How do general end-of-life treatment goals and values relate to specific treatment preferences? A population-based study

Abstract

Background:

There is a lack of research on the relationship between general end-of-life goals and values and preferences for specific life-sustaining treatments.

Aim:

To examine agreement between Dutch older people’s general end-of-life goals and specific life-sustaining treatment preferences.

Design:

Participants identified general end-of-life goals in an interview and preferences for four life-sustaining treatments in hypothetical cancer and dementia scenarios in a separate questionnaire. Agreement between general goals and specific treatment preferences was calculated.

Setting/participants:

In total, 1818 older people from 11 representative Dutch municipalities participated in the study.

Results:

In total, 1168 (response rate 73%) answered questions on general end-of-life and specific treatment preferences. Agreement between a desire to live as long as possible, irrespective of health problems, and a preference for life-sustaining treatments ranged from 51% to 76% in cancer and 41% to 60% in dementia scenarios, depending on the treatment. Agreement between a desire for a shorter life, if without major health problems, and a preference to forgo treatments ranged from 61% to 79% in cancer and 75% to 88% in dementia scenarios.

Conclusion:

For a sizable minority of participants, specific treatment preferences did not agree with their general end-of-life goals. The more frequent desire to forgo treatments in case of dementia than cancer suggests that physical deterioration is more acceptable than cognitive decline. The findings underline the importance of discussing general care goals, different end-of-life scenarios and the risks and burdens of treatments to frame discussions of more specific treatment preferences.

Keywords

Life support care decision-making palliative care health communication advance directives advance care planning

What is already known about the topic?

There is a lack of research on the relationship between general end-of-life goals and values and preferences for specific life-sustaining treatments.

What this paper adds?

This paper demonstrates that for a sizable minority of Dutch older people, specific life-sustaining treatment preferences did not agree with their general end-of-life goals.

Implications for practice, theory, or policy

Specific life-sustaining treatment preferences should be discussed within the context of general care goals, different end-of-life scenarios and the risks and burdens of treatments.

Introduction

End-of-life treatment preferences have typically been examined in two ways: determining specific life-sustaining treatment preferences^1–3 and identifying general goals or outcomes of care,^4–6 such as maintaining quality of life. Determining treatment preferences involves people in decisions about future care, strengthening autonomy. However, people may not thoroughly understand the risks and burdens of treatments. Moreover, because it is impossible to foresee all potential circumstances and treatments, knowledge of specific treatment preferences is often insufficient to guide all end-of-life decisions.⁷ Questions also remain about people’s ability, when relatively healthy, to predict preferred treatments during future incapacity.^8–10 In light of these shortcomings, some authors recommend that determining general end-of-life values and goals may be as, or more, useful in guiding decisions;^6,11,12 such information could orientate health care practitioners, who are perhaps better able to judge the potential outcomes of interventions.

Few studies, however, have investigated the relationship between general end-of-life goals and values and specific life-sustaining treatment preferences. Two studies that did found general care goals to be poor predictors of treatment preferences.^13,14 Schneiderman et al.¹³ examined the relationship between choice of three brief care instructions (general care goals) and specific life-sustaining treatment preferences. However, 45 participants chose the first care instruction, 11 the second and just 2 the third: limiting comparison and statistical power.¹³ Fischer et al.¹⁴ compared general care goals along with specific treatment preferences in different health scenarios. This study, however, was only conducted among physicians.¹⁴

Drawing on data from the Longitudinal Ageing Study Amsterdam (LASA)—a population-based study¹⁵—this article examines the relationship between older people’s general end-of-life goals and specific end-of-life treatment preferences: specifically, the relationship between general preferences for quantity versus quality of life and preferences for artificial hydration and nutrition, antibiotics, cardiopulmonary resuscitation (CPR), and mechanical respiration in hypothetical cancer and dementia scenarios. If general goals agree with specific treatment preferences, participants who preferred to have a longer life, irrespective of health problems, would be expected to prefer life-sustaining treatments, whereas those who prefer a shorter life, if without major health problems, would be expected to favor forgoing treatments.

