A preview of this full-text is provided by Karger Publishers.
Content available from Neurodegenerative Diseases
This content is subject to copyright.
Original Paper
Neurodegener Dis 2019;19:4–11
Physician and Surrogate Agreement with
Assisted Dying and Continuous Deep Sedation
in Advanced Dementia in Switzerland
Andrea Jutta Loizeau
a–c Simon M. Cohen
c Susan L. Mitchell
c, d
Nathan Theill
a, b, e Stefanie Eicher
a, b Mike Martin
a, b, f Florian Riese
a, e
a University Research Priority Program “Dynamics of Healthy Aging,” University of Zurich, Zurich, Switzerland;
b Center for Gerontology, University of Zurich, Zurich, Switzerland; c Hebrew SeniorLife Hinda and Arthur Marcus
Institute for Aging Research, Boston, MA, USA; d Department of Medicine, Beth Israel Deaconess Medical Center,
Boston, MA, USA; e Division of Psychiatry Research and Psychogeriatric Medicine, University of Zurich, Zurich,
Switzerland; f Department of Psychology, University of Zurich, Zurich, Switzerland
Received: July 29, 2018
Accepted after revision: February 18, 2019
Published online: April 23, 2019
Andrea Jutta Loizeau, PhD
Hebrew SeniorLife Hinda and Arthur Marcus Institute for Aging Research
1200 Centre Street
Boston, MA 02131 (USA)
E-Mail andrealoizeau @ hsl.harvard.edu
© 2019 S. Karger AG, Basel
E-Mail karger@karger.com
www.karger.com/ndd
DOI: 10.1159/000499113
Keywords
Dementia · Assisted dying · Sedation · End of life ·
Decision-making
Abstract
Background: Assisted dying and continuous deep sedation
(CDS) are controversial practices. Little is known about the
perceptions of physicians and surrogates about these prac-
tices for patients with advanced dementia. Objectives: To
describe and compare physician and surrogate agreement
with the use of assisted dying and CDS in advanced demen-
tia. Design, Setting, Subjects: Physicians (n = 64) and surro-
gates (n = 168) of persons with advanced dementia were re-
cruited as part of a randomized controlled trial in Switzer-
land that tested decision support tools in this population.
Methods: At baseline, the participants were asked about
their agreement with assisted dying and CDS in advanced
dementia using the following response options: “complete-
ly agree,” “somewhat agree,” “somewhat disagree,” “com-
pletely disagree,” and “do not know.” Multivariable logistic
regressions compared the likelihood that surrogates versus
physicians would completely or somewhat agree (vs. com-
pletely or somewhat disagree) with these practices. Results:
The physicians and surrogates, respectively, had a mean age
(SD) of 50.6 years (9.9) and 57.4 years (14.6); 46.9% (n = 30/64)
and 68.9% (n = 115/167) were women. A total of 20.3% (n =
13/64) of the physicians and 47.0% (n = 79/168) of the sur-
rogates agreed with assisted dying in advanced dementia.
Surrogates were significantly more likely to agree with this
practice than physicians (adjusted odds ratio, 3.87; 95% CI:
1.94, 7.69). With regard to CDS, 51.6% (n = 33/64) of the phy-
sicians and 41.9% (n = 70/169) of the surrogates agreed with
this practice, which did not differ significantly between the
groups. Conclusions: The surrogates were more agreeable
to considering assisted dying in the setting of advanced de-
mentia than the physicians, and about half of the partici-
pants in both groups reported CDS to be an appropriate op-
tion for this population. © 2019 S. Karger AG, Basel
End-of-Life Practices in Advanced
Dementia
5
Neurodegener Dis 2019;19:4–11
DOI: 10.1159/000499113
Introduction
Dementia afflicts more than 50 million people world-
wide and is one of the most common causes of death in
Switzerland [1, 2]. In advanced dementia, patients have
profound cognitive and functional deficits, and they ex-
perience clinical complications that may cause discom-
fort and a poor quality of life [3–5].
Assisted dying is a controversial practice that has
been used in certain contexts for patients with life-lim-
iting disease [6]. Assisted dying is when a physician ei-
ther administers drugs that cause a person to die (i.e.,
euthanasia) or provides lethal drugs for self-administra-
tion (i.e., physician-assisted suicide) [7]. In the case of
euthanasia, the physician takes an active role in com-
pleting the request of the patient incapable of complet-
ing the final act, for instance, by administering a lethal
injection. In assisted suicide, the physician solely makes
lethal means available to the patient, but the completion
of the final act is left to the patient. Currently, physician-
assisted dying is legal in only a few states in the USA as
well as in a limited number of countries, including Swit-
zerland [6, 8, 9]. However, only the Netherlands, Bel-
gium, and Luxembourg allow assisted dying for people
with advanced dementia [10]. This request must be
made in an advance directive by a patient with dementia
when decision-making capacity is still intact. Only lim-
ited research exists on the perceptions of physicians and
surrogates about the use of assisted dying in advanced
dementia [11–15].
