Social, health and lifestyle-related determinants of older adults’ preferences for place of death in South Tyrol, Italy – a cross-sectional survey study

Abstract

Background

As the global aging population expands, understanding older adults’ preferences for place of death becomes pivotal in ensuring person-centered end-of-life care.

Objective

This study aimed to investigate the influence of sociodemographic, health, and lifestyle-related factors on end-of-life care preferences of older adults in South Tyrol, Italy.

Methods

Employing a cross-sectional design, a population-based survey was conducted with a stratified probabilistic sample of adults aged ≥ 75 years in South Tyrol (Autonomous Province of Bolzano/Bozen, Italy). From a randomly selected sample of 3,600 older adults, participants were invited to respond to a questionnaire that included items on older adults’ preferences for place of death and socio-demographic and health- and lifestyle-related factors, including frailty (e.g., PRISMA-7). Descriptive and multinomial logistic regression analyses were performed.

Results

The majority (55.3%) of the 1,695 older adults (participation rate: 47%) expressed a preference for dying at home and 12.7% indicated a desire for specialized end-of-life care in a healthcare facility. However, 27.9% refrained from disclosing their end-of-life care preferences. The factors influencing these preferences concerning the place of death included age, native language, educational level, living situation, and community. Compared to the preference of dying at own home or home of family or friends, older adults aged ≥ 85 years (OR = 0.57, P = 0.002) and living in an urban area (OR = 0.40, P < 0.001) were less likely to prefer dying at a hospital, palliative care unit, or hospice. Older adults living alone (OR = 1.90, P < 0.001), Italian-speaking (OR = 1.46, P = 0.03), and those with an educational level above high school (OR = 1.69, P = 0.002) were more likely to prefer dying at a hospital, palliative care unit, or hospice.

Conclusions

End-of-life care preferences among older adults in South Tyrol were associated with socio-demographic, yet not health- and lifestyle-related factors. Recognizing and integrating these preferences is essential for developing, implementing, and evaluating interventions to promote advance care planning and provide effective, patient-centered end-of-life care.

Background

According to the World Health Organization (WHO), the global population of individuals aged ≥ 60 years is projected to double between 2020 and 2050, reaching 2.1 billion, whereas the number of people aged ≥ 80 years is expected to triple, reaching 426 million [1]. Older individuals require a comprehensive array of services aimed at preventing, slowing, or reversing declines in their physical and mental capacity. These services must be tailored to meet the specific needs of each person, adopt a person-centered approach, coordinate effectively among various health and social care providers, and avoid financial hardship for individuals receiving care. A critical component of this care is ensuring that end-of-life services align with the preferences and choices of older adults regarding the place of care and death [2]. While some older adults may prefer to pass away at home, others may opt for institutional settings such as hospitals or hospice facilities. Understanding these preferences is vital for providing person-centered end-of-life care and shaping healthcare policies and services that cater to the needs of an aging population.

The preferences of older adults concerning their place of death have received significant attention. Surveys and prioritization research indicate that the majority of individuals with advanced illnesses desire to pass away at home, although a notable minority may have other preferences [3,4,5]. However, evidence suggests that home deaths are not universally preferred among older adults [6], and the likelihood of home deaths decreases with age [7]. Future projections suggest that, rather than increasing, the number of home deaths among older adults may actually decline [5]. While research in various countries suggests that preferences for the place of death remain relatively constant in 70–90% of cases [6, 8, 9], the preferences of older adults vary and can change over time [10].

Older adults’ preferences regarding their place of death may be influenced by a complex interplay of individual, familial, cultural, and health care system factors. These factors include health status, symptom burden, socioeconomic status, cultural and religious beliefs, and family support. Health-related factors, such as the presence of terminal illnesses and the need for specific medical interventions, can significantly affect older adults’ decisions regarding end-of-life care at home, in a hospital, or at a hospice facility. Furthermore, cultural and religious beliefs play a vital role, with variations in preferences stemming from diverse perspectives on death and the afterlife. The healthcare system and policy environment in a given region can also affect the preferences of older adults regarding place of death. Factors such as availability and accessibility of palliative care services, insurance coverage, and reimbursement policies can influence the feasibility of delivering care at home or in an institution. Consequently, these determinants have a substantial influence on decisions made by older adults and their families regarding end-of-life care settings and place of death.