Methods

Sample

To date, the LASA study incorporates two nationally representative cohorts of older people, selected via random stratified sampling from 11 Dutch municipality registers.¹⁵ Men and older participants were deliberately oversampled (to maintain members of these high-mortality groups throughout the study).¹⁵ The first cohort (aged 55–85 years) was recruited in 1992/1993, and the second (aged 55–64 years) in 2002/2003. Data were collected at 3-year intervals through a face-to-face interview, self-administered questionnaire, and medical interview.¹⁵ Described in more detail elsewhere, the study has a low drop-out rate, most of which is due to mortality.¹⁵ The Medical Ethics Committee of the VU University Medical Center approved the study, and participants gave written informed consent.

Measurement instruments

Socio-demographics (age, gender, religious affiliation, and education) and health characteristics (presence of chronic disease) were ascertained from population registries and face-to-face interviews, respectively. General end-of-life goals were measured during the face-to-face interview via a question on quantity versus quality of life (Box 1). During this interview, a self-administered questionnaire was left for participants to complete and hand-in during a subsequent medical interview. In the questionnaire, participants were asked to indicate preferences for artificial hydration and nutrition, antibiotics, CPR, and mechanical respiration in case of cancer and dementia (Box 1). Vignettes concerning cancer and dementia were chosen because previous research has shown distinct life-sustaining treatment preferences if faced with physical or cognitive decline.^16–19 Questions were piloted with 10 older people looking at whether the questions were understandable and acceptable for the respondent group. For this, the “think aloud” method was used.²⁰ Where necessary questions were adjusted.

Box 1.

Questions on quality versus quantity of life and medical end-of-life treatment preferences.

Question on general care goals (asked during the main face-to-face interview)

When you think about the future, which do you prefer:

To live as old as possible, irrespective of health problems;

To have a shorter life, if without major health problems.

Question on specific end-of-life treatment preferences (asked in the self-administered questionnaire)

Scenario 1.

You are diagnosed with cancer, which has spread throughout your body. The disease is incurable. For some days you have not been responsive to any form of communication. You show clear signs of severe pain and agitation.

Scenario 2.

You suffer from dementia and no longer recognize your family or friends. You refuse to eat and drink and you are becoming increasingly withdrawn. It is no longer possible to discuss treatment options with you.

Preferences regarding four medical end-of-life treatments

	Yes	Probably yes	Probably no	No
Would you, in the above situation, wish to receive artificial hydration and nutrition (through a nasal gastric tube)?	1	2	3	4
Would you, in the above situation, wish to be given antibiotics if you got pneumonia?	1	2	3	4
Would you, in the above situation, want CPR in the event of a cardiac arrest?	1	2	3	4
Would you, in the above situation, want mechanical respiration?	1	2	3	4

Analysis

Sample characteristics, treatment preferences, and agreement with general goals were described using true numbers and adjusted percentages or means (weighted for sex and age, standard: Dutch population 01/01/2009²¹). Differences in treatment preferences between participants with different general end-of-life goals were assessed using Pearson’s chi-squared test.

Results

In the 2008/2009 data collection cycle, there were 1818 participants (985 from the first cohort and 833 from the second cohort), and 1601 participants participated in a face-to-face interview and were subsequently invited to participate in the self-administered questionnaire. Of these, 1168 (73%) answered questions on specific treatment preferences in the questionnaire and a question about general end-of-life goals during the interview. Participants who did not answer the questionnaire (n = 118), or who did not respond to end-of-life preferences questions (n = 315), were older (p < 0.001), more frequently female (p < 0.001), church members (p < 0.001), and had lower education (p < 0.001). Ethnicity did not differ between the two groups; however, only 1% were non-White Dutch. Participants’ socio-demographic and health characteristics are shown in Table 1.

Table 1.

Sample characteristics (n = 1168).