Another controversial practice for terminally ill pa-
tients is continuous deep sedation (CDS) until death, a
last-resort palliative practice to alleviate suffering [16,
17]. Medications are administered until death to low-
er the level of consciousness and relieve refractory
symptoms of patients with a life expectancy of typically
less than 2 weeks [18]. To date, CDS has predominant-
ly been used for advanced cancer patients who are ex-
periencing extreme physical suffering in their final
weeks of life [16, 19]. CDS has rarely been used in ad-
vanced dementia, a condition predominantly charac-
terized by cognitive impairments, with some, but typi-
cally less, physical suffering than metastatic cancer [3,
5, 20].
To better understand the perceptions of physi-
cians and surrogates about the use of assisted dying
and CDS in advanced dementia, we analyzed base-
line data from a randomized controlled trial, the
DEMentia FACT boxes (DEMFACT) Study, conducted
in the Swiss German region of Switzerland. The objec-
tives of this report were to (1) describe physician
andsurrogate agreement with the use of assisted dying
and CDS in advanced dementia, and (2) compare the
agreement with the use of these practices in these two
groups.
Subjects and Methods
Data Source
The data were retrieved from baseline assessments made in
the DEMFACT Study, conducted between June 2016 and Octo-
ber 2016 in the Swiss German region of Switzerland. DEMFACT
was a randomized controlled trial that evaluated FACT box de-
cision support tools for treatment decision-making regarding
advanced dementia among 64 physicians and 168 surrogates
(relatives of dementia patients, n = 100; professional guardians,
n = 68) who were randomized to either an intervention arm (n
= 114) or a control arm (n = 118) [21]. The participants in the
intervention and control arms were mailed written question-
naires at baseline and 1 month later that included questions
about the use of treatments for advanced dementia patients
based on hypothetical scenarios. At the 1-month follow-up, the
intervention participants received the decision support tools,
whereas the control participants were given no additional infor-
mation.
The main DEMFACT Study compared changes in prespecified
outcomes between the baseline and follow-up assessments in the
intervention arm relative to the control arm. In this report, com-
bined data from the baseline questionnaires in both arms (i.e., be-
fore receiving the decision support tools in the intervention arm)
were analyzed to describe agreement with the use of continuous
sedation and assisted dying in advanced dementia. The baseline
questionnaires were completed between June 2, 2016, and July 31,
2016.
Population
To understand the impact of the DEMFACT intervention on
various decision-makers, the participants included physicians as
well as surrogates (relatives of dementia patients and professional
guardians) who were potentially responsible for the care of ad-
vanced dementia patients. Physicians were identified from the
mailing lists of the Swiss Association for Palliative Medicine, Care
and Support, and the Swiss Geriatric Medicine Society. Surrogates
were identified either through the Alzheimer Association of the
canton of Zurich or from the Swiss Association of Professional
Guardians mailing list. The detailed recruitment procedure is de-
scribed elsewhere [21]. Eligible participants were mailed a consent
form, which they were asked to sign and return to the research
team.
Data Elements
Data collection was procedurally identical across physician
and surrogate participants. All variables in this report were col-
lected at the DEMFACT baseline assessment using a written
questionnaire sent and returned by mail (approx. 60 min to
complete).
Loizeau/Cohen/Mitchell/Theill/Eicher/
Martin/Riese
Neurodegener Dis 2019;19:4–11
6
DOI: 10.1159/000499113
The outcomes were participants’ agreement with the use ofas-
sisted dying and CDS in advanced dementia. Agreement with as-
sisted dying was ascertained using the following question: “The
current legislation prohibits physician-assisted suicide (assisted
dying) for persons lacking the capacity to consent. This prevents
persons with advanced dementia from accessing the services of
assisted dying organizations. Independent of the current legal sit-
uation, would you personally support theuse of assisted suicide
(one form of physician-hastened death) for persons with ad-
vanced dementia?” Agreement withthe use of CDS was measured
as follows: “At the end oflife, it is possible to use medications,
administered until death, to relieve the symptoms of a person with
advanced dementia and put them into a permanent, artificial
sleep (continuous sedation). Would you personally support the
use of continuous sedation for persons with advanced dementia?”
