Nursing Home Residents Dying With Dementia in Flanders, Belgium: A Nationwide Postmortem Study on Clinical Characteristics and Quality of Dying

Abstract

Objectives:
There is a lack of large-scale, nationwide data describing clinical characteristics and quality of dying of nursing home residents dying with dementia. We set out to investigate quality of end-of-life care and quality of dying of nursing home residents with dementia in Flanders, Belgium.

Design/setting/participants:
To obtain representativity, we conducted a postmortem study (2010) using random cluster sampling. In selected nursing homes, all deceased residents with dementia in a period of 3 months were reported. For each case, a structured questionnaire was filled in by the nurse most involved in care, the family physician, and the nursing home administrator. We used the Cognitive Performance Scale and Global Deterioration Scale to assess dementia. Main outcome measures were health status, clinical complications, symptoms at the end of life, and quality of dying.

Measurements:
Health status, clinical complications, symptoms at the end of life, and quality of dying.

Results:
We identified 198 deceased residents with dementia in 69 nursing homes (58% response rate). Age distribution was the same as all deceased residents with dementia in Flanders, 2010. Fifty-four percent had advanced dementia. In the last month of life, 95.5% had 1 or more sentinel events (eg, eating/drinking problems, febrile episodes, or pneumonia); most frequently reported symptoms were pain, fear, anxiety, agitation, and resistance to care. In the last week, difficulty swallowing and pain were reported most frequently. Pressure sores were present in 26.9%, incontinence in 89.2%, and cachexia in 45.8%. Physical restraints were used in 21.4% of cases, and 10.0% died outside the home. Comparing stages of dementia revealed few differences between groups regarding clinical complications, symptoms, or quality of dying.

Conclusion:
Regardless of the dementia stage, many nursing home residents develop serious clinical complications and symptoms in the last phase of life, posing major challenges to the provision of optimum end-of-life care.

Full text

Original Study
Nursing Home Residents Dying With Dementia in Flanders, Belgium:
A Nationwide Postmortem Study on Clinical Characteristics and Quality of Dying
An Vandervoort MA
a
,
*, Lieve Van den Block PhD
a
,
b
, Jenny T. van der Steen PhD
c
, Ladislav Volicer PhD
d
,
Robert Vander Stichele PhD
a
,
e
, Dirk Houttekier PhD
a
, Luc Deliens PhD
a
,
f
a
End-of-Life Care Research Group, Ghent University & Vrije Universiteit Brussel, Brussels, Belgium
b
Department of Family Medicine, Vrije Universiteit Brussel, Brussels, Belgium
c
EMGO Institute for Health and Care Research and Expertise Center for Palliative Care, Department of General Practice & Elderly Care Medicine, VU University Medical Center,
Amsterdam, Netherlands
d
University of South Florida, School of Aging Studies, Tampa, FL
e
Heymans Institute of Pharmacology, Ghent University, Ghent, Belgium
f
EMGO Institute for Health and Care Research, Expertise Center for Palliative Care, Department of Public and Occupational Health, VU University Medical Center, Amsterdam,
Netherlands
Keywords:
End-of-life care
nursing homes
dementia
quality of dying
abstract
Objectives: There is a lack of large-scale, nationwide data describing clinical characteristics and quality of
dying of nursing home residents dying with dementia. We set out to investigate quality of end-of-life
care and quality of dying of nursing home residents with dementia in Flanders, Belgium.
Design/Setting/Participants: To obtain representativity, we conducted a postmortem study (2010) using
random cluster sampling. In selected nursing homes, all deceased residents with dementia in a period of
3 months were reported. For each case, a structured questionnaire was filled in by the nurse most
involved in care, the family physician, and the nursing home administrator. We used the Cognitive
Performance Scale and Global Deterioration Scale to assess dementia. Main outcome measures were
health status, clinical complications, symptoms at the end of life, and quality of dying.
Measurements: Health status, clinical complications, symptoms at the end of life, and quality of dying.
Results: We identified 198 deceased residents with dementia in 69 nursing homes (58% response rate). Age
distribution was the same as all deceased residents with dementia in Flanders, 2010. Fifty-four percent had
advanced dementia. In the last month of life, 95.5% had 1 or more sentinel events (eg, eating/drinking
problems, febrile episodes, or pneumonia); most frequently reported symptoms were pain, fear, anxiety,
agitation, and resistance to care. In the last week, difficulty swallowing and pain were reported most
frequently. Pressure sores were present in 26.9%, incontinence in 89.2%, and cachexia in 45.8%. Physical
restraints were used in 21.4% of cases, and 10.0% died outside the home. Comparing stages of dementia
revealed few differences between groups regarding clinical complications, symptoms, or quality of dying.
Conclusion: Regardless of the dementia stage, many nursing home residents develop serious clinical
complications and symptoms in the last phase of life, posing major challenges to the provision of
optimum end-of-life care.
Copyright Ó2013 - American Medical Directors Association, Inc.
An increasing number of elderly people will suffer from dementia
in Europe, from about 6 million in 2010 to about 14 million in 2050.
1
Epidemiological evidence shows dementia as a risk factor for nursing
home admission.
2,3
In Belgium, 43% of people with dementia live in
nursing homes, rising to 76% of those with advanced dementia.
4
Several barriers to providing high-quality end-of-life care for
people with dementia have been described; for example, dementia
has a relatively unpredictable course compared with cancer,
5,6
and
most patients with dementia are not able to communicate their
preferences, leaving them at increased risk of prolonged suffering at
the end of life.
7
Hence, management of problems in the final stages
poses challenges for nursing homes in particular because although
There are no conflicts of interest.
This study was supported by a major grant from Vrije Universiteit Brussel (GOA
HW, VUB 2007) and is part of the “Dying Well With Dementia”study, the Fund for
Scientific Research in Flanders, Belgium (postdoctoral grant to L. Van den Block). The
Vrije Universiteit Brusseland the Fund for Scientific Research in Flanders did not have
any role in the design and conduct of the study; collection, management, analysis, and
interpretation of the data; or in preparation, review, or approval of this manuscript.
In Belgium, the study protocol and anonymity procedures were approved by the
ethical review boardof the University Hospital of the Vrije Universiteit Brussel. Patient
anonymity was preserved and physician confidentiality maintained through the
registration and data entry processes.
* Address correspondence to An Vandervoort, MA, End-of-Life Care Research
Group, Ghent University & Vrije Universiteit Brussel, Laarbeeklaan 103, 1090 Jette,
Brussels, Belgium.
E-mail address: An.Vandervoort@vub.ac.be (A. Vandervoort).
JAMDA
journal homepage: www.jamda.com
1525-8610/$ - see front matter Copyright Ó2013 - American Medical Directors Association, Inc.
http://dx.doi.org/10.1016/j.jamda.2013.01.016
JAMDA xxx (2013) 1e8