Socio-demographic characteristics		n	Adjusted sample mean (SD) or proportion of sample
Age (range 61–99 years)		1168	70.21 (7.49)
Gender	Male	560	50
	Female	608	50
Ethnicity	Dutch/the Netherlands	1158	99
	Other	10	1
Religious affiliation	None	519	46
	Protestant	317	26
	Catholic	316	27
	Other	16	2
Education	Low	254	21
	Medium	416	36
	High	498	44
Chronic disease	No chronic disease	266	25
	Has one or more chronic diseases	902	75

SD: standard deviation.

Missing values: subjective health 2 (0.2%); pain 7 (0.6%); depressive symptoms 1 (0.1%).

True n, adjusted valid percentages, and adjusted mean reported (adjusted = weighted for age and sex). Percentages have been rounded to whole numbers, therefore some totals are not exactly 100%.

End-of-life preferences

In total, 13.5% (n = 156) of the participants wanted to live as long as possible, irrespective of health problems, whereas 86.5% (n = 1012) preferred a shorter life, if without major health problems. There were significant differences (p < 0.001) in preferences for artificial hydration and nutrition, antibiotics, CPR, and mechanical respiration in hypothetical cases of cancer and dementia between participants who sought to extend life and who sought to maximize quality of life (Tables 2 and 3).

Table 2.

Relationship between preferences for four life-sustaining treatments and longevity/health preferences in case of cancer (n = 1168).

Cancer scenario		All participants		Preference for quality versus quantity of life
Medical end-of-life treatment	Preference for treatment	n	%^a	As old as possible, irrespective of health problems		Shorter life, if without major health problems		p value*
				n	%^a	n	%^a
Artificial hydration and nutrition	Yes/probably yes	481	40	107	69	374	36	p < 0.001
Artificial hydration and nutrition	No/probably no	687	60	49	31	638	64
Antibiotics	Yes/probably yes	514	43	120	76	394	38	p < 0.001
	No/probably no	654	57	36	24	618	62
Cardiopulmonary resuscitation	Yes/probably yes	311	25	83	51	228	21	p < 0.001
Cardiopulmonary resuscitation	No/probably no	857	75	73	49	784	79
Mechanical respiration	Yes/probably yes	507	43	109	69	398	39	p < 0.001
	No/probably no	661	57	47	31	614	61
Preference for all treatments	Wants all treatments	254	21	73	46	181	17
Preference for all treatments	Wants one, two, or three treatments	399	35	59	38	340	34	p < 0.001
	Wants no treatments	515	45	24	17	491	49

Percentages are weighted to population, age, and sex distribution.

Test for differences: Pearson’s chi-squared.

Table 3.

Relationship between preferences for four life-sustaining treatments and longevity/health preferences in case of dementia (n = 1168).

Dementia scenario		All participants		Preference for quality versus quantity of life
Medical end-of-life treatment	Preference for treatment	n	%^a	As old as possible, irrespective of health problems		Shorter life, if without major health problems		p value*
				n	%^a	n	%^a
Artificial hydration and nutrition	Yes/probably yes	290	25	86	56	204	20	p < 0.001
	No/probably no	878	75	70	44	808	80
Antibiotics	Yes/probably yes	325	27	95	60	230	22	p < 0.001
	No/probably no	843	73	61	40	782	78
Cardiopulmonary resuscitation	Yes/probably yes	197	16	66	41	131	12	p < 0.001
Cardiopulmonary resuscitation	No/probably no	971	84	90	59	881	88
Mechanical respiration	Yes/probably yes	355	30	94	60	261	25	p < 0.001
Mechanical respiration	No/probably no	813	70	62	40	751	75
Preference for all treatments	Wants all treatments	175	14	60	38	115	11	p < 0.001
Preference for all treatments	Wants one, two, or three treatments	257	22	50	32	207	21
	Wants no treatments	736	64	46	30	690	69

Percentages are weighted to population, age, and sex distribution.

Test for differences: Pearson’s chi-squared.

Agreement between a preference for a longer life and the desire for a treatment ranged from 51% (CPR) to 76% (antibiotics) in the cancer scenario (Table 2) and from 41% (CPR) to 60% (antibiotics and mechanical respiration) in the dementia scenario (Table 3). Of those who preferred a longer life, 46% wanted all life-sustaining treatments, 38% preferred one to three treatments, and 17% wanted no treatments in the case of cancer (Table 2). In the case of dementia, of those who preferred a longer life, 38% wanted all treatments, 32% wanted one to three treatments, and 30% wanted no treatments (Table 3).