For both questions, the participants were asked to select one of the
following response options: “completely agree,” “somewhat
agree,” “somewhat disagree,” “completely disagree,” or “do not
know.”
Other participant data assessed at baseline included demo-
graphics (age, gender, nationality [Swiss, German, and other], re-
ligion [Protestant or Catholic, other religion, no religion, and re-
fused to answer], and educational level [≥high school vs. other])
and whether participants had previously decided about the use of
antibiotics and/or artificial hydration for a person with advanced
dementia.
Analysis
The analyses were performed using R version 3.3.2
(Boston,MA, USA). Means with standard deviations (SDs) and
frequencies describe continuous and categorical variables, re-
spectively.
Logistic regression was used to examine the association be-
tween decision-maker type (surrogate vs. physician) and agree-
ment with the following practices in advanced dementia (out-
comes): (1) assisted dying and (2) CDS. For each practice, the
outcome was dichotomized as “agree” (i.e., “completely agree”
and “somewhat agree”) versus “disagree” (i.e., “completely dis-
agree” and “somewhat disagree”) and excluded “do not know”
responses. Covariates considered a priori to be possibly associ-
ated with supporting the use of assisted dying or CDS [13] in-
cluded participants’ demographic characteristics (age [dichoto-
mized at the median], gender, and religion [dichotomized as
“no religion” vs. “any,” excluding “refused to answer”]) and
prior decisions about the use of antibiotics and/or artificial hy-
dration in advanced dementia (dichotomized as “any prior
major treatment decision” vs. “none”). Bivariable analyses ex-
amined the unadjusted associations between each covariate and
the outcome. Variables associated with the outcome at p < 0.10
in the unadjusted analyses were entered into a multivariable
model. The final adjusted model included those variables sig-
nificantly associated with the outcome at p < 0.05. Adjusted odds
ratios (AORs) with 95% confidence intervals (CIs) were com-
puted. A sensitivity analysis examined whether the association
differed as a function of subgroup (relatives vs. proxies), since
perceptions might not have been homogeneous across sub-
groups.
Results
Subject Characteristics
Of the 3,860 individuals approached for participa-
tion, 254 (6.6%) contacted the research team indicating
their willingness to participate, and all were eligible for
enrollment. Prior to study completion, 15 participants
(5.9%) stopped responding to e-mails and/or phone
calls (physicians, n = 9/74 [12.2%]; surrogates, n = 6/180
[3.3%]) and 7 participants (2.8%) withdrew (physicians,
n = 1/74 [1.4%]; surrogates, n = 6/180 [3.3%]). The final
sample included the remaining 64 physicians and 168
surrogates.
The baseline characteristics of the physicians and sur-
rogates are shown in Table 1. The physicians’ mean age
(SD) was 50.6 years (9.9); 46.9% (n = 30/64) were women
and 72.1% (n = 44/61) were Swiss. The surrogates’ mean
age (SD) was 57.4 years (14.6); 68.9% (n = 115/167) were
women and 94.6% (n = 157/166) were Swiss. A total of
68.8% (n = 44/64) of the physicians and 65.5% (n =
108/165) of the surrogates reported being either Protes-
tant or Catholic. A total of 90.6% (n = 58/64) of the physi-
Table 1. Characteristics of the physicians and surrogates (n = 232)
Characteristics Physicians
(n = 64)
Surrogates
(n = 168)
Mean age ± SDa, years 50.6±9.9 57.4±14.6
Age >55 yearsb21 (32.8) 91 (54.5)
Femalea30 (46.9) 115 (68.9)
Nationalitya
Swiss 44 (72.1) 157 (94.6)
German 13 (21.3) 7 (4.2)
Other 4 (6.6) 2 (1.2)
Religiona
Protestant or catholic 44 (68.8) 108 (65.5)
Other 9 (14.1) 11 (6.7)
No religion 10 (15.6) 38 (23.0)
Refused to answer 1 (1.6) 8 (4.8)
Educationa
High school or higher 64 (100) 161 (97.0)
Any prior major treatment
decision in dementiaa, c 58 (90.6) 42 (25.3)
Values denote n (%) unless specified otherwise. a The total
number of missing values by characteristic was: age, n = 1; female,
n = 1; nationality, n = 5; religion, n = 3; education, n = 2; and any
prior major treatment decision, n = 2. bMedian age. c Decision-
makers reported whether they had or had not previously made any
major decision about the use of antibiotics and/or artificial
hydration for a person with advanced dementia.
End-of-Life Practices in Advanced
Dementia
7
Neurodegener Dis 2019;19:4–11
DOI: 10.1159/000499113
cians had previously made a major treatment decision for
a person with advanced dementia, whereas only 25.3%
(n= 42/166) of the surrogates had ever made such a deci-
sion.
