Changes in trend of antipsychotics prescription in patients treated with cholinesterase inhibitors after warnings from Italian Medicines Agency. Results from the EPIFARM-Elderly Project.

Abstract

The objective of the study was to assess the trend of antipsychotic prescription in elderly patients taking cholinesterase inhibitors (ChEIs) from 2002 to 2008 and the changes subsequent to two main official warnings issued by the Italian Medicines Agency to restrict their use. Elderly patients aged 65-94years who received at least one prescription of ChEIs between 1 January 2002 and 31 December 2008 were selected. We used data on prescriptions from the Lombardy Region Drug Administrative Database (Italy). The first prescription of one ChEI was used as the index day to calculate the prescription of an antipsychotic. The prescription of atypical antipsychotics in patients exposed to ChEIs declined from 21.0% in 2002 to 14.6% in 2008 (OR 0.92; 95%CI:0.90, 0.94; p<0.001), while the prescribing prevalence of typicals slightly increased (OR 1.08; 95%CI:1.03, 1.13; p=0.001). In relation to the two warnings, the prevalence of patients who received a prescription of antipsychotics was significantly lower in 2005 than 2004 (23.1% vs. 28.0%; OR 0.79; 95%CI:0.73-0.86; p<0.001) and in 2007 than 2006 (19.4% vs. 23.0%; OR 0.79; 95%CI:0.73-0.86; p<0.001). After the first safety warning the prevalence of prescriptions for risperidone and olanzapine dropped significantly, and there was a significant increase for quetiapine. Haloperidol prescriptions increased, especially after the second warning. Despite regulatory warnings issued to discourage the use of antipsychotics, they are still frequently prescribed to patients taking ChEIs. Awaiting further studies to clarify their therapeutic role, physicians should prescribe antipsychotics very cautiously and only after careful risk-benefit assessment.

Full text

Page 1

Changes in trend of antipsychotics prescription in
patients treated with cholinesterase inhibitors after
warnings from Italian Medicines Agency. Results from
the EPIFARM-Elderly Project
Carlotta Franchia,⁎, Mauro Tettamantib, Alessandra Marengonic,
Francesca Bonomettic, Luca Pasinaa, Laura Cortesid, Ida Fortinoe,
Angela Bortolottie, Luca Merlinoe, Ugo Luccab,
Emma Rivab, Alessandro Nobilia
aLaboratory for Quality Assessment of Geriatric Therapies and Services, Mario Negri Institute for Pharmacological Research,
Milan, Italy
bLaboratory of Geriatric Neuropsychiatry, Mario Negri Institute for Pharmacological Research, Milan, Italy
cDepartment of Medical and Surgery Sciences, University of Brescia, Geriatric Ward, Spedali Civili, Brescia, Italy
dGeneral Practice Research, Mario Negri Institute for Pharmacological Research, Milan, Italy
eRegional Health Ministry, Lombardy Region, Milan, Italy
Received 14 October 2011; received in revised form 1 December 2011; accepted 18 December 2011
KEYWORDS
Cholinesterase inhibitors;
Antipsychotic drugs;
Treatment trends;
Elderly
Abstract
The objective of the study was to assess the trend of antipsychotic prescription in elderly pa-
tients taking cholinesterase inhibitors (ChEIs) from 2002 to 2008 and the changes subsequent
to two main official warnings issued by the Italian Medicines Agency to restrict their use. Elderly
patients aged 65–94 years who received at least one prescription of ChEIs between 1 January
2002 and 31 December 2008 were selected. We used data on prescriptions from the Lombardy
Region Drug Administrative Database (Italy). The first prescription of one ChEI was used as the
index day to calculate the prescription of an antipsychotic. The prescription of atypical antipsy-
chotics in patients exposed to ChEIs declined from 21.0% in 2002 to 14.6% in 2008 (OR 0.92;
95%CI:0.90, 0.94; pb0.001), while the prescribing prevalence of typicals slightly increased (OR
1.08; 95%CI:1.03, 1.13; p=0.001). In relation to the two warnings, the prevalence of patients
who received a prescription of antipsychotics was significantly lower in 2005 than 2004 (23.1%
vs. 28.0%; OR 0.79; 95%CI:0.73–0.86; pb0.001) and in 2007 than 2006 (19.4% vs. 23.0%; OR
⁎ Corresponding author at: Laboratory for Quality Assessment of Geriatric Therapies and Services, Mario Negri Institute for Pharmacological
Research, Via Giuseppe La Masa, 19, 20156 Milan, Italy. Tel.: +39 02 39014580; fax: +39 02 39001916.
E-mail address: carlotta.franchi@marionegri.it (C. Franchi).
0924-977X/$ - see front matter © 2012 Elsevier B.V. and ECNP. All rights reserved.
doi:10.1016/j.euroneuro.2011.12.009
www.elsevier.com/locate/euroneuro
European Neuropsychopharmacology (2012) 22, 569–577