Agreement between a preference for a shorter, better quality life and the desire to forgo treatments ranged from 61% (mechanical respiration) to 79% (CPR) in the cancer scenario (Table 2) and from 75% (mechanical respiration) to 88% (CPR) in the dementia scenario (Table 3). Of those who preferred a shorter life, 17% wanted all life-sustaining treatments, 34% preferred one to three treatments, and 49% wanted no treatments in the cancer scenario (Table 2). In the dementia scenario, of those who preferred a shorter life, 11% wanted all treatments, 21% wanted one to three treatments, and 69% wanted no treatments (Table 3).

Discussion

This study reveals a majority agreement between general end-of-life goals and specific treatment preferences. For a sizable minority of participants, however, general goals did not agree with specific life-sustaining treatment preferences. Poor understanding of the risks and burdens of life-sustaining treatments or different understandings of their purpose may cause disagreement between general and specific end-of-life preferences. For example, a qualitative study²² revealed that patients often perceived life-sustaining treatments differently than their physicians—as acceptable to meet short-term goals (e.g. to reassess their medical condition or to say goodbye to loved ones) but not long-term dependence.

Regardless of general end-of-life goals, more participants preferred forgoing treatments in case of dementia compared with cancer. This suggests that physical deterioration is more acceptable than cognitive decline and supports findings from previous research.^16
–19,23

It also supports the interpretation that patients may view treatments as acceptable for meeting short-term goals, which are easier to make in physical than cognitive decline.

In the dementia scenario, agreement between general care goals and specific treatment preferences was highest among those who preferred a shorter, better quality life. This finding resonates with longitudinal research showing preferences for refusing life-sustaining treatments to be more stable over time than preferences for accepting treatments.^24
–26

Strengths of the study include the use of a large population-based representative sample and the separate measurement of general end-of-life goals and specific treatment preferences (minimizing question order²⁷ and item context effects²⁸). There were, however, limitations. Participants were not provided descriptions of treatments or associated burdens; therefore, it is unclear how informed preference choices were. The study also reports the preferences of older people, rather than people nearing death. Future research might recruit participants according to specific end-of-life illness trajectories,^29,30 and explore participants’ experiences of treatments, to differentiate between hypothetical and personal end-of-life scenarios.

Non-responders’ and respondents’ characteristics differed, reflecting the challenges older and less-educated participants face answering self-administered questionnaires or the confronting nature of the questions. Patients with these characteristics frequently receive substandard patient–physician communication,^31,32 future research examining reasons for non-response might help identify patient-based barriers to end-of-life communication.

Conclusion

Although general end-of-life goals were predictive of specific life-sustaining treatment preferences for the majority of participants, for a sizable minority general goals and specific treatment preferences did not agree. Advance care planning involves patients, families and health care practitioners considering end-of-life values and goals and articulating preferences.³³ This study underlines the importance of promoting a stepwise process of discussing general care goals before specific treatment preferences to frame discussions within the context of general goals, different end-of-life scenarios and the risks and burdens of treatments. A Dutch initiative “Guide to timely talk about the end of life,” for example, promotes and provides practical tips on general practitioner (GP)–patient end-of-life discussions for anyone who would like them, including older people and those with life limiting illnesses and covering both general and personal end-of-life scenarios, in GP³⁴ and patient³⁵ leaflets.

Footnotes

Acknowledgements

Collaborators (on behalf of EURO IMPACT): Lieve Van den Block^a, Zeger De Groote^a, Sarah Brearley^e, Augusto Caraceni^g,h, Joachim Cohen^a, Anneke Francke^b, Richard Harding^c,d, Irene J Higginson^c,d, Stein Kaasa^f, Karen Linden^k, Guido Miccinesiⁱ, Bregje Onwuteaka-Philipsen^b, Koen Pardon^a, Roeline Pasman^b, Sophie Pautex^j, Sheila Payne^e, Luc Deliens^a,b.