many reside at home for a large part of their illness, many end up in
nursing homes as they reach the final phases of dementia.
8
However, large-scale and nationwide empirical studies describing
how people with dementia are currently dying in nursing homes are
lacking. Earlier studies were often limited in scope and populations,
investigating those who progress to a very advanced stage of
dementia only, in specific regions or institutions, or focusing on type
or place of care rather than on clinical characteristics and quality of
dying.
6,8e15
One important study from the Boston, MA, area (Choices,
Attitudes, and Strategies for Care of Advanced Dementia at the End-
of-Life [CASCADE]) reported the clinical trajectory of nursing home
residents with advanced dementia.
16
The investigators found that
distressing symptoms were common, as were clinical complications,
such as eating problems, pneumonia, and febrile episodes.
17
However,
several gaps in our knowledge of how well people with dementia are
dying remain, especially of those not progressing to an advanced
stage. Also, to provide population-based insights regarding their
quality of dying, large-scale representative studies are needed. To
describe the final phase of life in a representative sample of deaths, it
has been recommended to use a retrospective study design.
18,19
In this study, we present findings from what is, to our knowledge,
the first European nationwide population-based study describing the
end of life of nursing home residents with dementia. Our main
objectives were to study how many deceased nursing home residents
had dementia in Flanders, Belgium, and to describe their clinical
characteristics and quality of dying.
Methods
Study Design
A retrospective cross-sectional study was conducted in Flanders,
the Dutch-speaking part of Belgium, where 6.3 million people (58% of
the Belgian population) live. Structured questionnaires were completed
by the nurse most involved in care for the resident, the family physi-
cian (FP), and the nursing home administrator regarding deceased
nursing home residents with dementia, in a representative sample
of Flemish nursing homes (ie, high-care nursing homes for elderly
people). There are several reasons why we used a retrospective rather
than a prospective design to study nursing home residents dying with
dementia.
18,19
A retrospective design more easily allows identification
of a clear study population. It also allows identification of a population-
based sample of deaths on a nationwide level, whereas prospective
studies are often not representative because of censoring.
18,19
Study Facilities
A random cluster sampling procedure was used, starting with
a random sample of Flemish nursing homes. To be representative,
homes were randomly sampled and stratified by region (5 provinces)
and subsequently by bed capacity (up to or more than 90 residents)
and ownership (public, private/nonprofit, private/profit), factors
found to be related to end-of-life care quality in nursing homes in
previous research.
20,21
If an institution declined to participate,
another was randomly selected until the targeted number per strata
was reached.
Study Subjects
All deaths were recorded between May and October 2010.
A 2-stage screening protocol was used to identify eligible subjects.
As a first step, the nursing home administrator identified all
residents dying, in the nursing home or elsewhere, in the preceding
3 months and identified residents who met one of the criteria used by
the Belgian health insurance system to allocate financial resources
22
:
the criteria for (1) “category Cdementia”(ie, “being completely care
dependent or needing help for bathing, dressing, eating, toileting,
continence and transferring plus being disoriented in time and
space”) OR (2) disorientation in time and space (3or“almost daily
a problem with disorientation in time and space”). These broad first-
selection criteria minimized the risk of missing eligible residents with
dementia but risked including false positives. Therefore, FPs and
nurses of residents fulfilling these criteria were sent questionnaires in
which additional eligibility criteria were surveyed.
In the second step, to exclude false positives or residents who did
not have dementia according to the FP or nurse, additional eligibility
criteria required that the FP or nurse indicate that the resident “had
dementia”or “was diagnosed with dementia.”
Data Collection
A letter introducing the research project and asking for partici-
pation was sent to the board of directors and the “coordinating and
advisory physician”(responsible for end-of-life care in Belgian
nursing homes) of each selected nursing home. In cases of no reply,
the researcher contacted the nursing home after 2 weeks. In each
participating home, the board of directors appointed 1 contact person
for the study (ie, nursing home administrator or head nurse). The
researcher visited each nursing home and assisted the contact person
in identifying eligible residents, the nurse who was “most closely
involved in resident’s care,”and the resident’sFP.
The nurse, FP, and nursing home administrator were sent
a structured standardized questionnaire no later than 3 months after
the death. If the identified nurse was not willing to cooperate, the
head nurse was asked to fill in the questionnaire. Nonresponders
received a reminder after 3 weeks, with the administrator as
a mediator to guarantee anonymity.
Measurements
After-death questionnaires surveyed sociodemographic charac-
teristics, health status, clinical complications, and quality of dying.
A feasibility study preceding the main study tested design and
measurements. Because nursing home filing systems differed con-
siderably, the questionnaires of nurses and FPs were completed based
on the respondent’s memory.
FP questionnaires:
- resident’s health status: cause of dementia, coexisting
conditions
Nurse questionnaires:
- whether or not there was a family caregiver/friend involved
in the resident’s care
- time of onset of dementia
- resident’s health status:
Bat the time of admission
-
stage of dementia
-
ability to express wishes and competency in medical
decision making
Bfunctional and cognitive status 1 month before death
-
Bedford Alzheimer Nursing Severity Scale
(BANS-S)
23
-
Global Deterioration Scale (GDS),
24
classifying
dementia into 7 stages based on deficits in cognition
and function
16
-
Cognitive Performance Scale (CPS),
25
a validated
measure that assigns residents to cognitive perfor-
mance categories
16, 25
A. Vandervoort et al. / JAMDA xxx (2013) 1e82