Page 2

0.79; 95%CI:0.73–0.86; pb0.001). After the first safety warning the prevalence of prescriptions
for risperidone and olanzapine dropped significantly, and there was a significant increase for
quetiapine. Haloperidol prescriptions increased, especially after the second warning. Despite
regulatory warnings issued to discourage the use of antipsychotics, they are still frequently
prescribed to patients taking ChEIs. Awaiting further studies to clarify their therapeutic role,
physicians should prescribe antipsychotics very cautiously and only after careful risk–benefit
assessment.
© 2012 Elsevier B.V. and ECNP. All rights reserved.
1. Introduction
Non-cognitive neuropsychiatric symptoms such as aggres-
sion, agitation, depression and delusion are common in pa-
tients with dementia (Lyketsos et al., 2000, 2002; Finkel,
2001; Sink et al., 2004). Though there are no FDA-approved
treatments for dementia-related behavioral disturbances
and limited data support their favorable benefit–risk profile
in these subjects (Lee et al., 2004), antipsychotics are
widely used to reduce aggression and psychosis. Convention-
al antipsychotics had long been used to treat these symp-
toms, but from the 1990s they have been replaced by the
“atypicals”, a new class considered safer than the “typi-
cals”, particularly because of their lower rate of parkinson-
ism and tardive dyskinesia (Lopez et al., 2003). Since 2002,
however, alerts about their safety have appeared in scientif-
ic reports and from regulatory agencies, showing increased
risks of adverse cerebrovascular events (Wooltorton, 2002,
2004), cognitive decline (Ballard and Waite, 2006), and mor-
tality (Schneider et al., 2005; Schneeweiss et al., 2007), es-
pecially in elderly people and patients with dementia (FDA
Center for Drug Evaluation and Research, 2005; Fick et al.,
2003; Schneider et al., 2006; Rochon et al., 2008). In 2004,
the Italian Medicines Agency (AIFA: Agenzia Italiana del
Farmaco) publishedthefirst
(Ministero della Salute, 2004). Then, after the publication
of Food and Drug Administration (FDA) analyses demonstrat-
ing an increased risk of mortality in patients with dementia
treated with atypical antipsychotics (US Department of
Health and Human Services, 2005), in July 2005 AIFA rein-
forced its previous warning (Agenzia Italiana del Farmaco,
2005), updating it in December 2006 (second safety warning)
(Agenzia Italiana del Farmaco, 2006), and introducing
specific rules for obtaining these drugs under the Italian
National Health Service (NHS) when prescribed to treat
non-cognitive neuropsychiatric symptoms in elderly patients
with dementia (Agenzia Italiana del Farmaco, 2005) [see
Box 1]. In the same period a specific warning on the in-
creased risk of cerebrovascular events was added on the
safety labels of some antipsychotics (Pillole.org website,
2005).
Only a few studies have analyzed the pattern of use of
typical and atypical antipsychotics and the effect of regula-
tory warnings about their prescription to elderly patients
with dementia (Valiyeva et al., 2008; Kales et al., 2011),
with different results. One study (Valiyeva et al., 2008)
highlighted the limited impact of a Canadian regulatory
warning on the overall prescription rate of antipsychotics
among patients with dementia, while another (Kales et al.,
2011), showed that the use of both typical and atypical
officialsafety warning
antipsychotics for patients with dementia began to decline
before the implementation of the FDA black box warning in
2003, and the 2005 FDA advisory was associated with signif-
icant acceleration in the decline of atypical use.
The aim of this study was to assess the trend of antipsy-
chotic prescription in Italian community-dwelling elderly pa-
tients taking ChEIs from 2002 to 2008 and the changes
subsequent to two main official warnings issued by AIFA to
restrict their use because of safety concerns on the increas-
ing risk of mortality and serious adverse cerebrovascular
events.
2. Experimental procedures
2.1. Study design and data sources
Thisisadescriptiveanalysisofadministrativeclaimsfromcommunity-
dwelling elderly people aged 65 to 94 years living in Milano, Lecco and
Brescia (three cities with a total of nearly 3 million inhabitants, in
Lombardy Region, Northern Italy), who received at least one prescrip-
tion of ChEIs (Anatomical-Therapeutic-Chemical [ATC] code N06DA)
with or without prescriptions of antipsychotic agents (ATC code
N05A) between 1 January 2002 and 31 December 2008. Data were
obtained from the Lombardy Region Drug Administrative Database,
which stores all prescriptions covered by the Italian NHS, issued
to the nearly 10 million individuals living in the Region (16.2% of the
Italian population). The structure of this database, routinely updated
Box 1 Summary of AIFA warnings and regulations:
1. Physicians of authorized centers must prescribe an antipsy-
chotic drug under their own responsibility if they consider
that taking the drug is essential and the non-pharmacological
interventions didn't improve the symptoms.
2. At the beginning the dosage administered must be the lowest
clinically effective; if treatment isn't effective it should be
interrupted, otherwise prolonging the therapy must be reas-
sessed periodicallyandattheremissionofsymptomsthetreat-
ment must be gradually stopped.
3. It is not possible to administer two different kinds of antipsy-
chotics contemporaneously and with benzodiazepines.
4. Be careful in administration of antipsychotics to patients
affected by cardiovascular disease.
5. Monitor effectiveness and safety of drugs and report
promptly any adverse reaction
6. Physician must complete a treatment plan at the begin-
ning and and every 2 months to assess the risk/benefit
profile of the antipsychotic used.
7. AIFA set up a registry of patients with dementia treated
with antipsychotic drugs.
570C. Franchi et al.