Agreement with Assisted Dying
The distribution of physician responses about the use
of assisted dying in advanced dementia was as follows:
completely agree, 4.7% (n = 3/64); somewhat agree, 15.6%
(n = 10/64); somewhat disagree, 26.6% (n = 17/64); com-
pletely disagree, 50.0% (n = 32/64); and do not know,
3.1% (n = 2/64). The distribution among the surrogates
was: completely agree, 20.2% (n = 34/168); somewhat
agree, 26.8% (n = 45/168); somewhat disagree, 23.2% (n=
39/168); completely disagree, 22.6% (n = 38/168); and do
not know, 7.1% (n = 12/168).
In the unadjusted analyses, participant variables as-
sociated with agreement with assisted dying at p < 0.10
were: age > 55 years; any prior major treatment decision
in dementia; and being a surrogate (vs. a physician) (Ta-
ble 2). After multivariable adjustment, only being a sur-
rogate (vs. a physician) remained significantly associat-
ed with a higher likelihood of agreeing with the use of
assisted dying (AOR, 3.87; 95% CI: 1.94, 7.69). When
analyzing surrogate subgroups, agreement was signifi-
cantly more likely among relatives than among proxies
(34.6 vs. 16.0%; AOR, 2.00; 95% CI: 1.04, 3.83) (Appen-
dix, Table 1).
Agreement with CDS
The distribution of physician responses about the use
of CDS in advanced dementia was as follows: completely
agree, 20.3% (n = 13/64); somewhat agree, 31.3% (n =
20/64); somewhat disagree, 31.3% (n = 20/64); complete-
ly disagree, 6.3% (n = 4/64); and do not know, 10.9% (n =
7/64). The distribution among the surrogates was: com-
pletely agree, 12.6% (n = 21/167); somewhat agree, 29.3%
(n = 49/167); somewhat disagree, 28.7% (n = 48/167);
completely disagree, 18.6% (n = 31/167); and do not
know, 10.8% (n = 18/167).
The only covariate associated with a greater likeli-
hood of supporting the use of CDS at p ≤ 0.10 was “no
religion” (vs. “any”) (Table 3). After adjusting for reli-
gion, being a surrogate (vs. a physician) remained not
significantly associated with agreement with the use of
CDS (AOR, 0.69; 95% CI: 0.36, 1.29). When analyzing
surrogate subgroups, agreement was significantly
morelikely among relatives than among proxies (34.9
vs. 12.1%; AOR, 2.82; 95% CI: 1.41, 5.64) (Appendix,
Table 2).
Discussion
In this report, a minority of the physicians (20%) and
about half of the surrogates (47%) supported the use of
assisted dying for patients with advanced dementia. Sur-
Table 2. Association between decision-maker type and agreement with the use of assisted dying in advanced dementia
Decision-makers
with the
characteristic
(n = 218),
n (%)
Decision-makers
agreeing with assisted
dying (n = 92), n (%)
Odds ratioa for agreement
with assisted dying (95% CI)
characteristic
present
characteristic
absent
unadjusted adjusted
Characteristic
Surrogate (vs. physician) 156 (71.6) 79 (36.2) 13 (6.0) 3.87 (1.94, 7.69)d3.87 (1.94, 7.69)
Covariates
Age >55 yearsb106 (48.8) 54 (24.9) 38 (17.5) 1.99 (1.15, 3.45)d
Femaleb133 (61.3) 62 (28.6) 30 (13.8) 1.57 (0.90, 2.76)
No religion (vs. any)b45 (21.4) 24 (11.4) 67 (31.9) 1.67 (0.86, 3.24)
Any prior major treatment decision
in dementiab, c 97 (44.9) 29 (13.4) 61 (28.2) 0.41 (0.23, 0.71)d
Decision-makers: 42.2% (n = 92/218) agreed with the use of assisted dying in advanced dementia. aLogistic regression was used in
all analyses, and unadjusted and adjusted odds ratios were computed. bThe total number of missing values was: age, n = 1; female, n =
1; no religion, n = 8 (missing values, n = 3; refused, n = 5); and any prior major treatment decision, n = 2. cDecision-makers reported
whether they had or had not previously made any major decision about the use of antibiotics and/or artificial hydration for a person
with advanced dementia. dVariables that were significant at p < 0.10 in the bivariable analyses and entered into the multivariable model.