-
sentinel events: ie, illnesses or events: pneumonia,
febrile episodes (other than pneumonia), problems
with eating or drinking, hip fracture, stroke, gastro-
intestinal bleeding, cancer, or other
17
-
quality of dying: presence of distressing symptoms
in last month of life measured by the Symptom
Management End-of-Life in Dementia (SM-EOLD)
26
Blast week before death
-
quality of dying: Comfort Assessment in Dying End-
of-Life in Dementia (CAD-EOLD)
26
and Quality of Life
in Late-Stage Dementia (QUALID)
27
-
stage of decubitus
28
-
degree of urine and feces incontinence
-
upper body or limb restraints
Bnutrition and hydration status at time of death
- type of death judged by the nurse: expected, expected
but faster than thought, neither expected nor unexpected,
unexpected
Nursing home administrator questionnaire:
- place of residence before nursing home admission, length
of stay, open or a closed unit, place of death
- last measurement of the Mini-Mental State Examination in
the patient’s administrative file
29
Statistical Analysis
Analyses were performed with PASW statistical software, 17.0
(SPSS Inc., Chicago, IL). Based on the CPS and GDS, residents were
classified into 3 groups (Table 1). Advanced dementia was defined as
by Mitchell et al.
16
Differences in distribution between very severe/
advanced dementia, severe dementia, and moderate/mild dementia
were calculated using Fisher Exact Tests. Medians and averages were
tested using Kruskal-Wallis and analysis of variance with post hoc
least significant difference (significance level P¼.05).
Ethical Aspects
The study protocol was approved by the Medical Ethical Committee
of UZ Brussel (University Hospital of Brussels). Anonymity was guar-
anteed for residents and respondents by using unique anonymous
numbers and a mediating contact person of the nursing home.
Results
Sample Description
Sixty-nine nursing homes (58% response rate) participated,
representative of all nursing homes in Flanders in terms of size,
region, and ownership because of the proportionate sampling. The
main reasons for not participating were lack of time because of high
work pressure, staff shortage, or overload of requests for research
participation. Response rates for FPs, nurses, and nursing home
administrators were 52.9%, 88.4%, and 95.0%, respectively. The
median time between death of the resident and completion of the
questionnaire was 65 days (interquartile range [IQR] 37e91 days)
for nurses, 82 days (IQR 48e137 days) for FPs, 134 days (IQR
45e104 days) for relatives, and 65 days (IQR 34e92 days) for the
nursing home administrator. Using data gathered by the professional
caregivers, nonresponse analysis showed no differences for important
resident (age, gender, length of stay, place of death, cognitive status,
disease severity) and care characteristics (end-of-life treatments,
nursing care) between participating and nonparticipating FPs, except
for the nurse’s perspective on end-of-life care and dying, in which
nurses perceived more consensus among relatives on care and
treatment of the resident (P¼.004) in cases in which the FP partic-
ipated than those in which the FP did not.
In participating homes, we identified 477 residents who had died
within the past 3 months, of whom 241 met the inclusion criteria in
the first step of the screening procedure (Figure 1). After screening
the returned questionnaires, in the second step we identified
205 deceased residents with dementia. Of all identified deaths, 46.7%
had dementia at the time of death. Seven of the 205 cases were
excluded because of too much missing data, leaving 198 for analysis:
106 (53.5%) cases of very severe or advanced dementia, 56 (28.3%) of
severe dementia, and 36 (18.2%) of moderate or mild dementia.
Analysis (not shown in tables) showed that age distribution was
representative (P¼.55) for the population of nursing home residents
dying with dementia insured by the 2 largest insurance companies in
Flanders that insure more than 70% of the population. Men were
overrepresented in our sample (P¼.02).
Mean age at time of death in our sample was 86.7 years, and more
than half (61.5%) were women (Table 2). Median length of nursing
home stay was 2.4 years, and median survival time after onset of
dementia was 3 years. Whereas differences in length of nursing home
stay were not statistically significant between dementia groups,
median survival time after onset of dementia was lower for residents
with mild/moderate dementia (1 year) compared with residents with
very severe/advanced dementia (4 years). Almost half of subjects
lived in a closed special care unit for dementia (47.9%), and 96.5% had
a family caregiver.
Health Status in Last Month of Life
Alzheimer disease was the cause of dementia in 50% of cases
(Table 3). Residents with dementia at time of death were at the time
of admission able or partly able to express their wishes in 84.3%
of cases and competent or partly competent for medical decision
making in 60.7%. Most (70.3%) had at least one coexisting condition
at time of death, of which cardiovascular diseases were reported most
frequently. They were also severely functionally and cognitively
disabled 1 month before death (mean BANS-S 20.8). Nursing severity
was lower in those with moderate/mild dementia (BANS-S P<.001).
Fewer coexisting conditions were reported in those with very se-
vere/advanced dementia (P¼.02).
Clinical Complications in Last Month of Life
Nurses reported that 95.5% of residents who died with dementia
had 1 or more sentinel events in the last month of life: eating or
drinking problems in 65.7%, febrile episodes in 42.9%, and pneumonia
in 32.3% of cases (Table 3). The only significant difference between
groups was that prevalence of eating/drinking problems in the last
month of life increased with severity of dementia (P¼.03).
Distressing Symptoms in Last Month of Life
Pain (mean 1.9), fear (mean 2.3), anxiety (mean 2.4), resistance to
care (mean 3.0), and agitation (mean 2.9) were most frequently
reported using the SM-EOLD scale (lower scores refer to more
Table 1
Classification of Dementia Severity 1 Month Before Death
Very severe or advanced
dementia
CPS 5 and GDS ¼7
Severe dementia (CPS 5 and GDS <7) or (CPS <5 and GDS ¼7)
Moderate or mild dementia CPS <5 and GDS <7
CPS, Cognitive Performance Scale; GDS, Global Deterioration Scale.
A. Vandervoort et al. / JAMDA xxx (2013) 1e83