EURO IMPACT aims to develop a multidisciplinary, multi-professional, and inter-sectorial educational and research training framework for palliative care research in Europe. EURO IMPACT is coordinated by Prof Luc Deliens and Prof Lieve Van den Block of the ^aEnd-of-Life Care Research Group, Ghent University and Vrije Universiteit Brussel, Brussels, Belgium. Other partners are: ^bEMGO Institute for health and care research, VU University Medical Center, Amsterdam, The Netherlands; ^cCicely Saunders Institute, King’s College London, London, UK; ^dCicely Saunders International, London, UK; ^eInternational Observatory on End-of-Life Care, Lancaster University, Lancaster, UK; ^fNorwegian University of Science and Technology, Trondheim, Norway; ^gEAPC Research Network, Trondheim, Norway; ^hFondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy; ⁱCancer Research and Prevention Institute, Florence, Italy; ^jEUGMS European Union Geriatric Medicine Society, Geneva, Switzerland; ^kSpringer Science and Business Media, Houten, The Netherlands. The authors would also like to thank all the participants, researchers, and fieldworkers involved in the LASA study.

Declaration of conflicting interests

The authors declare that there is no conflict of interest.

Funding

Ministry of Health, Welfare and Sports and VU University provided grants to The Longitudinal Aging Study Amsterdam, and European Union Seventh Framework Programme (FP7/2007-2013, under grant agreement no. 264697) provided grant to EURO IMPACT.

References

Philippart

Vesin

Bruel

. The ETHICA study (part I): elderly’s thoughts about intensive care unit admission for life-sustaining treatments. Intens Care Med 2013; 39: 1565–1573.

Barnato

Fisher

ES.

Racial and ethnic differences in preferences for end-of-life treatment. J Gen Intern Med 2009; 24(6): 695–701.

Gramelspacher

Zhou

Hanna

. Preferences of physicians and their patients for end-of-life care. J Gen Intern Med 1997; 12(6): 346–351.

Mack

Weeks

Wright

. End-of-life discussions, goal attainment, and distress at the end of life: predictors and outcomes of receipt of care consistent with preferences. J Clin Oncol 2010; 28(7): 1203–1208.

Davison

SN.

End-of-life care preferences and needs: perceptions of patients with chronic kidney disease. Clin J Am Soc Nephrol 2010; 5(2): 195–204.

Hawkins

Ditto

Danks

. Micromanaging death: process preferences, values, and goals in end-of-life medical decision making. Gerontologist 2005; 45(1): 107–117.

Teno

Licks

Lynn

. Do advance directives provide instructions that direct care? SUPPORT Investigators. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment. J Am Geriatr Soc 1997; 45(4): 508–512.

Fried

Byers

Gallo

. Prospective study of health status preferences and changes in preferences over time in older adults. Arch Intern Med 2006; 166(8): 890–895.

Sahm

Will

Hommel

Attitudes towards and barriers to writing advance directives amongst cancer patients, healthy controls, and medical staff. J Med Ethics 2005; 31(8): 437–440.

10.

Jox

Krebs

Bickhardt

. How strictly should advance decisions be followed? The patients’ opinion. Palliat Med 2008; 22(5): 675–676.

11.

Andorno

Biller-Andorno

Brauer

Advance health care directives: towards a coordinated European policy?

Eur J Health Law 2009; 16(3): 207–227.

12.

Doukas

McCullough

LB.

The values history: the evaluation of the patient’s values and advance directives. J Fam Pract 1991; 32(2): 145–153.

13.

Schneiderman

Pearlman

Kaplan

. Relationship of general advance directive instructions to specific life-sustaining treatment preferences in patients with serious illness. Arch Intern Med 1992; 152(10): 2114–2122.

14.

Fischer

Alpert

Stoeckle

. Can goals of care be used to predict intervention preferences in an advance directive? Arch Intern Med 1997; 157(7): 801–807.

15.

