Introduction

In the course of irreversible physical decline, people are often confronted with vital decisions [1, 2]. To assist the patients’ needs and decisions in the course of a disease such as amyotrophic lateral sclerosis (ALS) is a challenging issue for physicians, caretakers and families [3].

Amyotrophic lateral sclerosis is one of the most devastating neurological conditions with cardinal symptoms of progressive loss of mobility and verbal communication leading to death within 3–5 years. ALS is regarded as a rare disease; nevertheless, ALS has been prominent in the debate of assisted suicide and euthanasia [23]. In Oregon, about 0.4 % of terminal cancer patients requested lethal prescriptions, whereas 5 % of ALS patients did [24]. In the Netherlands, 20 % of ALS patients die as a result of life-shortening measures like euthanasia [1, 4]. In bordering Flanders, where only one out of two euthanasia cases are officially reported, a majority of those unreported cases requested euthanasia verbally [5]. Numbers of ALS patients’ initial verbal request for physician-assisted suicide shortly after diagnosis may be as high as 56 % [6], but only a minority finally request life-shortening treatments in Oregon [4, 7, 8]. In countries like Germany [9] and Switzerland [10], the numbers for actual request of life-shortening interventions is similarly low. Therapeutic treatments with life-prolonging and quality of life-assuring effects such as non-invasive ventilation may be used more frequently in these countries [9]. Whether good palliative care and provision of therapeutic treatments antagonizes desire for hastening death has been debated [4].

Previous studies reported that patients with ALS focus on the present state in decision making rather than anticipating future progression [11]. They often seem to adopt a wait-and-see strategy regarding treatment options [12] and gradually adjust to disabilities imposed by the condition [13]. An unbiased, patient-centered view and a scientifically based knowledge of the patients’ needs are a prerequisite for good counselling [13]. Instead, doctors, caregivers, and patients alike, lack knowledge about options and patients’ preferences [14]. Furthermore, psychological defence mechanisms to explore the actual needs of patients of health care practitioners prevails [4].

Decision making is influenced by a range of factors like individual preferences and available coping strategies [11, 15]. Quality of life may constitute a determinant of decision making in ALS and a poor quality of life is one of the most important reasons for requesting physician-assisted suicide in Oregon [8]. In the present study, we use the concept of hedonic quality of life, which refers to more transient feelings, such as life satisfaction and happiness opposed to eudaemonic health related QoL, which focuses on functioning, achievements and capabilities. The former can be measured with subjective and global quality of life which is known to be good in ALS [16, 17]. This type of QoL is not associated with loss of physical function [16, 17] and is highly underestimated by healthy persons [18]. Other possible determinants of decision making in ALS such as depression [1, 19], hopelessness [7], social support [7] and feeling of being a burden [20] have been reported by physicians and caregivers in some retrospective studies but not in others [1], whereas the presence of religious beliefs are usually determinants in patients who decide against euthanasia [1, 3]. Furthermore, cognitive impairments have been described in up to 50 % of ALS patients [21] and may interfere with decision making capabilities. The need for prospective studies to describe accurately the relationship of physical and psychological symptoms of ALS patients in the process of vital decision making is essential [4].

We report a prospective study in a large cohort of N = 93 ALS patients in advanced stages of the disease to investigate patients’ decisions and attitudes to treatments with a life-prolonging effect (non-invasive ventilation, invasive ventilation and percutaneous endoscopic gastrostomy) [14, 22] and desire for hastened death in the course of ALS. Furthermore, we aimed at determining factors leading to such decisions. Patients were interviewed three times within 1 year.

Methods

Design

Amyotrophic lateral sclerosis patients were consecutively recruited at the outpatient clinic of the department of neurology at the University of Ulm and at the Charité Berlin. They were assessed three times in the course of 1 year (initial interview, 6 and 12 months later; Fig. 1). All questionnaires were presented in each interview, except the extended cognitive evaluation, which was conducted at the initial interview only. The initial interview lasted about 2–4 h, the following interviews after 6 and 12 months lasted about 1.5–3 h. Interviews were guided by a trained psychologist (S.S. or S.N.) who filled out the questionnaires.