Page 3

for administrative and reimbursement purposes, has been described
in detail elsewhere (Nobili et al., 2011; Percudani et al., 2004).
Briefly, with a physician's prescription the patient can obtain the
drugs free of charge from retail pharmacies. Each local pharmacy
forwards these prescriptions to the Regional Health Authority for re-
imbursement. The Health Authority electronically stores prescrip-
tions in the Regional Drug Administrative Database, which also
holds information on the age and sex of patients. All data were man-
aged and analyzed using an anonymous patient code, in accordance
with current Italian laws on privacy, after obtaining authorization
from the Lombardy Region. In Italy, the NHS provides ChEIs and anti-
psychotics free of charge only to patients with Alzheimer's disease,
who receive an ad-hoc “treatment plan” completed by neurologists,
geriatricians, or psychiatrists from one of the authorized regional
Alzheimer Evaluation Units (AEUs) (National Health Ministry, 2000).
From the beginning of treatment, the therapeutic and safety moni-
toring plan for antipsychotic agents must include a periodic (every
2 months) clinical re-assessment of each patient to confirm the per-
sistence of neuropsychiatric disturbance, and the appropriateness
and safety of continuing treatment. Furthermore most of the guide-
lines restrict the use of antipsychotics in Alzheimer's disease to
subjects with severe behavioral disturbances who showed no im-
provement after non-pharmacological interventions, and they
should be used for less than 3 months or discontinued on remission
of non-cognitive neuropsychiatric symptoms (Agenzia Italiana del
Farmaco (AIFA) website, 2005, 2006). After reports of safety issues
and international regulatory advice on the risk of adverse cerebro-
vascular events (Wooltorton, 2002, 2004), cognitive decline
(Ballard and Waite, 2006), and mortality (Schneider et al., 2005;
Schneeweiss et al., 2007), especially in elderly people and patients
with dementia (FDA Center for Drug Evaluation and Research,
2005; Fick et al., 2003; Schneider et al., 2006; Rochon et al.,
2008), AIFA published two main official warnings: the first in 2004,
and the second in 2006 (Ministero della Salute, 2004; Agenzia
Italiana del Farmaco (AIFA) website, 2005, 2006).
2.2. Study population
We selected all patients aged 65–94 years who received at least one
prescription of ChEI, and starting from the date of the first ChEI pre-
scription (index day) searched all prescriptions for antipsychotic
agents. We excluded from this analysis anyone aged 95 or older be-
cause of the high rate of institutionalization (Italian nursing homes
have different accountability methods and drugs dispensed could
not be traced). However, before these subjects were removed, peo-
ple aged 95 or older were less than 1% of the study population. The
yearly prescription prevalence of antipsychotic use was calculated
as the proportion of patients who received at least one prescription
for an antipsychotic in each study year. Individuals treated with
drugs with ATC codes N05AX09, N05AH02, N05AX08, N05AH04,
N05AX12, N05AE04 were defined as taking atypical antipsychotics,
and those treated with drugs with ATC codes N05AA, N05AB,
N05AC, N05AD,N05AF were defined as taking typicals. In accordance
with a previous study (Nobili et al., 2011) the chronic (long-term) use
of a medication was classified as the consecutive prescription of at
least four packages of a medication containing the same drug (active
substance) in each year. Prevalence data for the first five most pre-
scribed antipsychotics (quetiapine, risperidone, olanzapine, halo-
peridol, clotiapine) were calculated. As the reduction in use of one
class of medications is typically associated with the increase in
other classes, we also evaluated the trend of prescriptions of antide-
pressant drugs in patients taking ChEIs, while benzodiazepine use
was not available in this database. To illustrate the effect of the
two main recommendations and safety warnings (in 2004 and in
2006) we examined the prevalence of prescription data in three dif-
ferentyears: in2002 (before the first AIFAwarning), 2005 (soonafter
the first major AIFA warning and before implementation of the
second warning) and 2008 (2 years after the second warning). For
each year, the intensity of drug treatment (by calculating 2–3, 4–5
and 6 or more prescriptions) and the mean time interval between
each prescription (less or more than 60 days) were also recorded.
To analyze the duration of exposure to an antipsychotic after the
first prescription, yearly cohorts of incident cases (who received at
least one prescription of antipsychotic agents in any 1 year consid-
ered, but not in the previous one) were followed up. The duration
of exposure was calculated by subtracting the day (date) of the
first antipsychotic prescription from the last one: patients with
more than 1 year of exposure were censored after 365 days. Then,
considering that one pill a day might be a standard treatment,
each prescribed package (containing 30 pills) was considered enough
for 30 days' treatment. Patients were considered consecutively trea-
ted if they had enough pills to take at least a pill a day for the period
between the first and last prescriptions (up to a maximum of 1 year).
Each patient was considered only once during the years examined.
2.3. Statistical analysis
The prescription prevalence of antipsychotics over the years was
studied in two steps. In the first step the calendar year was treated
as a continuous covariate, to look for a general trend. In the second
step the models were refitted using a dummy variable for each year
(taking 2002 as the reference year) to check the contributions of
specific years. Only results from the multivariable analyses are
reported, where the contribution was studied together with the fol-
lowing covariates: age, sex and number of concomitant medications
(different active principle other than ChEIs and antipsychotics). Age
and number of concomitant medications were categorized since
their effect on the prevalence of antipsychotic use was not linear.
Age was divided into three groups: 65–74 (reference group), 75–84
and 85 or more. Number of concomitant medications was split into
four groups, by quartiles: less than 3 drugs (reference group), 3 or
4 drugs, 5 to 7 drugs, 8 or more drugs.
The duration of antipsychotic therapy was plotted using a
Kaplan-Meier curve for each calendar year. Differences between the
calendar years were analyzed using the log-rank test (again using a
dummy variable for each year),considering only incident cases for
each year, calculating the time of exposure in the 365 days from the
date of first prescription of the antipsychotic. All statistical analyses
were done using Stata v 11.0, StataCorp. Since in analyses of preva-
lencepatientscouldbecountedmorethanonce,aclusteredsandwich
estimatorwasusedtocorrectfornon-independenceofdata.Ap-value
b0.05 was considered significant.
3. Results
3.1. Study population
During the study, nearly 600,000 elderly people aged 65–94 years
were available for the analysis: 10,826 distinct patients had at
least one prescription of a ChEI, and 2341 (21.6%) of these re-
ceived a prescription for at least one antipsychotic; there were
1519 women (64.9%). A total of 1822 subjects (77.8%) had only
one antipsychotic, 519 (22.2%) received two or more different an-
tipsychotics. Table 1 summarizes the main characteristics of the
patients included in the analyses, at baseline (year 2002), in the
year after the first safety warning (year 2005), and in the last
availableyearafterthesecondwarning(year2008).Table2shows
the distribution of patients in relation to the number of antipsy-
chotic prescriptions and number of days (mean time) between
each prescription. The majority received more than one prescrip-
tioninall3 years(77.2%in2002,81.6%in2005and65.2%in2008).
In 2005, after the first regulatory warnings, 36.4% were
571 Antipsychotic prescriptions in patients treated with cholinesterase inhibitors