symptoms) during the last month of life (Table 3). Presence of
symptoms did not differ significantly in stages of dementia.
Clinical Characteristics and Outcomes in Last Week of Life
In the last week of life (Table 4), low levels of comfort, measured
by CAD-EOLD scale (range 0 worst to 3 best), were reported for
difficulty swallowing (mean 1.9), pain (mean 2.0), lack of serenity
(mean 1.9), lack of peace (2.0), and lack of calm (2.0). Mean quality of
life during the last week (QUALID scale, range 11 best to 55 worst)
was 28.9, similar in all 3 groups; pressure sores were present in 26.9%
of cases, incontinence for urine or feces in respectively 89.2% and
85.6%, cachexia or severe cachexia in 45.8%, and dehydration or
severe dehydration in 38.6%. Physical upper body or limb restraints
were used in 21.4% of cases in the last week of life. Death was ex-
pected for 50.3% of those with very severe/advanced dementia, and
proportions of expected death were similar for the other dementia
stages. Place of death was hospital in 8.9% of cases. We found
statistically significant differences between stages of dementia with
respect to problems with choking (P<.001), swallowing (P<.001),
incontinence (P<.001), and dehydration in the last week of life
(P<.01), all being more common with more severe dementia.
Discussion
This study shows that nearly half of all nursing home residents in
Flanders, Belgium, die with dementia, and half of these have very
severe or advanced dementia at the time of death. Regardless of the
stage of dementia, many residents develop serious clinical compli-
cations and symptoms in the terminal phase of life.
To our knowledge, this is the first nationwide population-based
study describing the clinical characteristics and quality of dying
of nursing home residents with dementia. The random cluster
Dementia
according to FP
n = 7
Dementia
according to nurse
n = 85
Dementia
according to
nurse or FP
n = 113
Dementia according
to nurse or FP
n = 205
Deaths of residents in 69 nursing homes
n = 477
KATZ scale category Cdementia or
disorientation in time and space
n = 241
-236 cases not meeting
KATZ scale category Cd or
disorientation in time and space †
-19 cases, non response
(no questionnaire received
from FP and nurse)
Questionnaire received
from FP or nurse
n = 222
Questionnaire
received from
FP only
n = 9
Questionnaire
received from
nurse only
n = 95
Questionnaire
received from both
FP and nurse
n = 118
-17 cases, no dementia
according to nurse or FP
198 cases
-7 cases, questions regarding
primary outcomes measures
not filled in
Very severe or
Advanced Dementia ‡
n = 106
Severe
Dementia ‡
n = 56
Moderate or
Mild Dementia ‡
n = 36
Fig. 1. Overview of data gathered in this study. The age distribution of the sample was representative for nursing home residents dying with dementia in Flanders, Belgium, in 2010.
y
Cd or category Cdementia. KATZ scale: Katz S: Assessing self-maintenance: activities of daily living, mobility, and instrumental activities of daily living. J Am Geriatr Soc 1983,
31:721-727.
z
Very Severe or Advanced Dementia ¼CPS 5þGDS 7. Severe Dementia ¼CPS 5þGDS <7 OR CPS <5þGDS 7. Moderate or Mild dementia ¼CPS <5 AND GDS<7.
A. Vandervoort et al. / JAMDA xxx (2013) 1e84

sampling procedure, high response rates, and representativity of age
distribution contribute to the quality of the data. Nonresponse anal-
yses showed no differences between residents whose FPs did or did
not respond, except for the item of consensus among relatives on
care. Using a retrospective design, we could identify a population-
based sample in this setting. The use of a retrospective research
design has been proven to be superior to a prospective design in
identifying population-based information.
18,19
In Flemish nursing
homes, where residents with and without dementia live together and
move between different care levels according to their needs, selection
of those with dementia is difficult.Therefore,weuseda2-step
screening protocol to select the study population and validated
clinical scales to identify subgroups. Hence, we were able to study all
deceased nursing home residents with dementia and compare
advanced dementia with other stages. Using a combination of
different observers, we could provide a good epidemiological view of
the circumstances of dying with dementia in Flanders.
Nevertheless, the study has several limitations. Because a retro-
spective research design was used, memory bias cannot be ruled out;
however, this was limited by minimizing the time between death and