Fig. 1
figure 1

Study design

Full size image
Fig. 2
figure 2

Progression of possible determinants of decisions in the course of ALS. ALS-FRS-r ALS functional rating scale mean score; no depression, incidence of patients without clinically relevant depression; global and individual quality of life mean score; shaded area and dashed lines indicate theoretical progress of factors before our study based on clinical experience of interviewers S.S. and S.N., remaining area depicts actual study data

Full size image

Participants

N = 93 ALS patients participated in this study at first interview (Table 1). Sample size was determined for a power of 80 % and α = 0.05 for medium effects (ε = 0.4) which were expected due to previous findings [15]. For the interview after 6 months (N = 43) and 12 months (N = 31), dropouts were either due to death (N = 15, equivalent to 16 % after 12 months) or physical/mental inability (N = 47, equivalent to 51 % after 12 months) to participate. Patients were diagnosed according to the El Escorial criteria (Table 1). Patients were moderately to severely impaired according to the ALS functional ratings scale (revised form, ALSFRS-R) [25]. All had been informed prior to the study regarding life-sustaining and quality of life-assuring techniques such as non-invasive ventilation (NIV), invasive ventilation via tracheostoma (IV) or percutaneous endoscopic gastrostomy (PEG) by the treating physician.

Table 1 Background data of patients at initial interview
Full size table

None of the participants had any neurological or psychiatric diagnosis (other than ALS) and none used psychologically active medication.

Exclusion criteria were major cognitive impairments assessed by mini mental state examination (MMSE <22) [26]. Patients were interviewed by a certified clinical psychologist. The responsible physician confirmed that the patient had sufficient time to adapt to the diagnosis (at least 6 months). The study was approved by the ethics committees of the University of Ulm and the University of Berlin and has, therefore, been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. All participants gave informed consent prior to their inclusion in the study.

Decision status

A 120-item questionnaire (decision status, treatment status and determinants of decisions) regarding NIV, IV and percutaneous PEG was used (response options for decision status “decided for”, “decided against”, “undecided”, options for present treatments status “with NIV/IV/PEG” or “without NIV/IV/PEG”). Questions included “Did you seek information how to shorten life?” and “Should euthanasia be allowed?" (response options “yes” or “no”). Determinants asked about patients' “feeling of being a burden” (on a Likert scale from 1 = “not at all” to 4 = “yes indeed”).

Furthermore, the desire to hasten death was evaluated with the schedule of attitudes towards hastened death (SAHD [27]; range 0–20; >10 indicates clinically significant desire for hastened death).

Determinants of decision

Quality of life and depression

Global quality of life (QoL) was assessed with the anamnestic comparative self-assessment (ACSA [28]; range −5 for as bad as possible and +5 as good as possible), subjective QoL was assessed with the schedule for the evaluation of subjective quality of life (SEIQoL [29]; range 0 for as bad as possible and 100 as good as possible). Depression was measured with the Allgemeine Depressionsskala (ADSK [30]; German version of the Center for Epidemiologic Studies Depression Scale, CES-D, range 0–60; threshold >16).

Social support

Perceived social support was evaluated with the emotional scale of the social support scale (Soziale Unterstützungsskala SOZU K 22 [31]; 5-point Likert scale range from 1 = 'not at all' to 5 = 'yes indeed').

Cognition

Cognitive performance was assessed with explorative interview and MMSE [26] (range 0–30, threshold <26 for mild dementia).

Additionally, at first interview in N = 58 ALS patients verbal fluency was assessed with the Regensburg fluency test (Regensburger Wortflüssigkeitstest) [32], attention and frontal inhibition was assessed with a version free of motor requirements of the D2-test [33] (in-house computer adaptation) [34].

Statistics

Repeated measures ANOVA with time as factor was performed separately for independent variables, “desire for decisions for life-prolonging treatments (PEG/NIV/IV)” and “desire for hastened death (SAHD)”; and dependent variables, physical function (ALS-FRS-r), quality of life (SEIQoL, ACSA), depression (ADSK), social support (SOZU); and the factor “feeling of being a burden”.

Kruskal–Wallis ANOVA was performed for decisions for life-prolonging treatments (PEG/NIV/IV) and dependent variable physical disability (ALS-FRS-r), age, quality of life (SEIQoL, ACSA), depression (ADSK), social support (SOZU) and the factor “feeling of being a burden”.

