In-hospital care prior to assisted and unassisted suicide in swiss older people: a state-level retrospective study

Suicide is a major public health issue, affecting all ages and is expected to grow as life expectancy increases [1, 2]. The incidence of suicide is particularly high among older males aged 65 and older [3]. The rate of suicide in the oldest group (ages 85+) is over four times higher than overall rate of suicide in Switzerland or in the US; moreover, this age group shows an increase in suicide rates in almost all countries [3, 4].

Elderly persons have a higher risk of completed suicide compared with younger age groups. Their frailer health status should not be considered as the main contributing factor to this excess of mortality. Elderly use more deadly methods and are less likely to be found and rescued because of the scarcity of warning signs when planning and carrying out the lethal act [5].

Whereas for the majority of the countries in the world, assisted suicide is not an option, in Switzerland, this process is permitted. In this country, assisted suicide counts for about 1% of deaths, with people 65 years and older having the highest rates of assisted suicide [4, 6, 7]. A person can be granted access to assisted suicide when the following two conditions are met: to have decisional capacity and to be able to perform the fatal act oneself. In addition, other persons implicated in this process should have no selfish motive. The Swiss legal framework permits right-to-die societies to assess and decide on the acceptability of a request, without restrictions regarding the source of suffering.

The Swiss Academy of Medical Sciences ethical guideline for end-of-life care issued in 2013 required that the underlying disease implies that the end-of-life is expected to be near. However, since 2014, right-to-die organizations in Switzerland have considered requests for assisted suicide, in the case of unbearable suffering or age-related changes, even in the absence of terminal illness [8, 9].

Since assisted suicide became relatively easier to access in Switzerland, we anticipated that, over the years, an increasing number of people with chronic illnesses would carry out assisted suicide to end their suffering. We additionally hypothesized that there would be a decrease in unassisted suicides with the widening of indications by right-to-die organizations in 2014.

Suicide in late life is known to be complex and multifactorial. Some factors have been found to influence both types of suicides in the elderly, such as constricted social networks (loneliness, social exclusion, bereavement), mood disorders and cognitive impairment [10‐13]. It has been found in recent systematic reviews that physical illness and functional impairment are important suicide risk factors in the elderly. Furthermore personality disorders that lead to an inability to adapt to the changes occurring in late life are more likely to be found in older adults who attempt or complete suicide than in their younger counterparts [14‐18]. These findings suggest that the elderly who contemplate suicide would be particularly sensitive to any worsening of their quality of life. Older age suicide may have rational grounds, following a process of deliberation and holistic self-assessment of irremediable impairment of current or future ones-quality of life [19]. Indeed, elderly persons may contemplate suicide due to being tired of life or considering that their life is complete. Neoliberalism, technology, and changing attitudes related to the legalization of assisted death in some countries are sociological trends thought to influence rational suicidal thoughts [20].

However, elderly prone to unassisted and assisted suicide differ in many ways.

Unassisted suicide in the elderly often follows impulsive decision-making during major depressive or mixed illness episodes [21]. Unassisted suicide is known to be strongly associated with previous suicide attempts, mood disorders (e.g. depression, bipolar disorder), vulnerability in the context of borderline or other personality disorders, presence of impulsive-aggressive traits, hopelessness and substance use [22‐26]. Individuals with mental and behavioral disorders generally report having a poor perceived quality of life, as they are at greater risk of experiencing marginalization, poverty and stigma, and show a higher incidence of suicide than the general population [21, 22, 27, 28].

On the other hand, the typical assisted suicide decedent profile has been described as a single or divorced woman with tertiary education, without religious affiliation, affected by cancer or any other chronic illness [29]. Persons who choose to opt for assisted suicide are supposed to make an informed choice about how and when to die. They are evaluated for their decisional capacity beforehand and are imposed a time for reflection before making their final decision. Thus, the decision to end one’s life via assisted suicide may become an option, especially when facing or fearing chronic illnesses’ consequences. Indeed, chronic illnesses can leave patients in a debilitative state, that no longer enables the necessary self-determination to end their own lives. Fears of indignity, of being a burden, of losing autonomy or losing control as well as tiredness of living have also been reported as reasons for requesting assisted suicide [12, 13].

We thus anticipated that assisted suicide would be associated with more frequent somatic health problems leading to intense in-hospital care, whereas unassisted suicide would more frequently occur in cases with mental disorders. This divergence would explain older suicide victims’ different utilization of health care. Increased health care intensity would increase in assisted suicide but not unassisted suicide, since it adds to individuals’ experienced burden. Moreover, higher utilization of health care facilities could increase the opportunity for suicide prevention by screening and treating more elderly patients with severe mental illness who would carry out impulsive suicidal acts.

Late-life suicide remains understudied although the characterization of individuals would be central to reduce the progression of suicidal thoughts to acts and to improve elderly people’s quality of life. Assisted suicide research has focused on sociodemographic characteristics and general diagnoses, mostly disregarding individuals’ previous experience with healthcare. Examining health care experience seems nevertheless important as fears of medical treatments, hospitalizations and nursing home stays based on prior experience appear to be dramatic triggers of the contemplation phase of assisted suicide [30].

There is strong evidence for an increasingly high and aggressive intensity of care during the last 12 months of life [31‐38]. High intensity of care can be experienced as impacting quality of life negatively, as being futile and inconsistent with the patient’s wishes [39]. Hospital care is also associated with an increased risk of depression in older patients [40‐42]. Medicalization at the end of life may ultimately alter the patient’s desire to live and influence their wish or decision to die [43]. Furthermore, high levels of care may cause older adults to fear their lives would be prolongated against their will [39, 43].