the completion of questionnaires and by focusing on the final month
of life. Also, even though representativity was difficult to test because
of lack of population data, men appear overrepresented in our sample.
Finally, when interpreting the results, we take into account that we
needed to rely on FPs’and nurses’reports and perception of health
status, clinical complications, and distressing symptoms, which may
differ from what residents themselves would have reported; however,
self-reporting was not possible in this patient population.
30
From a public health point of view, our study provides important
data concerning the magnitude of the problem of dying with
dementia in nursing homes. Of all deaths in Flemish nursing homes,
about half concern residents with some form of dementia, most of
whom experience serious limitations and complications at the end of
life. Moreover, median nursing home stay was 2.4 years, and median
survival time after onset of dementia was 3.0 years. Considering the
future increase in people developing dementia,
31
our results confirm
that dying well with dementia in nursing homes is one of the most
important public health challenges.
Also, many residents dying with moderate/mild or severe
dementia appear to encounter end-of-life problems similar to those
with very severe/advanced dementia. As was previously observed
in the United States,
16
nursing home residents with advanced
dementia suffer from important clinical complications. However,
we additionally found that the prevalence of clinical complications
in the last month of life (eg, pneumonia or other infections), levels
of comfort, and quality of life in the last week did not differ
Table 2
Deceased Nursing Home Residents With Dementia in Flanders, Belgium: Description of the Sample (n ¼198)
Residents With
Dementia, n (%)
Very Severe or Advanced
Dementia, n (%)
Severe Dementia, n (%) Moderate or Mild
Dementia, n (%)
PValue*
Total n (%)
y
198 (100.0) 106 (53.5) 56 (28.3) 36 (18.2)
Age, y
<90 118 (63.4) 62 (63.9) 34 (60.7) 22 (66.7) .83
90 68 (36.6) 11 (36.1) 22 (39.3) 35 (33.3)
Mean SD 86.7 7.0 86.3 7.2 87.0 7.4 87.3 5.9
Gender
Male 72 (38.5) 39 (39.8) 22 (39.3) 11 (33.3) .81
Female 115 (61.5) 59 (60.2) 34 (60.7) 22 (66.7)
Motivation for nursing home admission
z
Behavioral problems 37 (19.1) 25 (24.5) 7 (12.5) 5 (13.9) .14
Physical disease 52 (27.2) 26 (25.5) 10 (18.5) 16 (45.7) .02
Burden in family caregiver 68 (35.4) 44 (43.1) 17 (30.9) 7 (20.0) .03
Problems with care 45 (23.9) 27 (26.5) 14 (25.9) 4 (12.5) .25
Lack of self-management 92 (48.9) 42 (41.2) 31 (57.4) 19 (59.4) .07
Cognitive problems 31 (16.0) 16 (15.4) 12 (21.4) 3 (8.8) .29
Other 11 (5.6) 3 (2.8) 5 (8.9) 3 (8.6) .12
Median length of nursing home stay, median y
(interquartile range)
2.4 (1.2e4.6) 2.4 (1.2e4.5) 3.2 (1.4e5.2) 2.1 (0.5e3.2) .16
Median survival after onset of dementia, median y
(interquartile range)
3.0 (2.0e5.0) 4.0 (2.0e5.0) 4.0 (2.0e5.0) 1.0 (0.8e2.0) <.001
Type of unit at time of death
Living in closed special care unit for dementia 90 (47.9) 52 (52.5) 25 (44.6) 13 (39.4) .28
Living in an open unit not delivering specialist dementia care 82 (43.6) 36 (36.4) 28 (50.0) 18 (54.5)
Other type of unit
x
16 (8.5) 11 (11.1) 3 (5.4) 2 (6.1)
Having an involved family caregiver
z
Child 145 (73.2) 76 (71.7) 40 (71.4) 29 (80.6) .61
Partner 35 (17.7) 19 (17.9) 12 (21.4) 4 (11.1) .49
Other 61 (30.8) 31 (29.2) 18 (32.1) 12 (33.3) .86
None 7 (3.5) 1 (2.8) 3 (5.4) 3 (2.8) .78
CPS 1 month before death
Intact, Borderline intact, Mild impairment (score 0e1e2) 8 (4.1) d9 (1.9) 7 (20.5) <.001
Moderate impairment (score 3) 27 (13.9) d6 (11.1) 21 (61.8)
Moderately severe impairment (score 4) 9 (4.6) d3 (5.6) 6 (17.6)
Severe impairment (score 5) 61 (31.4) 33 (31.1) 28 (51.9) d
Very severe impairment (score 6) 89 (45.9) 73 (68.9) 16 (29.6) d
GDS 1 month before death
Stage 7 105 (57.7) 96 (100.0) 9 (17.3) d<.001
CPS, Cognitive Performance Scale; GDS, Global Deterioration Scale.
*Differences between Very Severe/Advanced dementia, Severe dementia, and Moderate/Mild dementia are tested using Fisher Exact test, analysis of variance, or Kruskall-
Wallis test.
y
Missing values: age n ¼12, gender n ¼11, motivation for nursing home admission n ¼11, length of stay n ¼14, survival time n ¼69, type of unit n ¼10, involved caregiver
n¼0, CPS n ¼4, GDS n ¼16.
z
Multiple answers possible.
x
Other type of unit, ie, living in an open special care unit or living in a closed unit not delivering specialist dementia care.
A. Vandervoort et al. / JAMDA xxx (2013) 1e85