Loizeau/Cohen/Mitchell/Theill/Eicher/
Martin/Riese
Neurodegener Dis 2019;19:4–11
8
DOI: 10.1159/000499113
rogates were significantly more likely to agree with the
use of this practice than physicians. About half of the phy-
sicians (52%) and a slightly lower proportion of the sur-
rogates (42%) agreed with the use of CDS for these pa-
tients, which was not a significant difference between the
groups.
This study confirms and extends the limited existing
research on the perceptions of physicians and surrogates
about assisted dying in advanced dementia [11–15]. The
use of assisted dying has increased in countries where it
is legal (The Netherlands and Belgium) over the last de-
cade, but much less frequently for patients with demen-
tia than for those with terminal illnesses that do not im-
pact decision-making abilities [6, 10, 22]. Prior research
has shown that most physicians are opposed to the use
of this practice in dementia [12–15]. Irrespective of the
presence of an advance directive, physicians are reluctant
to perform physician-assisted suicide for patients lacking
decision-making capacity [12]. A Dutch study on 1,456
physicians found that physicians were more willing to
perform assisted dying for cancer patients with intact
cognition (85%) than for advanced dementia patients
who had requested euthanasia in an advance directive
prepared when they were still capable of making medical
decisions (33%) [14]. Similar proportions were found in
a recent Canadian study on 136 physicians [15]. Notably,
this study also showed that agreement increased to 71%
when the hypothetical scenario specified that the demen-
tia patients were in their last days of life. One chief con-
cern is that preferences change over time [23], and that
patients who imagine a future state with advanced de-
mentia as one not worth living, once in that state, may
appear to still retain a desire to live. Although surrogates
may share this concern, their more favorable view of the
use of assisted dying may be driven by a relatively great-
er aversion to their loved one experiencing the poor
quality of life, suffering, and indignities of advanced de-
mentia [11, 13]. The greater likelihood of surrogates
agreeing to assistance in dying is due to the fact that their
participation, unlike physicians, is not required to com-
plete the act.
Our findings build on the very limited data on the use
of CDS in advanced dementia [3, 15, 20], and they reveal
that there is no consensus among either physicians or
surrogates about the appropriateness of using CDS in
this population. This finding corroborates qualitative
research showing that both groups of key decision-mak-
ers have mixed feelings about the use of this practice
among terminally ill patients [24, 25]. To date, CDS has
predominantly been used and studied in advanced can-
cer, a condition where decision-making capacity re-
mains intact and the dying process is most commonly
Table 3. Association between decision-maker type and agreement with the use of continuous deep sedation in advanced dementia
Decision-makers
with the
characteristic
(n = 206), n (%)
Decision-makers agreeing
with continuous deep
sedation (n = 103), n (%)
Odds ratioa for agreement with
continuous deep sedation (95% CI)
characteristic
present
characteristic
absent
unadjusted adjusted
Characterstic
Surrogate (vs. physician) 149 (72.3) 70 (34.0) 33 (16.0) 0.64 (0.35, 1.19) 0.69 (0.36, 1.29)e
Covariates
Age >55 yearsb101 (49.3) 55 (26.8) 48 (23.4) 1.39 (0.81, 2.42)
Femaleb128 (62.4) 64 (31.2) 39 (19.0) 0.97 (0.55, 1.71)
No religion (vs. any)b41 (20.9) 27 (13.8) 74 (37.8) 2.11 (1.03, 4.33)d2.19 (1.06, 4.51)
Any prior major treatment
decision in dementiab, c 88 (43.1) 47 (23.0) 55 (27.0) 1.27 (0.73, 2.22)
Decision-makers: 50.0% (n = 103/206) agreed with the use of continuous deep sedation in advanced dementia. aLogistic regression
was used in all analyses, and unadjusted and adjusted odds ratios were computed. bThe total number of missing values was: age, n = 1;
female, n = 1; no religion, n = 10 (missing values, n = 2; refused, n = 8); and any prior major treatment decision, n = 2. cDecision-makers
reported whether they had or had not previously made any major decision about the use of antibiotics and/or artificial hydration for a
person with advanced dementia. dThe only variable that was significant at p < 0.10 in the bivariable analyses and entered into the
multivariable model. eAfter adjusting for religion, the association between being a surrogate and agreement with the use of continuous
deep sedation remained not significant.
End-of-Life Practices in Advanced
Dementia
9
Neurodegener Dis 2019;19:4–11
DOI: 10.1159/000499113
accompanied by intractable physical pain and suffering
[16, 19]. In contrast to these patients, advanced demen-
tia patients cannot participate in decision-making or re-
port the source of their discomfort due to their serious
cognitive impairments [5]. Therefore, it may be chal-
lenging for physicians to assess the intensity of a pa-
tient’s suffering, making it difficult to justify the use of
CDS in advanced dementia. Furthermore, pain and agi-
tation in advanced dementia are often caused by other
medical complications, such as infections, and may be
effectively controlled with standard palliative medica-
tions [3, 4].
