Devanand & Schultz, 2011), stroke and even death

(Ballard et al., 2011; Schneider et al., 2005) in older peo-

ple with dementia.

Another major concern of nurses and GPs was the risk

of recurring behavioral problems following antipsychotic

discontinuation. Cohen-Mansfield et al. (1999) similarly

found that 54% of the nursing staff thought drug with-

drawal would lead to deterioration in behavior. Mean-

while, several studies investigated the withdrawal of

antipsychotics in people with dementia. The majority of

these studies indicate that antipsychotics can be discontin-

ued without detrimental effects on behavior (Ballard

et al., 2008; Cohen-Mansfield et al., 1999; Ruths, Straand,

Nygaard, & Aarsland, 2008). In our own pilot study

(Azermai et al., 2013), we discontinued antipsychotics

abruptly in 40 patients with limited relapse of behavioral

problems and re-initiation of antipsychotics. Other studies

suggest that the baseline behavioral problems are indica-

tive for the success of antipsychotic discontinuation.

Ballard et al. (2008) indicated that antipsychotic with-

drawal can be successful, especially in those patients with

low antipsychotic dose and less severe neuropsychiatric

symptoms. A recent Cochrane review presented a similar

conclusion (Declercq et al., 2013). It is estimated that

two-thirds of older people with dementia receive antipsy-

chotics unnecessary for symptoms for which antipsy-

chotics are not licensed and not effective (Banerjee,

2009). An actual attempt at discontinuation of antipsy-

chotics is needed to differentiate between people for

whom antipsychotics have no added value and those for

whom the benefits outweigh the risks.

GPs mostly shared the same barriers as nurses but,

despite their smaller involvement in the direct care of resi-

dents, GPs had a higher (nonsignificant) mean barrier sum

score. Surprisingly, we found that GPs perceive insuffi-

cient nonpharmacological treatment options as an

Figure 1. Description of barriers to antipsychotic discontinuation by the GPs and the nurses for a total of 51 and 113 residents, respec-tively. p-values calculated for 51 residents for which both nurses and GPs filled out the questionnaires, using the nonparametricWilcoxon test for two related samples.

Aging & Mental Health 351

Dow

nloa

ded

by [

UO

V U

nive

rsity

of

Ovi

edo]

at 0

2:41

15

Oct

ober

201

4

important barrier, while nurses did not share this opinion.

We did not explore the reasons for this variance, but it is

rather important. If GPs do not believe in nonpharmaco-

logical interventions, than the consult with a doctor by a

nurse when confronted with behavioral problems will

more likely result in psychotropic treatment. Therefore,

GPs rather than nurses may be the ones needing more

education and persuasion on the efficacy of nonpharmaco-

logical interventions, which may prevent all too rapid

initiation of antipsychotic treatment.

To reduce antipsychotic prescribing and related

adverse events, it is imperative that we gain more insight

in the concerns of caregivers associated with antipsy-

chotic discontinuation. More research on this topic is

needed to design practical guidance on pharmacological

and nonpharmacological alternatives, as well as education

on best practice for antipsychotic discontinuation. To

implement discontinuation programs, complex multidisci-

plinary interventions should be offered taking existing

barriers into account.

Limitations and strengths

Our study has a number of limitations. First, our study is

exploratory, an exploration of a topic not well studied.

We conducted the research in a limited selection of nurs-

ing homes, with little room for comparison of institutional

characteristics.

For this study, we developed a structured question-

naire based on a prior expert meeting, which was neces-

sary to explore the topic. The questionnaire was checked,

improved and pretested in a small selection of nurses,

GPs, and pharmacists. However, in the future a larger

study should be performed in depth to fully explore the

barriers associated with antipsychotic use and misuse in

nursing homes using qualitative methods, and a thorough

validation of questionnaires.

The sample size of our study included 113 residents

for which 13 nurses and 28 GPs filled in the question-

naires. We had a lower response rate from the GPs, which

affected our ability to compare nurses and GPs for the

same residents.

A limited sample of nurses filled out the questionnaire

for multiple residents. However, in this study we focused

on the importance of case-specific information potentially

excluding repetition of answers. Despite the researchers’

explanation to informants on the importance of case-spe-

cific information (i.e. each questionnaire represents a

selected resident), informants may still have clustered

their answers with insufficient differentiation as a poten-

tial consequence.