Huisman

Poppelaars

van der Horst

. Cohort profile: the longitudinal aging study Amsterdam. Int J Epidemiol 2011; 40(4): 868–876.

16.

Pearlman

Cain

Starks

. Preferences for life-sustaining treatments in advance care planning and surrogate decision making. J Palliat Med 2000; 3(1): 37–48.

17.

Van Mil

AHM

van Klink

RCJ

Huntjens

. Cardiopulmonary resuscitation preferences in Dutch community-dwelling and hospitalized elderly people: an interaction between gender and quality of life. Med Decis Making 2000; 20(4): 423–429.

18.

Coppola

Bookwala

Ditto

. Elderly adults’ preferences for life-sustaining treatments: the role of impairment, prognosis, and pain. Death Stud 1999; 23(7): 617–634.

19.

Phipps

True

Harris

Approaching the end of life: attitudes, preferences, and behaviors of African-American and white patients and their family caregivers. J Clin Oncol 2003; 21(3): 549–554.

20.

Collins

Pretesting survey instruments: an overview of cognitive methods. Qual Life Res 2003; 12(3): 229–238.

21.

Statistics Netherlands. Dutch population, sex, age and marital status, http://statline.cbs.nl/StatWeb/?LA=en (accessed 7 December 2010).

22.

Rodriguez

Young

AJ.

Patients’ and healthcare providers’ understandings of life-sustaining treatment: are perceptions of goals shared or divergent?

Soc Sci Med 2006; 62(1): 125–133.

23.

Vandrevala

Hampson

Daly

. Dilemmas in decision-making about resuscitation—a focus group study of older people. Soc Sci Med 2006; 62(7): 1579–1593.

24.

Gready

Ditto

Danks

. Actual and perceived stability of preferences for life-sustaining treatment. J Clin Ethics 2000; 11(4): 334–346.

25.

Carmel

Mutran

EJ.

Stability of elderly persons’ expressed preferences regarding the use of life-sustaining treatments. Soc Sci Med 1999; 49(3): 303–311.

26.

Wittink

Morales

Meoni

. Stability of preferences for end-of-life treatment after 3 years of follow-up: the Johns Hopkins Precursors Study. Arch Intern Med 2008; 168(19): 2125–2130.

27.

Schwarz

Hippler

H-J.

Subsequent questions may influence answers to preceding questions in mail surveys. Public Opin Q 1995; 59(1): 93–97.

28.

Podsakoff

MacKenzie

Lee

J-Y

. Common method biases in behavioral research: a critical review of the literature and recommended remedies. J Appl Psychol 2003; 88(5): 879–903.

29.

Murray

Kendall

Boyd

. Illness trajectories and palliative care. BMJ 2005; 330(7498): 1007–1011.

30.

Murray

Boyd

Kendall

. Dying of lung cancer or cardiac failure: prospective qualitative interview study of patients and their carers in the community. BMJ 2002; 325(7370): 929.

31.

Verlinde

De Laender

De Maesschalck

. The social gradient in doctor-patient communication. Int J Equity Health 2012; 11: 12.

32.

Greene

Adelman

RD.

Physician-older patient communication about cancer. Patient Educ Couns 2003; 50(1): 55–60.

33.

Tulsky

JA.

Beyond advance directives. JAMA 2005; 294(3): 359–365.

34.

Koninklijke Nederlandsche Maatschappij tot bevordering der Geneeskunst (KNMG). Tijdig spreken over het levenseinde [Timely discussion of the end-of-life], http://www2.catharinaziekenhuis.nl/site/loader/loader.aspx?DOCUMENTID=cde6e435-9d9e-49dd-9e12-5a80a8bba15f (2012, accessed 16 May 2014).

35.

Koninklijke Nederlandsche Maatschappij tot bevordering der Geneeskunst (KNMG). Spreek op tijd over uw levenseinde [Speak on time about your life’s end], http://www.anbo.nl/sites/default/files/uploads/knmg-publieksbrochure-spreek-op-tijd-over-uw-levenseinde-12-06-2012-21.pdf (2012, accessed 16 May 2014).