Based on these results, the authors have made the hypothesis that, the level of care may be experienced by older patients as a burden or as repeated negative events altering their quality of life and thus may be an important influencing factor for carrying out assisted suicide.

Prior hospital care intensity may be a positive predictor of assisted suicide relatively to unassisted suicide. If confirmed, this association would further increase with age, somatic diseases and hospitalizations.

We tested the hypotheses in a state level retrospective study conducted in Geneva, Switzerland. The study employed data from individually linked hospital medical and forensic records to assess the strength of the associations between high intensity of care (for mental health or somatic reasons) received by patients aged 65 years or older prior to their assisted suicide or unassisted suicidal act and to determine whether these associations are modified by patients’ characteristics and diseases or by the type of hospital care experienced.

Figure 1 shows the progression of suicide deaths of older adults in Geneva between 2010 and 2016. During the study period, 359 people died in the group aged 65 and more in Geneva; the majority being assisted suicides (n = 269, 75%). Deaths by assisted suicide globally increased by 200%. The proportion of assisted suicide among overall suicides significantly increased in years 2015–2016 compared to 2010–2014 (152 of 228, 66.6%, versus 117 of 131, 89.3%, P < .0001); the highest increase of numbers of assisted suicide deaths being observed in 2015.

As hypothesized, our study shows a substantial progression in assisted suicide deaths between 2010 and 2016, for both men and women 65 years and older, in Geneva, with a peak in 2015. This peak most likely reflects the expansion of eligibility criteria for requesting assisted suicide to non-terminal diseases decided by right-to-die organizations in 2014. The count of assisted suicides in retirement homes reached a peak the same year, thus participating to this phenomenon. The social impact of the increase in assisted suicide on families should not be disregarded as some studies has shown a higher prevalence of post-traumatic stress disorders and depression among family members [46].

Elderly assisted suicide decedents were more likely to receive highly intensive care than peers who died by unassisted suicide, with these positive associations being stronger among the oldest women suffering from somatic illnesses living at home or in a nursing home as previously reported [47]. First, the oldest elderly may be more tired of living than their younger counterparts, and thus more at risk of rational suicide. The oldest old living at home may suffer from isolation and social exclusion, that are well known risk factors for assisted suicide [4]. In the other hand, the fact that we found more assisted suicide than unassisted suicide in nursing homes, could be explained by a narrower contact with caregivers as well as the generally worse level of functioning of residents that prevent carrying out a suicidal act on their own.

Mental and behavioural disorders as well as poor use of hospital care were more likely to be found in the unassisted suicide group compared to the assisted suicide group. This result is consistent with the fact that previous reports showed that patients with mental disorders are at risk to receive poorer access and quality care as a result of mental illness-related stigma [28, 48]. Furthermore, help-seeking behaviour is affected among patients with mental illness [48]. Individuals with mental and behavioral disorders often lack the necessary level of planning and organization to engage in assisted suicide procedures when they consider ending their life, and therefore would engage in more impulsive suicide attempts even when an assisted way of dying is available [21].

As mental health disorders are rarely found in assisted suicide decedents records, screening and treating mental illness (e.g. depressive state) would not be sufficient to reduce assisted suicide death. Further research should therefore examine whether it is of interest to look for the determinants of poor quality of life in this group to enhance their autonomy.

Somatic diseases such as cancer and non-malignant chronic diseases were more frequently associated with assisted suicide than unassisted suicide as previously reported [49]. Diseases of the circulatory system were the most prevalent diagnoses present in both groups whereas cancer was the second main diagnosis found in assisted suicide decedents. Cardiovascular diseases such as coronary heart disease, hypertension and stroke are strongly linked with depression and loneliness, which could be linked with both types of suicide [50].

Hospitalization rates of people aged 70 and more were at 20% in Geneva state according to 2016 statistics, which is comparable to the values found in our study for the unassisted suicide group (18.8%). In our study, this percentage was by comparison about three times lower than the rate observed among people who died by assisted suicide (57.9%) [51].

The association between assisted suicide and hospitalization could have reflected the physical, psychological, existential, social and/or financial burden associated with severe somatic diseases. Patients with advanced cancer as well as those facing non-oncologic diseases are indeed at high risk of receiving intensive levels of care at the end of life [52].

However, the discrepancy in hospital care utilization between people who died by assisted versus unassisted suicide did not appear to be fully accounted for by the difference observed in the diagnoses nor the other tested characteristics. As hospitalization and surgery are independently associated with assisted suicide, our findings augment existing evidence that these markers of high level of care impact on the wish to live negatively.

This phenomenon could be influenced by changes in the quality of life during hospitalization, particularly after surgery or other contacts with health care professionals.

It has been indeed demonstrated that hospitalizations put elderly patients at risk of disability, depression and cognitive loss. Moreover, delirium which is very common especially after surgery, may be persistent after hospitalizations [53‐57].

No evidence was found to support associations between the type of suicide and emergency department unit admissions without hospitalizations or ICU admissions. One explanation might be that emergency department ambulatory care is due to unstable situations where patients find support and receive reassuring punctual care that does not have a lasting negative impact on their quality of life. In addition, ICU admissions, which are proven to impact late-life quality of life negatively, were events that occurred too rarely in our study to confirm their potential negative impact on quality of life and the desire to live meaningfully [42].

Compared to individuals who carried out unassisted suicide, older patients who opted for assisted suicide receive significantly more intense hospital care during their last year of life, independently of having cancer, other somatic diseases and/or mental and behavioural disorders. Suicide research should no further ignore the global consequences of intense care in older people.