Conclusion

The message that antipsychotics should and can be dis-

continued without detrimental behavioral effects has not

yet reached the floor. Risks, problems, and barriers associ-

ated with the use and misuse of antipsychotics for BPSD

in nursing homes are not well investigated. Despite sev-

eral limitations, our small, exploratory study brings an

important insight into the barriers associated with discon-

tinuing antipsychotics in nursing homes, as indicated by

the main actors: nurses, and GPs. Nevertheless, there is

need for larger qualitative and quantitative studies to fully

explore this problem, and to develop multifaceted, multi-

disciplinary interventions.

Acknowledgements

We would like to thank the management, nursing staff, GPs, aswell as the residents from the nursing homes for their willingnessto participate and their cooperation. We also thank the Univer-sity of Antwerp, Department of Nursing Science, and in particu-lar Kristien Van Rompaey for her valuable contribution incollecting the data.

References

Azermai, M., Elseviers, M., Petrovic, M., van Bortel, L., &Vander Stichele, R.H. (2011). Assessment of antipsychoticprescribing in Belgian nursing homes. International Psycho-geriatric, 23, 1240–1248.

Azermai, M., Petrovic, M., Elseviers, M.M., Bourgeois, J., VanBortel, L.M., & Vander Stichele, R.H. (2012). Systematicappraisal of dementia guidelines for the management ofbehavioural and psychological symptoms. Ageing ResearchReviews, 11(1), 78–86.

Azermai, M., Petrovic, M., Engelborghs, S., Elseviers, M.M.,Van der Mussele, S., Debruyne, H., . . . Vander Stichele, R.H. (2013). The effects of abrupt antipsychotic discontinua-tion in cognitively impaired older persons: A pilot study.Aging & Mental Health, 17(1), 125–132.

Ballard, C., Creese, B., Corbett, A., & Aarsland, D. (2011).Atypical antipsychotics for the treatment of behavioral andpsychological symptoms in dementia, with a particular focuson longer term outcomes and mortality. Expert Opinion onDrug Safety, 10, 35–43.

Ballard, C., Lana, M.M., Theodoulou, M., Douglas, S.,McShane, R., Jacoby, R., . . . Juszczak, E. (2008). A rando-mised, blinded, placebo-controlled trial in dementia patientscontinuing or stopping neuroleptics (the DART-AD trial).PLoS Medicine, 5, e76.

Banerjee S. (2009). The use of antipsychotic medication for peo-ple with dementia: Time for action. A report for the Ministerof State for Care Services, Department of Health.

Briesacher, B.A., Limcangco, M.R., Simoni-Wastila, L., Doshi,J.A., Levens, S.R., Shea, D.G., & Stuart B. (2005). The qual-ity of antipsychotic drug prescribing in nursing homes.Archives of Internal Medicine, 165, 1280–1285.

Bronskill, S.E., Anderson, G.M., Sykora, K., Wodchis, W.P.,Gill, S., Shulman, K.I., & Rochon P.A. (2004). Neurolepticdrug therapy in older adults newly admitted to nursinghomes: Incidence, dose, and specialist contact. Journal ofthe American Geriatrics Society, 52, 749–755.

Brookmeyer, R., Johnson, E., Ziegler-Graham, K., & Arrighi,H.M. (2007). Forecasting the global burden of Alzheimer’sdisease. Alzheimer’s & Dementia, 3, 186–191.

Burke, A.D., & Tariot, P.N. (2009). Atypical antipsychotics inthe elderly: A review of therapeutic trends and clinical out-comes. Expert Opinion on Pharmacotherapy, 10, 2407–2414.

Cody, M., Beck, C., & Svarstad, B.L. (2002). Challenges to theuse of nonpharmacologic interventions in nursing homes.Psychiatric Services, 53, 1402–1406.

Cohen-Mansfield, J., Lipson, S., Werner, P., Billig, N., Taylor,L., & Woosley, R. (1999). Withdrawal of haloperidol, thio-ridazine, and lorazepam in the nursing home: A controlled,double-blind study. Archives of Internal Medicine, 159,1733–1740.