This study has several limitations that merit discus-
sion. First, our participation rate was low, and thus our
findings cannot be generalized to eligible nonpartici-
pants. Second, generalizability is also limited to the
Swiss German region of Switzerland and to hypothetical
situations. The perceptions about end-of-life practices
may vary in other regions of Switzerland and in real-life
situations. In addition, it cannot be ruled out that physi-
cians unaffiliated with associations promoting palliative
care may hold different views on end-of-life practices.
Third, the questions may not have been detailed enough
to detect nuances in participants’ perceptions [26]. It is
likely that participants’ opinions would have been dif-
ferent if they had been given more information, such as
a statement indicating that the patient had made a writ-
ten request for assisted dying before losing decision-
making capacity. Likewise, providing participants with
background information and definitions of different
end-of-life practices such as physician-assisted suicide
and euthanasia in the case of incompetent patients
would likely have altered the responses. Lastly, the sta-
tistical power may have been insufficient to detect sig-
nificant differences in secondary outcomes, notably dif-
ferences in perceptions about CDS. Future work should
replicate these findings in a clinical cluster-randomized
trial.
This study expands upon the limited existing litera-
ture on the perceptions of physicians and surrogates
about the use of assisted dying and CDS for patients with
advanced dementia. Physicians were more opposed to
the use of assisted dying than were the surrogates, and
about half of the participants in both groups perceived
CDS as an appropriate option in advanced dementia. In
general, there is a clear need to improve end-of-life care
in this patient population. However, due to the practical,
ethical, and legal issues associated with performing as-
sisted dying and CDS for patients who lack decision-
making capacity, it remains unclear whether these prac-
tices would help accomplish this goal. Early goals of care
discussions can promote decisions that align with the pa-
tient’s preferences and the use of standard palliative
medications [27, 28].
Acknowledgments
The investigators wish to thank the data management
team(Daniel Gfeller and Kathrin Heiniger), the organizations’
staff who helped with recruitment, Marius C. Vollberg for sta-
tistical guidance, and the participants who gave their time to this
study.
Statement of Ethics
The ethics commission of the canton of Zurich approved the
study’s conduct (KEK-ZH-No. 2015-0626). All participants pro-
vided written informed consent. The authors have no ethical con-
flicts to disclose.
Disclosure Statement
The authors have no conflicts of interest to declare.
Funding Sources
This work was supported by the following grants: the Research
in Palliative Care funding program (PC 22/14), which was offered
jointly by the Swiss Academy of Medical Sciences, the Stanley
Thomas Johnson Foundation, and the Gottfried and Julia Bangert-
er-Rhyner Stiftung; the Swiss National Science Foundation
(P1ZHP3_171747 to A.J.L.); and the United States National Insti-
tutes of Health – National Institute on Aging (NIH-NIA R01
AG032982, NIH-NIA R01 AG043440, and NIH-NIA
K24AG033640 to S.L.M.).
Author Contributions
F.R. and A.J.L. had full access to all data in the study and take
responsibility for the integrity of the data and the accuracy of the
data analyses. Study concept and design: A.J.L., N.T., S.E., M.M.,
and F.R.; acquisition of data: A.J.L., S.M.C., S.L.M., N.T., S.E.,
M.M., and F.R.; analyses and interpretation of data: A.J.L., S.M.C.,
S.L.M., N.T., S.E., M.M., and F.R.; drafting of manuscript: A.J.L.,
S.M.C., S.L.M., and F.R.; critical revision of manuscript for impor-
tant intellectual content: A.J.L., S.M.C., S.L.M., N.T., S.E., M.M.,
and F.R.; statistical analyses: A.J.L., S.M.C., S.L.M., N.T., M.M.,
and F.R.; administrative, technical, or material support: A.J.L.,
S.E., and F.R.; and study supervision: A.J.L., N.T., S.E., M.M., and
F.R.