Using multivariate logistic regression, we examined which variables predicted desire for hastened death. Variables were psychological determinants such as quality of life (ACSA, SEIQoL), depression (ADSK) and feeling to be a burden as well as external determinants such as physical function (ALS-FRS-r), age and social support (SOZU). Only participants with complete data sets were included. As data were frequently not normally distributed, non-parametric statistics were used.

The statistical analysis was performed with Statistical Package for the Social Sciences 16.0. The significance level was adjusted two-tailed at p = 0.05.

Results

Vital decisions

Initially, up to half of the patients decided on life-sustaining treatment. About half were undecided (Table 2).

Table 2 Decisions and determinants of decision
Full size table

Of those with undecided or negative attitudes, there was a significant increase in decisions for IV and PEG towards acceptance in the course of 1 year. For NIV the increase was not significant.

The desire for hastened death was low (Fig. 3) and decreased during the first 6 months (F = 6.72, p = 0.01*). It remained low throughout the study despite the fact that physical function declined (Table 2; Fig. 2). None of the other possible determinants of decision (QoL, depression, social support, feeling of being a burden) changed significantly in the course of a year (Fig. 2). Indicators of psychosocial adaptation such as high subjective quality of life (SEIQoL score >70 on a scale of 100 indicates a good QoL), and neutral to positive global quality of life (ACSA) and low depression (ADSK) were stable throughout the study (Table 2; Fig. 2).

Fig. 3
figure 3

Decline in mean desire for hastened death of patients within 12 months SAHD schedule for the attitudes towards hastened death. Mean score of SAHD >10 indicates clinically relevant desire for hastened death (solid line); maximum value = 20

Full size image

About one-third of the patients (N = 35) responded “yes” for the item “Did you seek information how to shorten life?” and more than two-thirds (N = 63) responded “yes” for the item “Should euthanasia be allowed?"

Determinants of decisions

Kruskal–Wallis ANOVA was performed for decisions for life-prolonging treatments and actual treatment choices (PEG/NIV/IV) and determinants (psychological, external and desire for hastened death).

Psychological determinants

Subjective and global quality of life and depression were not significantly different between patients who decided for life-prolonging decisions (NIV/IV/PEG) and those who decided against. The variable “feeling of being a burden” was significantly different between patients who decided for NIV, PEG and IV (trend) compared to those who decided against those treatments (Table 3).

Table 3 Determinants of vital decisions
Full size table

QoL and depression were no significant predictors for desire for hastened death.

The variable “feeling of being a burden” was a significant predictor for desire for hastened death (F = 16.22, p < 0.001).

External determinants: physical disability, age, social support

Physical disabilities were significantly predominant in patients who decided for NIV and for PEG but not for IV compared to those patients who decided against (Table 3).

Physical disabilities were not significant predictors for desire for hastened death.

Neither age nor perceived social support were predictors for decisions towards life-sustaining treatments nor for the desire for hastened death.

Determinants and actual treatment choices

Physical disabilities were significantly different between people with ventilation and PEG, and those without.

Patients with ventilation and PEG had a significantly higher “feeling of being a burden” compared to those without. The other determinants were not different between groups (Table 3).

Desire for hastened death and actual treatment choices

Desire for hastened death and decisions for invasive life-prolonging treatments (IV/PEG) were negatively associated: Those who decided for life-prolonging treatments (IV/PEG) had a significantly lower desire for hastened death than those who decided against. Desire for hastened death and decisions for NIV were not associated (Table 3).

Discussion

Decisions

The present study is the first prospective study on decisions to prolong or shorten life in a large population of ALS patients. ALS patients showed no end-of-life-oriented despair but decided for life-sustaining treatments indicating positive attitudes towards life. Half of the patients expressed the explicit desire for non-invasive life-prolonging treatments even at the initial interview. In the course of 1 year, there was a significantly increasing number of patients deciding for invasive life-prolonging treatments (IV/PEG), particularly in those patients who had previously either negative or undecided attitudes towards life-prolonging treatments. This illustrates the “wait-and-see” strategy [11, 12] of ALS patients as they adapt and adjust their preferences concerning requests of invasive treatments. Those who decided for invasive life-prolonging treatments had a significantly lower desire for hastened death than those who decided against. Therefore, decisions for invasive life-sustaining treatments antagonize desire to shorten life or vice versa [4].