Page 4

continuously (mean time between prescriptions b60 days) and in-
tensively (≥6 prescriptions) treated with antipsychotics, a much
higher percentage than in 2002 (19.7%). This had fallen to 24.5%
during 2008. Meanwhile, the percentage of patients who received
only one prescription of antipsychotics rose from 22.8% in 2002 to
34.8% in 2008.
3.2. Prescription of antipsychotics in patients taking
ChEIs
Fig. 1 shows the prevalence of patients aged 65–94 years treated
withChEIswhoreceivedaprescriptionofatleastoneantipsychot-
icagent.Theoverallprevalencefellfrom24.2%in2002to18.1%in
2008(yearOR0.93;95%CI:0.92–0.95;pb0.001).The prescription
prevalence of atypical antipsychotics in patients exposed to ChEIs
declinedfrom21.0%in2002to14.6%in2008(yearOR0.92;95%CI:
0.90–0.94; pb0.001), while that of typicals rose slightly (year OR
1.08; 95%CI: 1.03–1.13; p=0.001). Atypicals were prescribed on
average four times more than typicals. In relation to two warn-
ings, there was a statistically significant drop in the prevalence
of patients who received a prescription of antipsychotics in 2005
compared to 2004 (23.1% vs 28.0%; OR 0.79; 95%CI: 0.73–0.86;
pb0.001) and 2007 compared to 2006 (19.4% vs 23.0%; OR 0.79;
95%CI: 0.73–0.86; pb0.001). No difference was observed be-
tween age groups in the overall prevalence of treatment, or for
atypicals and typicals. Men were treated slightly more than wom-
en with atypicals (OR 1.12; 95%CI: 1.01–1.24; p=0.026), but no
difference emerged for typicals. Patients receiving eight or more
drugs in addition to ChEIs and antipsychotics had a higher proba-
bility of taking an antipsychotic (OR 1.38; 95%CI: 1.23–1.55;
pb0.001). Results were similar for atypicals (OR 1.31; 95%CI:
1.16–1.48; pb0.001) and typicals (OR 1.84; 95%CI: 1.43–2.37;
pb0.001).Incoincidencewiththereductioninuseofantipsychot-
ic drugs overtime we noticed an increase in antidepressant use in
patientstakingChEIsfrom2002(39.5%)until2006(53.1%),whena
plateau was reached. If we evaluate the data of antidepressant
use in demented patients in the light of the AIFA warnings, we
didn't find a significant increase either in 2005 compared to 2004
or in 2007 compared to 2006.
3.3. The first five most prescribed antipsychotics
The five antipsychotics most prescribed to patients treated with
ChEIs who received at least one antipsychotic were quetiapine,
risperidone, olanzapine, haloperidol and clotiapine (Fig. 2). The
prevalence of patients taking at least one prescription of que-
tiapine and haloperidol increased over the years, respectively
Table 1
baseline (year 2002), in the years after the first safety
warning, and in the last years available after the second
safety warning.
Main characteristics of patients analyzed at
Years2002 2005 2008
Patients taking ChEIs N=2590
(%)
N=4156
(%)
N=4621
(%)
Sex
Men850
(32.8)
1740
(67.2)
1312
(31.6)
2844
(68.4)
1489
(32.2)
3132
(67.8)
Women
Age groups (years)
65–69
70–74
201 (7.8)
444
(17.2)
835
(32.2)
694
(26.8)
332
(12.8)
78 (3.0)
6 (0.2)
259 (6.2)
558
(13.4)
1194
(28.7)
1362
(32.8)
599
(14.4)
170 (4.1)
14 (0.3)
226 (4.9)
549
(11.9)
1182
(25.6)
1502
(32.5)
947
(20.5)
190 (4.1)
25 (0.5)
75–79
80–84
85–89
90–94
N95
PatientstakingChEIsandAPS
Overall 627
(24.2)
197
(23.2)
430
(24.7)
78.0±6.1 79.1±5.8 78.0±5.8
780
(18.8)
263
(20.0)
517
(18.2)
837
(18.1)
273
(18.3)
564
(18.0)
Men
Women
Mean age
Mean number of drugs*
Overall
Men
Women
Mean number of chronic drugs (at least four packs)a
Overall
Male
Female
6.5±4.0
6.9±4.1
6.3±3.9
6.5±4.0
6.9±4.0
6.4±3.9
7.2±4.2
7.7±4.3
7.0±4.2
3.7±2.2
4.0±2.4
3.6±2.2
4.2±2.5
4.5±2.6
4.1±2.5
4.73±2.7
5.1±2.9
4.6±2.7
aExcluding cholinesterase inhibitors (ChEIs) and antipsychotic
agents (APS).
Table 2
by number of prescriptions for antipsychotics received and
mean time between prescriptions.
Distribution of patients (number and percentage)
Mean
time
(days)
Number of prescriptions Total
12–34–5≥6
2002
0-5973
22.8%
24
7.5%
51
15.9%
75
23.4%
13
4.0%
28
8.7%
41
12.8%
63
19.7%
68
21.2%
131
40.9%
100
31.2%
147
45.9%
320
100%
≥60
Total73
22.8%
2005
0–59 57
18.4%
27
8.5%
24
7.6%
51
16.1%
14
4.4%
14
4.4%
28
0.9%
115
36.4%
65
20.6%
180
56.9%
156
49.4%
103
32.6%
316
100%
≥60
Total57
18.4%
2008
0–59145
34.8%
72
17.3%
45
10.8%
117
28.0%
34
8.1%
15
3.6%
49
11.7%
102
24.5%
4
0.96%
106
25.4%
208
49.9%
64
15.3%
417
100%
≥60
Total 145
34.8%
572 C. Franchi et al.