significantly between groups, nor did the presence of distressing
symptoms in the last month. Hence, nursing homes face major
challenges in caring for residents dying with dementia regardless of
the stage.
Overall, the most common clinical characteristics in the last
week of life were incontinence for urine or feces, pressure sores,
and cachexia/dehydration. The most common clinical complications
in the last month of life were eating or drinking problems (65.7%),
Table 3
Health Status, Clinical Complications, and Distressing Symptoms in the Last Month of Life of Nursing Home Residents With Dementia (n ¼198)
Residents With
Dementia, n (%)
Very Severe or Advanced
Dementia, n (%)
Severe Dementia, n (%) Moderate or Mild
Dementia, n (%)
PValue*
Total n
y
198 106 56 36
HEALTH STATUS
Cause of dementia
z
Alzheimer dementia 49 (50.0) 34 (55.7) 7 (30.4) 8 (57.1) .12
Vascular dementia 34 (34.7) 20 (32.8) 10 (43.5) 4 (28.6) .62
Other type of dementia 11 (11.2) 9 (14.8) 2 (8.7) 0 (0.0) .35
Functional and cognitive status
Bedford Alzheimer Nursing Severity Scale (BANS-S) 1 month
before death
x
Mean (SD) 20.8 (3.9) 22.5 (2.8) 20.6 (3.5) 16.3 (3.8) <.001
k
Mini Mental State Examination (last measurement)
{
25 3 (1.5) 1 (0.9) 2 (3.6) 0 (0.0) .13
24-21 6 (3.0) 3 (2.8) 2 (3.6) 1 (2.8)
20-11 52 (26.3) 25 (23.6) 12 (21.4) 15 (41.7)
10 115 (58.1) 69 (65.1) 30 (53.6) 16 (44.4)
No measurement in patient file 22 (11.1) 8 (7.5) 10 (17.9) 4 (11.1)
Was able to express wishes at time of nursing home admission
Yes 77 (39.1) 31 (29.5) 24 (42.9) 22 (61.1) <.001
Partly 89 (45.2) 47 (44.8) 28 (50.0) 14 (38.9)
No 31 (15.7) 27 (25.7) 4 (7.1) 0 (0.0)
Was competent for medical decision-making at time of nursing home admission
Yes 51 (26.0) 23 (21.9) 13 (23.2) 15 (42.9) <.001
Partly 68 (34.7) 29 (27.6) 21 (37.5) 18 (51.4)
No 77 (39.3) 53 (50.5) 22 (39.3) 2 (5.7)
Co-existing conditions**
Malignant tumor 12 (10.8) 5 (7.6) 4 (16.0) 3 (15.0) .35
Cardiovascular 32 (28.8) 14 (21.2) 11 (44.0) 7 (35.0) .08
Respiratory 15 (13.5) 9 (13.6) 4 (16.0) 2 (10.0) .87
Neurological 17 (15.3) 12 (18.2) 3 (12.0) 2 (10.0) .67
Urogenital system 9 (8.1) 4 (6.1) 4 (16.0) 1 (5.0) .30
Other 18 (16.4) 10 (15.4) 4 (16.0) 4 (20.0) .94
None of the above 33 (29.7) 26 (39.4) 5 (20.0) 2 (10.0) .02
CLINICAL COMPLICATIONS
Sentinel events during last month of life registered by the nurse
Pneumonia 64 (32.3) 33 (31.1) 19 (33.9) 12 (33.3) .92
Febrile episode (other than pneumonia) 85 (42.9) 42 (39.6) 27 (48.2) 16 (44.4) .54
Eating or drinking problem 130 (65.7) 76 (71.7) 37 (66.1) 17 (47.2) .03
Other sentinel events 73 (36.9) 39 (36.8) 21 (37.5) 13 (36.1) >.99
No sentinel event 9 (4.5) 4 (3.8) 2 (3.6) 3 (8.3) .53
DISTRESSING SYMPTOMS
SM-EOLD during last month of life: range 0 (worst) to 5 (best) mean (SD)
yy
Pain 1.9 (1.9) 2.1 (2.0) 1.8 (1.9) 1.7 (1.7) .44
Shortness of breath 3.2 (2.0) 3.3 (1.9) 3.2 (2.1) 2.8 (2.1) .51
Skin breakdown 4.1 (1.8) 4.1 (1.7) 3.9 (2.0) 4.2 (1.8) .71
Calm 3.4 (2.0) 3.5 (2.0) 3.2 (2.1) 3.2 (2.1) .62
Depression 3.6 (2.0) 3.8 (1.9) 3.7 (2.0) 2.9 (2.1) .06
zz
Fear 2.3 (2.0) 2.2 (1.9) 2.3 (2.0) 2.6 (2.0) .47
Anxiety 2.4 (2.0) 2.3 (2.0) 2.7 (2.1) 2.3 (2.1) .55
Agitation 2.9 (2.1) 2.9 (2.1) 2.9 (2.1) 2.8 (2.0) .99
Resistiveness to care 3.0 (2.1) 2.7 (2.1) 3.2 (2.1) 3.5 (1.9) .09
*Statistically significant difference between Very Severe/Advanced dementia, Severe dementia, and Moderate/Mild dementia.
y
Missing values: cause of dementia n ¼100 (of which 85 because no questionnaire was received from the FP), BANS-S n ¼0, MMSE n ¼96, expressing wishes n ¼1,
competence at nursing home admission n ¼2, co-existing conditions n ¼87 (of which 85 because no questionnaire was received from the FP), sentinel events n ¼87, pain n ¼
8, shortness of breath n ¼19, skin breakdown n ¼9, calm n ¼29, depression n ¼18, fear n ¼4, anxiety n ¼9, agitation n ¼11, resistiveness to care n ¼6, total SM-EOLD n ¼10.
z
Multiple answers possible. Don’t know cause of dementia n ¼13.
x
Scores on the Bedford Alzheimer’s Severity Subscale range from 7 to 28; higher scores indicate greater functional disability.
k
Statistically significant difference found between Very Severe/Advanced dementia and Severe dementia, and between Very Severe/Advanced dementia and Moderate/
Mild dementia.
{
Assessment was on average 8 months before death.
**Multiple answers possible. The answer to this question was reported by FPs only. In 85 cases, we received no questionnaire from the FP; these cases were reported as
missing, so n ¼113.
yy
SM-EOLD. All items were (re)coded so that higher scores mean better symptoms management. Total score is constructed by summing the value of each item and ranges
from 0 to 45 with higher scores indicating better symptom control; missing SM-EOLD items were imputed with resident means in case there were 3 or fewer missing items.
zz
Statistically significant difference found between Very Severe/Advanced Dementia and Moderate/Mild dementia.
A. Vandervoort et al. / JAMDA xxx (2013) 1e86