While a substantial body of research has focused on identifying factors associated with the actual place of death [9], relatively little is known about the factors associated with preferences for the place of death. Understanding these factors is essential for tailoring healthcare services that respect individual choices and designing policies that promote high-quality, person-centered end-of-life care while accommodating the diverse preferences of older adults. South Tyrol, a region in northern Italy, provides a unique context for studying older adults’ preferences for place of death due to its distinct cultural (e.g., German- and Italian-speaking) and healthcare landscape. The Italian National Health Service (SSN) is a tax-funded National Health Service and highly decentralised, as each region is in charge of organizing and delivering health services to the population. Older adults and informal caregivers in South Tyrol have access to well-developed health and social care, including community care services, residential care, palliative care within hospitals and at home providing long-term and end-of-life care. Individuals who require assistance with daily living activities due to age-related frailty or chronic health conditions receive a care allowance as financial support (ranging from 571 to 1800 Euro monthly) to help maintain their independence and quality of life by enabling them to afford necessary care services. This study aimed to describe older adults’ preferences for place of death in the multilingual and multicultural regions of South Tyrol, Northern Italy, and to explore the sociodemographic, health and lifestyle-related factors associated with their preferred place of death (e.g., home, palliative care unit, or hospice).

Methods

Study design

This cross-sectional, population-based survey study was conducted jointly by the Provincial Institute of Statistics (ASTAT; Istituto Provinciale di Statistica – Landesinstitut für Statistik) and the Institute of General Medicine and Public Health in the Autonomous Province of Bolzano, South Tyrol, between March 1 and May 30, 2023.

Setting and sample

South Tyrol, the Autonomous Province of Bolzano, is part of the Trentino–Alto Adige region in Italy, next to Austria (total population: 534.912), with approximately 70% German-speaking, 25% Italian-speaking, and 5% other languages. The target population of the survey comprised approximately 51,000 individuals residing in South Tyrol, aged 75 years and older. Stratified probabilistic sampling was used in this study. The Provincial Institute of Statistics (ASTAT) randomly selected 3,600 study participants aged ≥ 75 years, stratified by age, sex, and residence, from the register of the current resident population in the whole province. The sample size calculation was based on the decision to collect data from at least 1000 older adults, assuming a response rate of 30% [11]. As mentioned in the invitation letter, the selected individuals were invited to participate when living at home and were able to complete the questionnaire independently or with the aid of a family member. Excluded from the survey were individuals permanently residing in senior living facilities or if institutionalization occurred during the data collection period and those who, due to health reasonswere unable to complete the questionnaire independently or with the aid of a family member.

Participant survey

The participant survey was designed collaboratively by ASTAT and the Institute of General Medicine and Public Health, building on the INSPIRE Population Survey conducted in Switzerland [11]. The German and Italian language versions, translated from ASTAT, were reviewed for language equity by the research group at the Institute for General Medicine and Public Health. Face validity for the survey items was determined by the research group. The final versions were approved by the local health and social authorities and pilot tested for clarity and readability.

Older adults’ preferences for place of death were assessed using a single item [2, 12]. Participants were asked the following question: ‘If you were in a situation of serious illness (e.g., cancer) with limited time to live, where do you think would be your preferred place to die?’ The response options were (1) in your own home, (2) in the home of a relative or friend, (3) in a palliative care unit or hospice, (4) in a hospital (but not a palliative care unit), and (5) I prefer not to say.

Participants’ sociodemographic characteristics included age (birth year), sex (male/female), native tongue (German/Italian/Ladin/Others), citizenship (Italy/other country), educational level (below school/high school or higher), community and region of origin (rural/urban), living situation (alone/with spouse or family member), children (yes/no), financial resources (excellent,/good/adequate/insufficient, t or low), and overall optimismyes/nono).