Page 5

from 4.9% in 2002 to 12.1% in 2008 (OR 1.16; 95%CI: 1.13–1.19;
pb0.001), and from 1.0% in 2002 to 3.2% in 2008 (OR 1.23;
95%CI: 1.16–1.30; pb0.001). The prevalence of patients taking
at least one prescription of risperidone dropped from 9.7% in
2002 to 0.8% in 2008 (OR 0.64; 95%CI: 0.61–0.67; pb0.001)
and for olanzapine from 7.4% in 2002 to 1.6% in 2008 (OR 0.71;
95%CI: 0.69–0.74; pb0.001). Clotiapine did not change much
over the study period. Comparing the prescription prevalence
in the year after the safety warning with the following year, pre-
scriptions of quetiapine increased in 2005 in comparison to 2004
(OR 1.21; 95%CI: 1.08–1.35; pb0.001), while after the second
safety warning there was a significant decrease (OR 0.78;
95%CI: 0.71–0.86; pb0.001). For risperidone, the prevalence
of patients treated in the year after the two safety warnings de-
creased, comparing 2005 and 2004 (OR 0.43; 95%CI: 0.36–0.53;
pb0.001) and 2007 and 2006 (OR 0.64; 95%CI: 0.48–0.86;
pb0.004). For olanzapine there was a significant decrease in
the prevalence of treated patients only comparing 2005 and
2004 (OR 0.43; 95%CI: 0.36–0.50; pb0.001). No other substan-
tial changes were observed for the other antipsychotics included
in the analysis comparing the years of the safety warning and
the year after.
3.4. Duration of the treatment
To investigate whether the two safety warnings influenced the
duration of exposure to antipsychotics, 2021 incident consecu-
tively treated cases (86.2% of overall study population) were an-
alyzed. Fig. 3 shows the one-year Kaplan–Meier survival curves
for the duration of exposure to antipsychotics in each year of
the study. There were significant differences between years
(pb0.0001 log-rank test): duration of treatment got progres-
sively longer from 2002 to 2005, but it decreased afterwards.
4. Discussion
From 2002 to 2008 more than 20% of ChEI treated
community-dwelling elderly people (aged 65–94 years) re-
ceived a prescription of at least one antipsychotic. In the
same period the overall prevalence of prescription of anti-
psychotics fell significantly from 24% to 18%. This fall was ob-
served for atypicals, whose prevalence declined from 21% in
2002 to 15% in 2008, while typicals rose slightly from 3% in
2002 to nearly 5% in 2008. During the same period a shift to-
wards quetiapine was noted with a significant increase from
2003 to 2006, when it reached the peak of 12%, but after the
second safety warning the prevalence of use declined.
Among atypicals the prevalence of prescription of risperi-
done and olanzapine fell significantly, while clotiapine did
not show any real change. Haloperidol, a typical drug, in-
creased from 1% to 3%. There was a significant decrease in
the prevalence of patients treated with ChEIs who received
a prescription of antipsychotics in the years immediately
after the two warnings issued by AIFA. After the first warn-
ing, the prevalence of prescription for risperidone and olan-
zapine dropped significantly and this decrease continued,
though less evidently after the second warning. For quetia-
pine, after the first warning there was a significant increase
in prescription prevalence, which began to decrease after
the second warning. Haloperidol showed an increase, espe-
cially after the second warning. The two warnings also had
an effect on duration of antipsychotic use, which progres-
sively declined from 2006. To our knowledge this is the
first study to analyze the prescription of typical and atypical
antipsychotic agents in community-dwelling Italian patients
treated with ChEIs and how two national regulatory warnings
influenced the patterns. A recent Italian study (Chiabrando
1st safety warning2nd safety warning
Figure 1
agent (APS), overall and by sex and class of antipsychotics.
Prevalence of patients aged 65–94 years treated with cholinesterase inhibitors receiving a prescription of an antipsychotic
573Antipsychotic prescriptions in patients treated with cholinesterase inhibitors

Page 6

et al., 2010) analyzed the prescriptive profile of atypical an-
tipsychoticsin392patientswithdementiaselectedfromthree
main AEUs at the University Hospital in Ferrara. Over the
3 year analysis, the proportion of patients receiving risperi-
done or olanzapine decreased from the first to the second
year, and was almost constant during the third year. The
numbers receiving quetiapine increased during the second
year and then fell during the third. After the warning in
2004, there was a foreseeable increase in prescriptions of
quetiapine to the detriment of risperidone and olanzapine.
Another study (Valiyeva et al., 2008), using prescription
drug claims data from the Ontario Drug Benefit database,
evaluated the impact of three warnings of serious adverse
events related to the use of atypical antipsychotics among
elderly patients with dementia, and found that each warning
had only a limited effect on prescription rates. However,
the overall prescription rate of antipsychotics rose between
thefirstwarningandtheendofthestudyperiod.Theseresults
contrast with ours, which found an overall tendency to a
decrease in the prescription of antipsychotics (especially atyp-
icals), and a significant reduction when the years before and
after the warning were compared, with a 21% reduction in
the risk of co-prescription in both cases. The impact of the US
Food and Drug Administration (FDA) advisory and subsequent
black boxwarning,inApril 2005,ontheuseofatypicalantipsy-
chotic medications, was assessed in a quasi-experimental,
interrupted time-series analysis of IMS Health's National
Disease and Therapeutic Index from 2003 to 2008 (Dorsey et
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
50
100150200250 300350
Days
Subjects, %
2005
2006
2004
2003
2007
2002
2008
Figure 3 One-year Kaplan–Meier survival curves for the probability of continuing the antipsychotic treatment, by subcohort.
0
2
4
6
8
10
12
14
200220032004 2005
Years
2006 20072008
Subjects, %
Quetiapine
Risperidone
Olanzapine
Haloperidol
Clotiapine
1st safety warning2nd safety warning
Figure 2
five antipsychotic agents most prescribed during the study period.
Prevalence of patients aged 65–94 years treated with cholinesterase inhibitors receiving a prescription of one of the first
574 C. Franchi et al.