Patients had no desire for hastened death and it declined even more in the course of ALS progression despite physical function decline. It was far below the threshold for clinically relevant desire for hastened death [27]. About two-thirds of the patients thought that euthanasia should be allowed. One-third of the patients actively searched for information on how to shorten life. But when physical disabilities progress, patients still value life, and only a small percentage of patients consider life-shortening treatments [4, 8–10, 19]. In other regions, e.g. in the Netherlands, up to 20 % of ALS patients request life-shortening treatments [3]. These numbers are incompatible with patients’ attitudes and decisions to end life in the present study and other studies in different regions [4, 10].

Determinants of decision

Quality of life and depression were not determinants for decisions for life-prolonging treatments nor for the desire to shorten life. The lack of an association between depression and requests for life-shortening treatments is in line with previous studies [1, 7, 8]. For QoL, this lack of association has not been shown before. An anticipated poor quality of life was one of the most important reasons for requesting physician-assisted suicide in Oregon [8] and probably in other locations as well. Yet, our prospective data do not support that low QoL is prevalent or predictive. Hedonic QoL was good and depression was low and stable throughout the study despite significant decrease in physical function [16, 17, 35]. A drop in quality of life and increase in depression may follow the diagnosis but after a few months a majority of patients achieve equilibrium, as our data suggest. Differences in QoL between patients are likely due to pre-existing individual differences [36].

We found the “feeling of being a burden” to be a significant determinant of decision against life-prolonging treatment and for the desire to hasten death. Furthermore, those patients with ventilation and/or PEG had a significantly stronger “feeling of being a burden” than those without. The issue of “feeling to be burden” seems to be predominant in patients with terminal illness [37] and this is the first prospective data providing evidence for the association of vital decision making and “feeling of being a burden” in a large cohort of ALS patients. In Maessen’s [1] retrospective study, caregivers did not report this item to be important in the case of deceased ALS patients but in Oregon, terminally ill patients (including six ALS patients) in favour of assisted suicide reported great distress due to the feeling of being a burden [7, 8]. In an environment in which proper physical (and cognitive) functioning is requested for social inclusion, people with disability and terminal illness are prone to experience themselves as being a burden because they do not fulfill these requirements.

Physical disabilities determined the desire for therapeutic treatments, meaning that the more patients were advanced in their disease, the higher the probability to ask for NIV and PEG. NIV is an established procedure that is regularly used in the early course of the disease [22]. PEG seems to have similar therapeutic implications [38], whereas for IV the determinants of decisions are more heterogeneous. Ventilated patients and/or with PEG had the same QoL and depression rate as non-ventilated patients and/or those without PEG. Thus, in line with previous results [17], ventilated patients experience a satisfying quality of life and are not depressed [39]. This is also true for patients with PEG. Cognitive impairments were excluded and did not account for any differences or lack of differences between groups. Therefore, the satisfactory QoL in patients with NIV/IV and/or PEG compared to those without is not a function of incapability to comprehend the impact of the diagnosis but rather a result of the adaptation processes.

Concerning the desire of hastened death, external factors like loss of physical functioning [40], age [1] and perceived social support were not predictors. This supports the notion that lack of social support is not necessarily relevant for the desire to shorten life [8] and the desire to shorten life does not increase while physical function declines or age increases [1]. As opposed to positive attitudes for life-prolonging treatments, which increased in the course of physical function decline, desire to hasten death and physical disabilities both declined in the course of ALS.

Conclusion

These data suggest that patients change their attitudes positively towards life and learn to cope with the disease [15]. Decisions are adapted accordingly. Possible determinants of decisions like QoL, depression and perceived social support are stable. Quality of life is not a determinant of negative or positive decisions and it is high in patients with invasive life-prolonging treatments. The “feeling of being a burden” is a predominant issue in ALS patients in decisions against life and needs to be specifically addressed in clinical counselling.

The data may enable clinicians and caretakers to best support patients in their decision making process. Hastened death decisions [8, 10] have to be circumvented by considering patient’s possible adaptation processes in the course of a fatal disease.