Loizeau/Cohen/Mitchell/Theill/Eicher/
Martin/Riese
Neurodegener Dis 2019;19:4–11
10
DOI: 10.1159/000499113
Appendix
Table 1. Association between being a relative versus a proxy and agreement with the use of assisted dying in advanced dementia
Surrogates with the
characteristic
(n = 156), n (%)
Surrogates agreeing with assisted
dying (n = 79), n (%)
Odds ratioa for
agreement with assisted
dying (95% CI)
characteristic present characteristic absent adjustedb
Characteristic
Relative (vs. proxy) 94 (60.3) 54 (34.6) 25 (16.0) 2.00 (1.04, 3.83)
Surrogates: 50.6% (n = 79/156) agreed with the use of assisted dying in advanced dementia. aLogistic regression was used in all
analyses, and unadjusted and adjusted odds ratios were computed. bAfter accounting for age, gender, religion, and prior decisions about
antibiotics/artificial hydration use, no variables were significant at p < 0.10 in the bivariable analyses, except for the association between
being a relative versus proxy and agreement with the use of assisted dying in advanced dementia.
Table 2. Association between being a relative versus a proxy and agreement with the use of continuous deep sedation in advanced
dementia
Surrogates with the
characteristic
(n = 149), n (%)
Surrogates agreeing with
continuous deep sedation (n = 70), n (%)
Odds ratioa for agreement
with continuous deep
sedation (95% CI)
characteristic present characteristic absent adjustedb
Characteristic
Relative (vs. proxy) 92 (61.7) 52 (34.9) 18 (12.1) 2.82 (1.41, 5.64)
Surrogates: 47.0% (n = 70/149) agreed with the use of continuous deep sedation in advanced dementia. aLogistic regression was used
in all analyses, and unadjusted and adjusted odds ratios were computed. bAfter accounting for age, gender, religion, and prior decisions
about antibiotics/artificial hydration use, no variables were significant at p < 0.10 in the bivariable analyses, except for the association
between being a relative versus proxy and agreement with the use of continuous deep sedation in advanced dementia.
References
1 Federal Statistical Office FSO [Internet].
Cause of death statistics: death and its
main causes in Switzerland, 2015 [accessed
2018 Apr 26]. Available from: https://www.
bfs.admin.ch/bfsstatic/dam/assets/3742832/
master.
2 World Health Organization [Internet].
Theepidemiology and impact of dementia:
current state and future trends [accessed
2018 Apr 26]. Available from: http://www.
who.int/mental_health/neurology/demen-
tia/dementia_thematicbrief_epidemiology.
pdf.
3 Hendriks SA, Smalbrugge M, Hertogh CM,
van der Steen JT. Dying with dementia: symp-
toms, treatment, and quality of life in the last
week of life. J Pain Symptom Manage. 2014
Apr; 47(4): 710–20.
4 Mitchell SL, Teno JM, Kiely DK, Shaffer ML,
Jones RN, Prigerson HG, et al. The clinical
course of advanced dementia. N Engl J Med.
2009 Oct; 361(16): 1529–38.
5 van der Steen JT, Deliens L, Koopmans RT,
Onwuteaka-Philipsen BD. Physicians’ per-
ceptions of suffering in people with dementia
at the end of life. Palliat Support Care. 2017
Oct; 15(5): 587–99.
6 Emanuel EJ, Onwuteaka-Philipsen BD, Ur-
win JW, Cohen J. Attitudes and practices of
euthanasia and physician-assisted suicide in
the United States, Canada, and Europe.
JAMA. 2016 Jul; 316(1): 79–90.
7 Radbruch L, Leget C, Bahr P, Müller-Busch C,
Ellershaw J, de Conno F, et al.; Board Mem-
bers of EAPC. Euthanasia and physician-as-
sisted suicide: a white paper from the Euro-
pean Association for Palliative Care. Palliat
Med. 2016 Feb; 30(2): 104–16.
8 Swiss Academy of Medical Sciences [Inter-
net]. Medical ethical guidelines: management
of dying and death [accessed 2018 Jun 7].
Available from: https://www.samw.ch/en/
Publications/Medical-ethical-Guidelines.
html.
9 Li M, Watt S, Escaf M, Gardam M, Heesters
A, O’Leary G, et al. Medical assistance in dy-
ing – implementing a hospital-based program
in Canada. N Engl J Med. 2017 May; 376(21):
2082–8.
10 Neil F [Internet]. Assisted dying practice in
Benelux: whitepaper 1 [accessed 2018 Apr
15]. Available from: http://www.dyingfor-
choice.com/resources/fact-files/assisted-dy-
ing-benelux-whitepaper-1.
11 Tomlinson E, Spector A, Nurock S, Stott J.
Euthanasia and physician-assisted suicide in
dementia: a qualitative study of the views of
former dementia carers. Palliat Med. 2015
Sep; 29(8): 720–6.
12 Kouwenhoven PS, Raijmakers NJ, van
Delden JJ, Rietjens JA, van Tol DG, van de
Vathorst S, et al. Opinions about euthanasia
and advanced dementia: a qualitative study
among Dutch physicians and members of the
general public. BMC Med Ethics. 2015 Jan;
16(7): 7.