Page 7

al., 2010). The authors found that the FDA warning was associ-
ated with a decrease in the use of atypical antipsychotics. The
decline begun within a month after the advisory and continued
at least through the end of 2008, and was particularly pro-
nounced for patients with dementia. Another time-series ana-
lyses using data from the US Veterans Affairs Registries
examined the effect of the 2005 US FDA warning on the use of
typical and atypical antipsychotics in outpatients with demen-
tia during three periods: a no-warning period (April 1999–June
2003), an early warning period (July 2003–March 2005), and a
post-black-box warning period (April 2005–September 2007)
(Kales et al., 2011). This study found that the overall use of an-
tipsychotics began to decline during the no-warning period,
and continued following the black box warning, with a signifi-
cant difference between the early and black box warning pe-
riods, while use of atypicals increased in the first period and
started to decline during the second, declining more sharply
during the last period.
Although these studies indicate that safety warnings do
affect the use of at risk drugs, the impact of these warnings
on the appropriateness and safe use of antipsychotics is not
definite (Lyketsos et al., 2002; Fick et al., 2003; Valiyeva
et al., 2008). We observed an overall decline in the use of
antipsychotics. The prescription pattern was different for
typicals, which showed a slight increase in prescription,
more evident after the second safety warning. For atypicals,
as shown in other studies, there was an overall decline in
prescriptions, which rose slightly in the year after each safe-
ty warning. However, considering the single atypical drugs,
we observed a significant shift from risperidone and olanza-
pine, the drugs most involved in safety warnings, to quetia-
pine, as in the Veterans Affairs study (Kales et al., 2011),
and to haloperidol. Although we cannot directly compare
the effects of specific safety warnings on the prescribing
rate of antipsychotics in different countries, the results
obtained by Lombardy Region could be explained by the
availability and efficiency of a thorough system for drug
monitoring, of specific disease registries, and of specific di-
agnostic and disease monitoring unit (the EAUs) at the level
of each Local Health Unit (LHU). The continuous monitoring
by the LHU on appropriate prescription of antipsychotics to
patients with dementia and the reinforcement of AIFA warn-
ings on the safe use of these medications, may be an effi-
cient way to implement strategies for a more appropriate
use of these drugs. It is difficult to ascertain a clear cause–
effect relationship between a warning and any change in
prescribing patterns, especially when in the same period
many other reports on the limited benefit and the increasing
risk of death and serious cerebrovascular adverse events of
these drug in dementia patients with behavioral distur-
bances were published or sent to physicians through differ-
ent health systems and by pharmaceutical companies. A
possible explanation to this decreasing use in antipsychotic
drugs could be the lack of physicians' belief on their “real”
efficacy in the light of low clinical benefits and the potential
high risks underlined by international regulatory warnings.
Starting from 2002, the first reports appeared in the medical
and pharmacological literature with epidemiological studies
on the safety risk of these drugs, particularly for risperidone
and olanzapine, addressing the need to restrict these medica-
tionstoselectedcases.Moreover,despiteclinicalresemblance
to symptoms (delusions, hallucinations, sleep disturbances) of
other psychiatric diseases, the mechanisms underlying behav-
ioral disturbances in dementia are still not well known and
the use of these drugs appears not firmly grounded. However,
theAIFA'sregistryof patients withdementia treatedwith anti-
psychotics and the need for a therapeutic plan for prescribing
these drugs certainly helped reinforce the safety warning. In-
stead we can't relate a general reduction of prescribing trend
for olanzapine and risperidone and an inverted trend for que-
tiapine and haloperidol with a change in market scenarios or
in costs, because in Italy risperidone and clotiapine have
been available as generic formulations only since 2007, while
olanzapine will be available as equivalent drug in 2011 and
quetiapine in 2012. (Assogenerici web site, 2011). However
we cannot exclude the possibility that changes in marketing
policiesofpharmaceuticalcompaniescouldhavehadaneffect
on antipsychotic prescription. Another effect was the shorter
durationoftherapyafterthefirstwarning,whichprogressively
reduced the number of patients treated chronically with these
medications. We did not find any age-related difference in an-
tipsychotic prescriptions: this probably means that the oldest
patients received the same amounts of antipsychotics despite
their higher risk of adverse drug reactions and the increasing
risk of serious cerebrovascular adverse events and mortality
in older patients with dementia (Schneider et al., 2005;
Schneeweiss et al., 2007; Rochon et al., 2008). We also found
arelationbetweenthenumbersofco-prescribeddrugsandan-
tipsychotic use. Therefore, since polypharmacy is generally
considered a proxy measure of multiple diseases (Hilmer and
Gnjidic, 2009) the use of antipsychotics in these patients
might imply a further risk for drug–drug interactions and ad-
verse events. Although we know that reductions in the use of
one class of medications is typically associated with increases
in other classes, the overall tendency to increasing use in anti-
depressants is probably linked to a general increase of antide-
pressantuseintheLombardyregion,asshownbyParabiaghiet
al. (Parabiaghi et al., 2011).
The results for single drugs and the positive patterns for
haloperidol and quetiapine might be explained because the
first warning focused only on risperidone and olanzapine.
This might have influenced clinicians' beliefs in safer profiles
of these drugs, shifting prescription to haloperidol and que-
tiapine, without any real scientific evidence of their better
benefit–risk profiles (Wang et al., 2005; BMJ group, 2007).
Although all guidelines indicate that non-pharmacological
therapy should be considered first-line treatment in these
patients, the lack of well-designed studies documenting
the efficacy of behavioral or alternative pharmacologic ap-
proaches limit the clinician's confidence in these therapies.
Thus, antipsychotics are still in clinical practice the treat-
ment of choice in the majority of cases, despite their limited
efficacy, the seriousness of their adverse events, and in
many patients their “off-label” use. Although we did not in-
vestigate the presence of behavioral disturbances, the pre-
scription of these drugs to patients treated with ChEIs is
presumably strongly related to these disturbances. All regu-
latory agencies have now urged physicians to apply a more
restrictive and critical approach to assess the appropriate-
ness and the benefit–risk ratio before prescribing these
drugs for dementia patients. The choice should consider
the approved indication, the type of behavioral disturbance
and comorbidity, the patient's risks or contraindications, the
presence of cardiovascular diseases, and co-prescription of
575 Antipsychotic prescriptions in patients treated with cholinesterase inhibitors

