Frequency and determinants of advance directives concerning end-of-life care in The Netherlands

Abstract

In the USA, the use of advance directives (ADs) has been studied extensively, in order to identify opportunities to increase their use. We investigated the prevalence of ADs and the factors associated with formulation of an AD in The Netherlands, using samples of three groups: the general population up to 60 years of age, the general population over 60 years of age, and the relatives of patients who died after euthanasia or assisted suicide. The associated factors were grouped into three components: predisposing factors (e.g. age, gender), enabling factors (e.g. education) and need factors (e.g. health-related factors). We found that living wills had been formulated by 3% of younger people, 10% of older people, and 23% of the relatives of a person who died after euthanasia or assisted suicide. Most living wills concerned a request for euthanasia. In all groups, 26–29% had authorized someone to make decisions if they were no longer able to do so themselves. Talking to a physician about medical end-of-life treatment occurred less frequently, only 2% of the younger people and 7% of the older people had done so. Most people were quite confident that the physician would respect their end-of-life wishes, but older people more so than younger people. In a multivariate analysis, many predisposing factors were associated with the formulation of an AD: women, older people, non-religious people, especially those who lived in an urbanized area, and people with less confidence that the physician would respect their end-of-life wishes were more likely to have formulated an AD. Furthermore, the enabling factor of a higher level of education, the need factor of contact with a medical specialist in the past 6 months, and the death of a marital partner were associated with the formulation of an AD.

Introduction

When a patient becomes incompetent it may be difficult for relatives and physicians to make end-of-life decisions when they are not sure what the patient would have wanted. However, if the patient has an advance directive (AD), or has talked about end-of-life preferences, this may make the decision-making easier and more in accordance with the patient's wishes.

Much attention has been paid to ADs in the USA, and research concerning ADs and how to stimulate patients to formulate them originates almost exclusively from the USA. We will first discuss previous research on ADs in the USA, and then move on to describe the situation in The Netherlands and the aims of this study.

In the USA, the federal Patient Self-Determination Act (PSDA) was implemented in 1991 to stimulate patients to express their wishes with regard to treatment at the end of their life. An important requirement of the PSDA was that health care organizations had to inform patients about their right to formulate ADs concerning health care. Two types of ADs in the USA are “living wills” and “durable power of attorney for health care”. In a living will the patients themselves specify which treatments they wish to receive and which treatments they would want to be foregone, and under which circumstances. In a durable power of attorney, the patients authorize another person to speak or decide on their behalf (Miles, Koepp, & Weber, 1996).

After the enactment of the PSDA the formulation of ADs increased considerably in some populations, and the PSDA led to better documentation of ADs in medical records. Two years before the enactment of the PSDA, 5% of nursing home residents had an AD, but 2 years after the enactment this had increased to 35% (Bradley, Wetle, & Horwitz, 1998). In a cohort of seriously ill hospitalized patients, 21% had an AD. This rate was similar before and after the PSDA, and even after an additional intervention. The documentation of ADs, however, did improve, only 6% of which were mentioned in the medical records before the PSDA, 35% were documented after the PSDA, and 78% after the PSDA with an additional intervention (Teno et al., 1997). It was considered somewhat disappointing in the USA that studies concerning ADs after the enactment of the PSDA showed only a limited increase in the prevalence of ADs in the overall population.

Existing evidence suggests that people are unable or unwilling to formulate ADs until they grow older or become ill. This observation is supported by several prevalence studies. The prevalence of ADs in adults was found to be 18% (DeLuca Havens, 2000), and approximately one third of older adults, seniors in health maintenance organizations and nursing home residents had an AD (Gordon & Shade, 1999; Hopp, 2000; Levin et al., 1999; Terry & Zweig, 1994). Approximately 40% of people with HIV infection or a chronic lung condition had an AD (Heffner, Fahy, Hilling, & Barbieri, 1996; Wenger et al., 2001) and 56% of people in the terminal stages of cancer had an AD (Virmani, Schneiderman, & Kaplan, 1994). Finally, a retrospective study of decedents showed that, at the time of their death, over 50% had an AD; up to 63% among deceased nursing home residents and 75% among hospice residents (Jacobson et al., 1996). Apparently people are more likely to formulate an AD when they think it more likely that they will need one.