febrile episodes (42.9%), and pneumonia (32.2%). These 3 compli-
cations are also cited as highly prevalent in US and Italian studies
limited to advanced dementia.
17,3 2
In the prospective Boston study
among nursing home residents with advanced dementia (323
residents of whom 177 died during the study), an eating or drinking
problem was reported in 85.8%, febrile episodes in 52.6%, and
pneumonia in 41.4% during the study period of 18 months.
17
Eating/
drinking problems were the most prevalent clinical complications
across all dementia stages and were particularly common among
residents with advanced dementia. This reinforces the need to pay
particular attention to this end-of-life complication. It might be one
of the most common end points for which family members need to
be prepared, particularly because far-reaching end-of-life decisions
concerning forgoing food or fluid might need to be made in the final
days of life. It can also be one of the main issues to discuss with
residents early in their disease course as part of an advance care
planning process.
Across all dementia stages, the most commonly reported symptoms
in the last month of life were pain, fear, and anxiety. Although com-
parison with other studies is difficult because of differences
in population and research procedures, these findings do not seem
surprising. In other studies from other European Union countries or
the United States,
32e36
people with dementia also appear to experience
a number of burdensome physical, emotional, or psychosocial symp-
toms. Interestingly, although some other studies report that anxiety
seems to decrease in the severe stages of dementia,
34
the nurses in
Table 4
Clinical Characteristics and Outcomes in the Last Week of Life of Nursing Home Residents With Dementia (n ¼198)
Residents With
Dementia
Very Severe or Advanced
Dementia
Severe Dementia Moderate or Mild
Dementia
PValue*
Total n
y
198 106 56 36
CAD-EOLD range during last week of life: 0 (worst) to 3 (best) mean (SD)
z
Discomfort 2.1 (0.7) 2.1 (0.7) 2.1 (0.8) 2.1 (0.7) .97
Pain 2.0 (0.7) 2.1 (0.7) 1.9 (0.7) 2.0 (0.7) .28
Restlessness 2.1 (0.7) 2.1 (0.8) 2.0 (0.8) 2.1 (0.7) .55
Shortness of breath 2.2 (0.8) 2.2 (0.7) 2.2 (0.8) 2.1 (0.8) .72
Choking 2.1 (0.8) 1.9 (0.7) 2.3 (0.8) 2.5 (0.7) <.001
x
Gurgling 2.3 (0.8) 2.3 (0.8) 2.3 (0.8) 2.5 (0.7) .58
Difficulty swallowing 1.9 (0.8) 1.8 (0.7) 2.0 (0.8) 2.4 (0.8) .001
k
Fear 2.1 (0.7) 2.0 (0.7) 2.2 (0.8) 2.1 (0.7) .63
Anxiety 2.2 (0.7) 2.2 (0.7) 2.2 (0.8) 2.1 (0.7) .74
Crying 2.7 (0.6) 2.6 (0.7) 2.7 (0.6) 2.8 (0.4) .57
Moaning 2.3 (0.7) 2.3 (0.7) 2.3 (0.7) 2.5 (0.6) .43
Serenity 1.9 (0.7) 1.9 (0.7) 1.9 (0.7) 2.0 (0.8) .96
Peace 2.0 (0.7) 2.0 (0.7) 2.0 (0.7) 2.0 (0.8) .95
Calm 2.0 (0.7) 2.1 (0.7) 2.1 (0.7) 1.9 (0.8) .44
Total 30.0 (6.0) 29.8 (6.4) 30.0 (5.6) 30.8 (5.7) .70
Quality of life in late-stage dementia (QUALID) scale during last week of life
{
Mean (SD) 28.9 (8.6) 29.4 (8.2) 29.1 (8.4) 27.0 (9.8) .34
Decubitus during last week of life n (%) 52 (26.9) 29 (28.2) 15 (27.3) 8 (22.9) .88
Urine incontinence during last week of life n (%)
Incontinent 174 (89.2) 101 (96.2) 52 (92.9) 21 (61.8) <.001
Feces incontinence during last week of life n (%)
Incontinent 167 (85.6) 100 (95.2) 47 (83.9) 20 (58.8) <.001
Nutrition status at time of death as evaluated by the nurse n (%)
Very cachectic 32 (17.9) 16 (16.2) 12 (23.5) 4 (13.8) .27
Cachectic 50 (27.9) 28 (28.3) 17 (33.3) 5 (17.2)
Normal 72 (40.2) 37 (37.4) 17 (33.3) 18 (62.1)
Adipose 17 (9.5) 13 (13.1) 3 (5.9) 1 (3.4)
Very adipose 8 (4.5) 5 (5.1) 2 (3.9) 1 (3.4)
Hydration status at time of death n (%)
Normal 39 (21.2) 14 (13.7) 10 (19.6) 15 (48.4) .01
Little dehydrated 74 (40.2) 47 (46.1) 18 (35.3) 9 (29.0)
Dehydrated 55 (29.9) 33 (32.4) 17 (33.3) 5 (16.1)
Very dehydrated 16 (8.7) 8 (7.8) 6 (11.8) 2 (6.5)
Use of upper body restraints or limb restraints during
last week of life n (%)
Used 41 (21.4) 26 (25.0) 10 (18.5) 5 (14.7) .42
Type of death n (%)
Expected 96 (50.3) 53 (51.0) 29 (54.7) 14 (41.2) .43
Other 95 (49.7) 51 (49.0) 24 (45.3) 20 (58.8)
Place of death n (%)
Nursing home 171 (90.0) 94 (93.1) 49 (87.5) 28 (84.8) .28
General hospital ward or intensive care unit 17 (8.9) 6 (5.9) 7 (12.5) 4 (12.1)
Palliative care unit 2 (1.1) 1 (1.0) 0 (0.0) 1 (3.0)
*Differences between Very Severe/Advanced dementia, Severe dementia, and Moderate/Mild dementia.
y
Missing values: discomfort n ¼19, pain n ¼9, restlessness n ¼15, shortness of breath n ¼12, choking n ¼16, gurgling n ¼18, difficulty swallowing n ¼11, fear n ¼13,
anxiety n ¼14, crying n ¼17, moaning n ¼16 serenity n ¼16, peace n ¼18, calm n ¼19, SM-CAD total n ¼16, QUALID n ¼4, decubitus n ¼5, urine incontinence n ¼3, feces
incontinence n ¼3, nutrition status n ¼19, hydration status n ¼14, restraints n ¼6, type of death n ¼7, place of death n ¼8.
z
CAD-EOLD. All items were (re)coded so that higher scores mean better symptoms management. The CAD-EOLD total score is constructed by summing the value of each
item. It ranges from 14 to 42 with higher scores indicating better symptom control; missing CAD-EOLD items were imputed with residents’means in case there were 4 or
fewer missing scores on the scale.
x
Statistically significant difference found between Very Severe/Advanced dementia and Severe dementia, and between Very Severe/Advanced dementia and Moderate/
Mild dementia.
k
Statistically significant difference found between Very Severe/Advanced dementia and Moderate/Mild dementia.
{
QUALID. Total scores range from 11 points to 55 points with lower scores reflecting a higher quality.
A. Vandervoort et al. / JAMDA xxx (2013) 1e87