Page 8

other psychiatric drugs, and should be discussed with the
caregiver or family. Although evidence from randomized
controlled clinical trials indicates a slight advantage in
terms of efficacy for atypicals, recent meta-analyses
(Ballard and Waite, 2006; Schneider et al., 2005), safety
warnings, and the results of the CATIE-AD trial (Schneider
et al., 2006) showed that the overprescription and the shift
towards newer antipsychotics in patients with dementia
and behavioral disturbances lack any basis of clinical evi-
dence, in addition to results of Wang et al. who underlined
that the use of typical antipsychotics is probably associated
with similar risks as atypicals (Wang et al., 2005).
5. Limitations
This study has several limitations inherent to the use of ad-
ministrative data: the inability to determine important
clinico-functional, cognitive and behavioral characteristics
of selected patients, in order to identify those with demen-
tia and assess the appropriateness of use of antipsychotic
agents. In terms of drug use, prescriptions filled are an im-
precise measure of actual drug exposure and may not reflect
daily use, especially when only one prescription is recorded.
Wehave noinformationon drugs not reimbursedbythe Italian
NHS, and over-the-counter medicines, so we cannot evaluate
co-prescription or a shift to benzodiazepines. Furthermore,
our selection criterion was the presence of at least one pre-
scription of a ChEI to identify patients with dementia. Thus
we may have underestimated the use of antipsychotics in the
whole population of demented patients. On the same ground
itisnot possiblefor ustosay if the use of cholinesterase inhib-
itors delays or has an impact on the prescription rate of anti-
psychotics. Finally, our sample refers only to patients living
in the Lombardy Region, which is not fully representative of
other Italian regions. However, despite these drawbacks our
results are in line with other studies both in terms of the over-
all prescription patterns for the period and for the effect of
safety warnings on antipsychotic prescription.
6. Conclusions
In conclusion, antipsychotics are still widely used for elderly
patients taking ChEIs; two safety warnings only slightly re-
duced the overall use of these drugs. Since physicians struggle
every day withthe treatment of neuropsychiatric symptomsin
dementia due to the lack of efficacious drugs, further studies
are urgently needed to clarify their safety profile.
Role of the funding source
This study was supported by grants from the Regional Health
Ministry of the Lombardy region (Progetto ‘EPIFARM-epidemiologia
dei farmaci’)
Contributors
Franchi designed the study, analyzed the data and wrote the manu-
script. Tettamanti undertook the statistical analysis and revised the
manuscript. Nobili designed the study and revised the manuscript.
Marengoni revised the paper before submission. All authors contrib-
uted to and approved the final manuscript.
Conflict of interest
The authors have no financial interest to disclose.
Acknowledgments
This study was supported by grants from the Regional Health Ministry
of the Lombardy Region (Progetto ‘EPIFARM-epidemiologia dei
farmaci’).
References
Agenzia Italiana del Farmaco (AIFA) website, 2005. http://www.
agenziafarmaco.gov.it/it/content/il-trattamento-farmacologico-
dei-disturbi-psicotici-pazienti-affetti-da-demenza. Accessed July
8, 2011.
Agenzia Italiana del Farmaco (AIFA) website, 2006. http://www.
agenziafarmaco.gov.it/it/content/trattamento-farmacologico-
dei-disturbi-psicotici-nei-pazienti-affetti-da-demenza-
aggiornamen. Accessed July 8, 2011.
Assogenerici web site, 2011. http://www.assogenerici.org/2011/
centrostudi5.asp?s=2&p=5&modulo=scadenze. Accessed November
11, 2011.
Ballard, C., Waite, J., 2006. The effectiveness of atypical antipsy-
choticsforthetreatmentofaggressionandpsychosisinAlzheimer's
disease. Cochrane Database Syst. Rev. (1), CD003476.
BMJ group, 2007. How safe are antipsychotics in dementia? Drug
Ther. Bull. 45, 81–85.
Chiabrando, G., Bianchi, S., Poluzzi, E., Montanaro, N., Scanavacca,
P., 2010.Profile ofatypical-antipsychoticsuseinpatients affected
by dementia in the University Hospital of Ferrara. Eur. J. Clin.
Pharmacol. 66, 661–669.
Dorsey, E.R., Rabbani, A., Gallagher, S.A., Conti, R.M., Alexander,
G.C., 2010. Impact of FDA black box advisory on antipsychotic
medication use. Arch. Intern. Med. 170 (1), 96–103.
FDA Center for Drug Evaluation and Research, 2005. FDA informa-
tion for healthcare professionals on antipsychotics [online]
Available from URL: http://www.fda.gov/Drugs/DrugSafety/
PostmarketDrugSafetyInformationforPatientsandProviders/
ucm124830.htm.
Fick, D.M., Cooper, J.W., Wade, W.E., Waller, J.L., Maclean, J.R.,
Beers, M.H., 2003. Updating the Beers criteria for potentially
inappropriate medication use in older adults. Arch. Intern. Med.
163, 2716–2724.
Finkel,S.I.,2001.Behavioralandpsychologicalsymptomsofdementia:
a current focus for clinicians, researchers, and caregivers. J. Clin.
Psychiatry 62, 3–6.
Hilmer, S.N., Gnjidic, D., 2009. The effects of polypharmacy in
older adults. Clin. Pharmacol. Ther. 85, 86–98.
Kales, H.C., Zivin, K., Kim, H.M., Valenstein, M., Chiang, C., Ignacio,
R.V., Ganoczy, D., Cunningham, F., Schneider, L.S., Blow, F.C.,
2011. Trends in antipsychotic use in dementia 1999–2007. Arch.
Gen. Psychiatry 68 (2), 190–197.
Lee, P.E., Gill, S.S., Freedman, M., Bronskill, S.E., Hillmer, M.P.,
Rochon, P.A., 2004. Atypical antipsychotic drugs in the treatment
ofbehaviouralandpsychologicalsymptomsofdementia:systematic
review. BMJ 329, 75. doi:10.1136/bmj.38125.465579.55.
Lopez, O.L., Becker, J.T., Sweet, R.A., Klunk, W., Kaufer, D.I.,
Saxton, J., DeKosky, S.T., 2003. Patterns of change in the
treatment of psychiatric symptoms in patients with probable
Alzheimer's disease from 1983 to 2000. J. Neuropsychiatry
Clin. Neurosci. 15 (1), 67–73.
576C. Franchi et al.