End-of-Life Practices in Advanced
Dementia
11
Neurodegener Dis 2019;19:4–11
DOI: 10.1159/000499113
13 Tomlinson E, Stott J. Assisted dying in de-
mentia: a systematic review of the interna-
tional literature on the attitudes of health pro-
fessionals, patients, carers and the public, and
the factors associated with these. Int J Geriatr
Psychiatry. 2015 Jan; 30(1): 10–20.
14 Bolt EE, Snijdewind MC, Willems DL, van der
Heide A, Onwuteaka-Philipsen BD. Can phy-
sicians conceive of performing euthanasia in
case of psychiatric disease, dementia or being
tired of living? J Med Ethics. 2015 Aug; 41(8):
592–8.
15 Bravo G, Rodrigue C, Arcand M, Downie J,
Dubois MF, Kaasalainen S, et al. Quebec phy-
sicians’ perspectives on medical aid in dying
for incompetent patients with dementia. Can
J Public Health. 2018 Dec; 109(5-6): 729–39.
16 Miccinesi G, Rietjens JA, Deliens L, Paci E,
Bosshard G, Nilstun T, et al.; EURELD Con-
sortium. Continuous deep sedation: physi-
cians’ experiences in six European countries.
J Pain Symptom Manage. 2006 Feb; 31(2):
122–9.
17 Ziegler S, Schmid M, Bopp M, Bosshard G,
Puhan MA. Continuous deep sedation until
death – a Swiss death certificate study. J Gen
Intern Med. 2018 Jul; 33(7): 1052–9.
18 Cherny NI, Radbruch L; Board of the Euro-
pean Association for Palliative Care. Europe-
an Association for Palliative Care (EAPC)
recommended framework for the use of seda-
tion in palliative care. Palliat Med. 2009 Oct;
23(7): 581–93.
19 McCarthy EP, Phillips RS, Zhong Z, Drews
RE, Lynn J. Dying with cancer: patients’ func-
tion, symptoms, and care preferences as death
approaches. J Am Geriatr Soc. 2000 May; 48(5
Suppl):S110–21.
20 Anquinet L, Rietjens JA, Vandervoort A, van
der Steen JT, Vander Stichele R, Deliens L, et
al. Continuous deep sedation until death in
nursing home residents with dementia: a case
series. J Am Geriatr Soc. 2013 Oct; 61(10):
1768–76.
21 Loizeau AJ, Theill N, Cohen SM, Eicher S,
Mitchell SL, Meier S, et al. Fact Box decision
support tools reduce decisional conflict about
antibiotics for pneumonia and artificial hy-
dration in advanced dementia: a randomized
controlled trail. Age Ageing. 2019 Jan; 48(1):
67–74.
22 Dierickx S, Deliens L, Cohen J, Chambaere K.
Euthanasia for people with psychiatric disor-
ders or dementia in Belgium: analysis of offi-
cially reported cases. BMC Psychiatry. 2017
Jun; 17(1): 203.
23 Emanuel EJ, Fairclough DL, Emanuel LL. At-
titudes and desires related to euthanasia and
physician-assisted suicide among terminally
ill patients and their caregivers. JAMA. 2000
Nov; 284(19): 2460–8.
24 Bruinsma SM, Rietjens JA, Seymour JE, An-
quinet L, van der Heide A. The experiences of
relatives with the practice of palliative seda-
tion: a systematic review. J Pain Symptom
Manage. 2012 Sep; 44(3): 431–45.
25 Ziegler S, Merker H, Schmid M, Puhan MA.
The impact of the inpatient practice of con-
tinuous deep sedation until death on health-
care professionals’ emotional well-being: a
systematic review. BMC Palliat Care. 2017
May; 16(1): 30.
26 Magelssen M, Supphellen M, Nortvedt P, Ma-
terstvedt LJ. Attitudes towards assisted dying
are influenced by question wording and or-
der: a survey experiment. BMC Med Ethics.
2016 Apr; 17(1): 24.
27 Mitchell SL, Palmer JA, Volandes AE,
Hanson LC, Habtemariam D, Shaffer ML.
Level of care preferences among nursing
home residents with advanced dementia. J
Pain Symptom Manage. 2017 Sep; 54(3):
340–5.
28 Loizeau AJ, Shaffer ML, Habtemariam DA,
Hanson LC, Volandes AE, Mitchell SL. Asso-
ciation of prognostic estimates with burden-
some interventions in nursing home residents
with advanced dementia. JAMA Intern Med.
2018 Jul; 178(7): 922–9.