Several studies have attempted to identify factors associated with the formulation of ADs, in order to achieve better understanding of the people who have ADs and those who do not, and to find opportunities to increase their use. We will arrange these factors according to the three following components: predisposing, enabling and need factors. These components were originally proposed by Andersen and Newman in 1973 (Andersen & Newman, 1973) as a framework for access to medical care, but have proven useful as a framework for factors associated with the formulation of ADs (Rosnick & Reynolds, 2003). Predisposing factors are demographic factors (e.g. age, gender) and factors concerning beliefs (e.g. religion, attitudes). Enabling factors could facilitate access to health care, or in this case more specifically the formulation of ADs (e.g. education, social support). Need factors are health-related factors, either evaluated or self-perceived, that could motivate someone to access health care or formulate an AD. Many predisposing factors were found to be associated with the formulation of ADs in one or more studies, such as: being older, being female, being white, coming from a rural area, being more religious, and not being married (Bradley et al., 1998; Bravo, Dubois, & Paquet, 2003; Buchanan, Bolin, Wang, Zhu, & Kim, 2004; DeLuca Havens, 2000; Elpern, Yellen, & Burton, 1993; Gordon & Shade, 1999; Hopp, 2000; Jacobson et al., 1996; Phipps et al., 2003; Rosnick & Reynolds, 2003; Terry & Zweig, 1994; Wenger et al., 2001). Enabling factors that were found to be associated with having formulated an AD were: higher education, higher socio-economic status, higher income, and higher social support (Bradley et al., 1998; Hopp, 2000; Ott, 1999; Rosnick & Reynolds, 2003; Wenger et al., 2001). Need factors that were found to be associated with having formulated an AD were: worse (self-perceived) health status, and residing in a nursing home or hospice (Elpern et al., 1993; Jacobson et al., 1996). Furthermore, several negative experiences were associated with having formulated an AD: death or illness of a loved one, and negative life events in general (Bradley et al., 1998; DeLuca Havens, 2000; Inman, 2002; Rosnick & Reynolds, 2003). We will classify these factors as need factors, since they are experiences that could influence the perceived need to formulate an AD.

In comparison with all the attention that ADs have received in the USA, the subject is very much underexposed in Europe. This may be because physicians in the USA are more likely to choose more aggressive treatment options than physicians in various European countries (Alemayehu et al., 1991; Koeck, Hemenway, Donelan, & Lipsitz, 1998; McKenzie, Moss, Feest, Stocking, & Siegler, 1998). In this context, ADs, which usually limit treatment, may be considered more necessary in the USA.

In The Netherlands, we have ADs that are similar to those in the USA, i.e. living wills and people can appoint a representative. People can formulate any AD, regardless of their current health condition. Several Dutch associations have issued standard forms which people can complete by marking situations in which they would want the AD to apply (e.g. in case of being comatose for a certain number of weeks) and/or treatments which they would want to be provided or forgone in some or all situations (e.g. artificial nutrition and hydration). People can also write an AD in their own words, and such a directive can have equal legal value. Considerable attention has recently been paid to the right of patients to refuse treatment. According to the Dutch constitution, this right has always existed, but it has been described in more detail in the Medical Treatment Contract Act, that came into force in 1995.

In The Netherlands we recognize a type of living will that does not exist in most other countries, the “advance euthanasia directive”, in which people can request euthanasia in specific situations of incompetence. Since 2002, such requests can legally be granted if the official requirements of due care for euthanasia or assisted suicide (EAS) are met. However, there is much debate in The Netherlands about whether or not in the case of EAS these requirements can ever be met, because one of the requirements is that the patient's suffering is unbearable, which is usually considered to be impossible if a patient is comatose or demented, or is considered to be incompetent for another reason. All cases of EAS have to be reported to review committees, and there have been no reports of cases in which the physician performed EAS on the basis of an advance euthanasia directive.

The aim of this study is to investigate the prevalence of ADs in The Netherlands and to determine which factors are associated with the formulation of ADs. Almost no data are available on the prevalence of ADs in Europe. Especially in The Netherlands, where people can have ADs concerning euthanasia besides other types of ADs, evidence is needed to see how many people have ADs, how often these concern euthanasia, and whether the same factors influence whether people have an AD, in spite of the differences between ADs in the USA and in The Netherlands. The prevalence will be analyzed in three study groups: the general population up to 60 years of age, the general population over 60 years of age, and the relatives of patients who have died after EAS. These relatives were included because we thought that they were more likely than the general population to have ADs, and to have informed their physician, as a result of their experience with EAS. We hypothesize that older people are more likely to have ADs than younger people, since we expect that people with a greater (perceived) risk of needing an AD will be more likely to have formulated one. We will also investigate the association between confidence that a physician will respect end-of-life preferences, and the formulation of ADs, an association not taken into account previously. This seems a relevant factor, because people who are not confident that physicians will respect their end-of-life preferences may attach more importance to the formulation of ADs. Furthermore, in the sample of the population over 60 years of age, we will evaluate several characteristics that were found to be associated with the formulation of ADs in the studies in the USA. We expect that each of the three types of factors, predisposing, enabling and need, are associated with the formulation of an AD, with a decisive role for need factors. By evaluating the factors associated with the formulation of an AD, it may be possible to find suggestions to stimulate the formulation of ADs.