Page 9

Lyketsos, C.G., Steinberg, M., Tschanz, J.T., Norton, M.C., Steffens,
D.C., Breitner, J.C., 2000. Mental and behavioral disturbances in
dementia: findings from the Cache County Study on Memory in
Aging. Am. J. Psychiatry 157, 708–714.
Lyketsos, C.G., Lopez, O., Jones, B., Fitzpatrick, A.L., Breitner, J.,
DeKosky, S., 2002. Prevalence of neuropsychiatric symptoms
in dementia and mild cognitive impairment: results from the
Cardiovascular Health Study. JAMA 288, 1475–1483.
Ministero della Salute, 2004. Chiarimenti riguardanti la prescrivibilità
dei farmaci antipsicotici atipici da parte delle UVA. Farmacovigi-
lanza News 7–8, 4.
National Health Ministry, 2000. Protocollo di monitoraggio dei piani
di trattamento farmacologico per la Malattia di Alzheimer. D.M.
del 20/07/2000, S.O. alla G.U. n° 204 del 01/09/2000, e D.M. del
14/03/2001, G.U. n° 86 del 12/04/2001Supplemento ordinario
alla G. U. n° 204 del 1° settembre 2000 - Serie Generale.
Nobili,A.,Franchi,C.,Pasina,L.,Tettamanti,M.,Baviera,M.,Monesi,
L., Roncaglioni, C., Riva, E., Lucca, U., Bortolotti, A., Fortino, I.,
Merlino, L., 2011. Drug utilization and polypharmacy in an Italian
elderly population: the EPIFARM-Elderly Project. Pharmacoepide-
miol. Drug Saf. 20, 488–496.
Parabiaghi,A.,Franchi,C., Tettamanti, M., Barbato,A.,D'Avanzo,B.,
Fortino, I.,Bortolotti, A., Merlino,L., Nobili, A.,2011. Antidepres-
santsutilizationamongelderlyinLombardyfrom2000to2007:dis-
pensing trends and appropriateness. Eur. J. Clin. Pharmacol. 67
(10), 1077–1083.
Percudani, M., Barbui, C., Fortino, I., Petrovich, L., 2004. Antide-
pressant drug use in Lombardy, Italy: a population-based study.
J. Affect. Disord. 83, 169–175.
Pillole.org website, 2005. http://www.pillole.org/public/aspnuke/
news.asp?id=2162&sid=412903068. Accessed July 8, 2011.
Rochon, P.A., Normand, S.L., Gomes, T., Gill, S.S., Anderson, G.M.,
Melo, M., Sykora, K., Lipscombe, L., Bell, C.M., Gurwitz, J.H.,
2008. Antipsychotic therapy and short-term serious events in
older adults with dementia. Arch. Intern. Med. 168 (10),
1090–1096.
Schneeweiss, S., Setoguchi, S., Brookhart, A., Dormuth, C., Wang,
P.S., 2007. Risk of death associated with the use of conventional
versus atypical antipsychotic drugs among elderly patients. CMAJ
176, 627–632.
Schneider, L.S., Dagerman, K.S., Insel, P., 2005. Risk of death with
atypical antipsychotic drug treatment for dementia: metaanalysis
of randomized placebo-controlled trials. JAMA 294, 1934–1943.
Schneider,L.S.,Tariot,P.N.,Dagerman,K.S.,Davis,S.M.,Hsiao,J.K.,
Ismail, M.S., Lebowitz, B.D., Lyketsos, C.G., Ryan, J.M., Stroup,
T.S., Sultzer, D.L., Weintraub, D., Lieberman, J.A., CATIE-AD
Study Group, 2006. Effectiveness of atypical antipsychotics
drugs in patients with Alzheimer's disease. N. Engl. J. Med. 355
(15), 1525–1538.
Sink, K.M., Covinsky, K.E., Newcomer, R., Yaffe, K., 2004. Ethnic
differences in the prevalence and pattern of dementia-related
behaviors. J. Am. Geriatr. Soc. 52, 1277–1283.
US Department of Health and Human Services, 2005. Public health
advisory: deaths with antipsychotics in elderly patients with be-
havioral disturbances. Food and Drug Administration Web site
http://www.fda.gov/Drugs/DrugSafety/
PostmarketDrugSafetyInformationforPatientsandProviders/
DrugSafetyInformationforHeathcareProfessionals/
PublicHealthAdvisories/ucm053171.htm. Accessed July 1, 2011.
Valiyeva, E., Herrmann, N., Rochon, P.A., Gill, S.S., Anderson, G.M.,
2008. Effect of regulatory warnings on antipsychotic prescription
rates among elderly patients with dementia: a population-based
time-series analysis. CMAJ 179 (5), 438–446.
Wang, P.S., Schneeweiss, S., Avorn, J., Fischer, M.A., Mogun, H.,
Solomon, D.H., Brookhart, M.A., 2005. Risk of death in elderly
users of conventional vs. atypical antipsychotic medications. N.
Engl. J. Med. 353 (22), 2335–2341.
Wooltorton, E., 2002. Risperidone (Risperdal): increased rate of
cerebrovascular events in dementia trials. CMAJ 167, 1269–1270.
Wooltorton, E., 2004. Olanzapine (Zyprexa): increased incidence of
cerebrovascular events in dementia trials. CMAJ 170, 1395.
577Antipsychotic prescriptions in patients treated with cholinesterase inhibitors