our study indicated fear and anxiety to be among the most prevalent
symptoms in residents with dementia regardless of the stage.
It is remarkable that survival time after diagnosis was shorter in
the case of mild/moderate dementia than in the case of severe or
advanced dementia. This might be explained by the fact that coex-
isting conditions, such as cardiovascular diseases, were more often
reported by the FPs among residents with mild/moderate dementia
than among the more advanced residents. Residents with dementia
seem to either die earlier with more comorbid conditions or die later
with more severe or advanced dementia.
Finally, we found that 21.4% of residents with dementia at time of
death were physically restrained (upper body or limb restraints)
during the last week of life. The use of these restraints is controversial
and often associated with stress, agitation, and a higher risk of falls.
37
Also, on average, 1 in 10 nursing home residents with dementia died
in a hospital, indicating that burdensome terminal transitions remain
an issue. Further research is needed concerning the best approach to
these problems.
Conclusion
In conclusion, considering our finding that half of nursing home
residents die with dementia, a number that will only increase in the
future, dying well with dementia in nursing homes is one of the most
important public health challenges for end-of-life care. Serious clin-
ical complications and distressing symptoms were reported regard-
less of the stage of dementia. This suggests that research addressing
the challenge of providing high-quality end-of-life care for residents
with dementia should not only focus on residents with advanced
dementia. It also suggests that end-of-life care practices in nursing
homes might benefit from a regular and systematic assessment of
symptoms and problems among residents with dementia, and from
preparing caregivers, residents, and their families for the clinical
complications that might occur at the end of life and the accompa-
nying end-of-life decisions.
Acknowledgments
We thank the umbrella organizations for nursing homes (Ver-
eniging van Vlaamse Steden en Gemeenten, Federatie Onafhankelijke
Seniorenzorg and the Federatie van Rustoorden van België), BVGG-
Crataegus, CRA-Domus, Belgische Vereniging voor Gerontologie en
Geriatrie, Expertisecentrum Dementie, Federation Palliative Care
Flanders, and the work group Palliatieve Zorg en Geriatrie for their
support given to this study; Eva Dumon for her support in data
collection; Departement R&D van de CM Landsbond and Johan Van
Overloop for providing population data; Jane Ruthven for English
editing; and all participating nursing homes and respondents for
providing the study data.
References
1. Mura T, Dartigues JF, Berr C. How many dementia cases in France and Europe?
Alternative projections and scenarios 2010e2050. Eur J Neurol 2010;17:
252e259.
2. Banaszak-Holl J, Fendrick AM, Foster NL, et al. Predicting nursing home
admission: Estimates from a 7-year follow-up of a nationally representative
sample of older Americans. Alzheimer Dis Assoc Disord 2004;18:83e89.
3. Jagger C, Andersen K, Breteler MM, et al. Prognosis with dementia in Europe:
A collaborative study of population-based cohorts. Neurologic Diseases in the
Elderly Research Group. Neurology 2000;54:S16eS20.
4. Scuvee-Moreau J, Kurz X, Dresse A. The economic impact of dementia
in Belgium: Results of the National Dementia Economic Study (NADES). Acta
Neurol Belg 2002;102:104e113.
5. Sachs GA, Shega JW, Cox-Hayley D. Barriers to excellent end-of-life care for
patients with dementia. J Gen Intern Med 2004;19:1057e1063.
6. Mitchell SL, Kiely DK, Hamel MB. Dying with advanced dementia in the nursing
home. Arch Intern Med 2004;164:321e326.
7. Aminoff BZ, Adunsky A. Dying dementia patients: Too much suffering, too little
palliation. Am J Hosp Palliat Care 2005;22:344e348.
8. Houttekier D, Cohen J, Bilsen J, et al. Place of death of older persons with
dementia. A study in five European countries. J Am Geriatr Soc 2010;58:
751e756.
9. Mitchell SL, Teno J, Miller SC, Mor V. A national study of the location of death
for older persons with dementia. J Am Geriatr Soc 2005;53:299e305.
10. Afzal N, Buhagiar K, Flood J, Cosgrave M. Quality of end-of-life care for
dementia patients during acute hospital admission: A retrospective study in
Ireland. Gen Hosp Psychiatry 2010;32:141e146.
11. Cordner Z, Blass DM, Rabins PV, Black BS. Quality of life in nursing home
residents with advanced dementia. J Am Geriatr Soc 2010;58:2394e2400.
12. Goodman C, Evans C, Wilcock J, et al. End of life care for community dwelling
older people with dementia: An integrated review. Int J Geriatr Psychiatry
2010;25:329e337.
13. Sampson EL, Jones L, Thune-Boyle IC, et al. Palliative assessment and advance
care planning in severe dementia: An exploratory randomized controlled trial
of a complex intervention. Palliat Med 2011;25:197e209.
14. Torke AM, Holtz LR, Hui S, et al. Palliative care for patients with dementia:
A national survey. J Am Geriatr Soc 2010;58:2114e2121.
15. van der Steen JT, Lane P, Kowall NW, et al. Antibiotics and mortality in patients
with lower respiratory infection and advanced dementia. J Am Med Dir Assoc
2012;13:156e161.
16. Mitchell SL, Kiely DK, Jones RN, et al. Advanced dementia research in the
nursing home: The CASCADE study. Alzheimer Dis Assoc Disord 2006;20:
166e175.
17. Mitchell SL, Teno JM, Kiely DK, et al. The clinical course of advanced dementia.
N Engl J Med 2009;361:1529e1538.
18. Earle CC, Ayanian JZ. Looking back from death: The value of retrospective
studies of end-of-life care. J Clin Oncol 2006;24:838e840.
19. Teno JM. Measuring end-of-life care outcomes retrospectively. J Palliat Med
2005;8:S42eS49.
20. Hillmer MP, Wodchis WP, Gill SS, et al. Nursing home profit status and quality
of care: Is there any evidence of an association? Med Care Res Rev 2005;62:
139e166.
21. McGregor MJ, Tate RB, Ronald LA, McGrail KM. Variation in site of death among
nursing home residents in British Columbia, Canada. J Palliat Med 2007;10:
1128e1136.
22. Aerneart A, Delesie L. Ijking van meetgegevens: riziv ADL-index. Acta Hospi-
talia 1999;4:19e31.
23. Bellelli G, Frisoni GB, Bianchetti A, Trabucchi M. The Bedford Alzheimer Nursing
Severity scale for the severely demented: Validation study. Alzheimer Dis
Assoc Disord 1997;11:71e77.
24. Reisberg B, Ferris SH, de Leon MJ, Crook T. The Global Deterioration Scale for
assessment of primary degenerative dementia. Am J Psychiatry 1982;139:
1136e1139.
25. Morris JN, Fries BE, Mehr DR, et al. MDS Cognitive Performance Scale.
J Gerontol 1994;49:M174eM182.
26. Volicer L, Hurley AC, Blasi ZV. Scales for evaluation of end-of-life care in
dementia. Alzheimer Dis Assoc Disord 2001;15:194e200.
27. Weiner MF, Martin-Cook K, Svetlik DA, et al. The quality of life in late-stage
dementia (QUALID) scale. J Am Med Dir Assoc 2000;1:114e116.
28. European Pressure Ulcer Advisory Panel. Guidelines on treatment of pressure
ulcers. EPUAP Review 1999;1:31e33.
29. Folstein MF, Robins LN, Helzer JE. The Mini-Mental State Examination. Arch
Gen Psychiatry 1983;40:812.
30. Bridge M, Roughton DI, Lewis S, et al. Using caregivers-as-proxies to retro-
spectively assess and measure quality of dying of palliative care clients. Am J
Hosp Palliat Care 2002;19:193e199.
31. Ferri CP, Prince M, Brayne C, et al. Global prevalence of dementia: A Delphi
consensus study. Lancet 2005;366:2112e2117.
32. Di Giulio P, Toscani F, Villani D, et al. Dying with advanced dementia in long-
term care geriatric institutions: A retrospective study. J Palliat Med 2008;11:
1023e1028.
33. Meeussen K, Van den Block L, Echteld M, et al. Older people dying
with dementia: A nationwide study. Int Psychogeriatr 2012;24:1581e1591.
34. Seignourel PJ, Kunik ME, Snow L, et al. Anxiety in dementia: A critical review.
Clin Psychol Rev 2008;28:1071e1082.
35. van der Steen JT, Gijsberts MJ, Muller MT, et al. Evaluations of end of life with
dementia by families in Dutch and US nursing homes. Int Psychogeriatr 2009;
21:321e329.
36. Zimmerman S, Anderson W, Brode S, et al. Comparison of characteristics
of nursing homes and other residential long-term care settings for people
with dementia. Rockville, MD: Agency for Healthcare Research and Quality;
2008.
37. Scherder EJ, Bogen T, Eggermont LH, et al. The more physical inactivity,
the more agitation in dementia. Int Psychogeriatr 2010;22:1203e
1208.
A. Vandervoort et al. / JAMDA xxx (2013) 1e88