352 M. Azermai et al.

Dow

nloa

ded

by [

UO

V U

nive

rsity

of

Ovi

edo]

at 0

2:41

15

Oct

ober

201

4

Cornege-Blokland, E., Kleijer, B.C., Hertogh, C.M., & vanMarum, R.J. (2012). Reasons to prescribe antipsychotics forthe behavioral symptoms of dementia: A survey in Dutchnursing homes among physicians, nurses, and family care-givers. Journal of the American Medical Directors Associa-tion, 13, 80–86.

Declercq, T., Petrovic, M., Azermai, M., Van der Stichele, R.,De Sutter, A.I., van Driel, M.L., & Christiaens, T. (2013).Withdrawal versus continuation of antipsychotics for behav-ioural and psychological symptoms in older people withdementia. Cochrane Database of Systematic Reviews, 3,CD007726. doi:10.1002/14651858.CD007726.pub2

Devanand, D.P., & Schultz, S.K. (2011). Consequences of anti-psychotic medications for the dementia patient. AmericanJournal of Psychiatry, 168(8), 767–769.

European Medicine Agency (EMA). (2004). Public statement onthe safety of Olanzapine. Retrieved from http://www.bfarm.de/SharedDocs/1_Downloads/DE/Pharmakovigilanz/am-sicher-akt/emea-zyprexa-st.pdf?__blob = publicationFile

Katz, S., & Akpom, C.A. (1976). A measure of primary socio-biological functions. International Journal of Health Serv-ices, 6, 493–508.

Lyketsos, C.G., Lopez, O., Jones, B., Fitzpatrick, A.L., Breitner,J., & DeKosky, S. (2002). Prevalence of neuropsychiatricsymptoms in dementia and mild cognitive impairment:Results from the cardiovascular health study. Journal of theAmerican Medical Association, 288, 1475–1483.

Medicines and Healthcare Products Regulatory Agency (MHRA).(2008). Conventional typical antipsychotics: Increased mortal-ity in dementia. Retrieved from http://www.mhra.gov.uk/Safetyinformation/DrugSafetyUpdate/CON087985

Medicines and Healthcare products Regulatory Agency(MHRA). (2009). Antipsychotics: Use in elderly peoplewith dementia. Retrieved from http://www.mhra.gov.uk/Safetyinformation/DrugSafetyUpdate/CON088116

Powell, R., & Single, H. (1996). Methodology matters - V: Focusgroups. International Journal for Quality in Health Care,8(5), 499–504.

Reisberg, B., Borenstein, J., Salob, S.P., Ferris, S.H., Franssen,E., & Georgotas, A. (1987). Behavioral symptoms inAlzheimer’s disease: Phenomenology and treatment. Jour-nal of Clinical Psychiatry, 48 (Suppl), 9–15.

Ruths, S., Straand, J., Nygaard, H.A., & Aarsland, D. (2008).Stopping antipsychotic drug therapy in demented nursinghome patients: A randomized, placebo-controlled study –the Bergen District Nursing Home Study (BEDNURS).International Journal of Geriatric Psychiatry, 23,889–895.

Schneider, L.S., Dagerman, K.S., & Insel, P. (2005). Risk ofdeath with atypical antipsychotic drug treatment fordementia: Meta-analysis of randomized placebo-controlledtrials. Journal of the American Medical Association, 294,1934–1943.

Sterke C.S., van Beeck E.F., van der Velde N., Ziere G., PetrovicM., Looman C.W., & van der Cammen T.J. (2012). Newinsights: Dose-response relationship between psychotropicdrugs and falls: A study in nursing home residents withdementia. Journal of Clinical Pharmacology, 52(6), 947–955.

US Food and Drug Administration (FDA). (2005). Public healthadvisory: Deaths with antipsychotics in elderly patients withbehavioral disturbances. Retrieved from http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/DrugSafetyInformationforHeathcareProfessionals/PublicHealthAdvisories/UCM053171

US Food and Drug Administration (FDA). (2008). Antipsychoticsare not indicated for the treatment of dementia-related psycho-sis. Retrieved from http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm124830.htm

Aging & Mental Health 353

Dow

nloa

ded

by [

UO

V U

nive

rsity

of

Ovi

edo]

at 0

2:41

15

Oct

ober

201

4