Health and lifestyle-related factors included self-reported health status (poor or moderate/good or very good), frailty (no = 0–3 points/yes = 4–7 points on PRISMA-7 [13]), , physical activity (2 h or more a week/less than 2 h a week/never),, use of home care assistance (e.g., from family, nursing team, and/or private family assistant) (yes/no), and living will (yes/no). Cronbach’s alpha as measure of internal consistency for the PRISMA-7 was 0.65 for both language versions.

Data collection

Letters were mailed to the randomly sampled participants to inform them about the study and to invite them to voluntarily participate by completing the survey alone or with the aid of a family member between March 1 and March 8, 2023. One month after the first letter (between April 15 and April 24, 2023), a second letter was sent to inform the participants about the study and invite them again to participate. The survey was completed through one of the following: (1) online self-completion, (2) paper-based self-completion, or (3) telephonic interviews with collaborators from ASTAT. The online survey was created using the LimeSurvey [14].

Statistical analysis

Descriptive statistics (e.g., frequency) were calculated to describe the measured variables. We applied multinomial logistic regression analysis to explore the association between three older adults’ preferences: (1) dying at home or home of relatives or friends; (2) dying within hospital, at a palliative care unit, or inpatient hospice; and (3) no response (that is, “I prefer not to say”) as dependent variables, and socio-demographics, and health- and lifestyle-related factors as dependent variable. A multicollinearity test was performed using the variance inflation factor (VIF), and the results showed no evidence of collinearity among the independent variables (Mean VIF = 1.4, maximum VIF = 2.44, minimum = 1.17). All analyses were performed using R Statistical Software (v4.3.1; [15]) in the RStudio environment (2023.6.2.561; [16]) with the packages tidyr [17] and nnet [18]. Statistical significance was set at P < 0.05.

Results

Sample characteristics

In total, 1,695 older adults completed the survey, reflecting a response rate of 47%. As described in Table 1, the majority of participants were female (51.9%), between 75 and 84 years of age (59.3%), living with partners/family (69.6%), had at least one child (79.3), and had the highest educational level below high school (77.8%). More than half of the participants reported their health status as poor or moderate (59.7%), being not frail (58.9%), and receiving home care assistance from family, nursing team, and/or private family assistant (66.1%). Only 10.1% reported having a living will; that is, a legal document specifying the preferences about treatments at the end of life if unable to make their own decisions.

Table 1 Characteristics of study participants (N = 1,695)

Preference for place of death

The most preferred place of death among the surveyed older adults was their own home, with 55.3% of respondents indicating a preference for this setting. Only a small percentage (0.5%) of respondents expressed a preference for passing away from the homes of a relatives or friends. A total of 12.7% of the respondents preferred a palliative care unit or inpatient hospice as their place of death, indicating a desire for specialized end-of-life care in a healthcare facility. A modest 3.5% of older adults favored the hospital as their place of death. A significant proportion (27.9%) chose not to disclose their preferences, possibly indicating a level of uncertainty or variability in their preferences (Table 2).

Table 2 Older adults’ preferences for place of death (N = 1,695)

Determinants for preferred place of death

Table 3 presents the results of the multinomial logistic regression analysis. Compared to older adults’ preference for dying at their home or the home of a relative or friend (reference category), several socio-demographic factors were associated with older adults’ preferences for dying at hospital, palliative care unit or inpatient hospice and preference not to say.

Table 3 Multinomial logistic regression analyses on preferences for dying at own home or the home of a relative or friend (reference group), hospital, palliative care unit or inpatient hospice and preference not to say (N = 1,695)

Individuals aged ≥ 85 years were less likely to prefer dying in a hospital, palliative care unit, or inpatient hospice than to those aged 75–84 years (OR = 0.57, P = 0.002). Italian-speaking older adults were more likely to prefer dying in a hospital, palliative care unit, or inpatient hospice compared to German-speaking counterparts (OR = 1.46, P = 0.03). Conversely, participants with Ladin or other native languages were significantly less likely to express this preference (OR = 0.25, P = 0.022). Living in a rural community was associated with lower odds of preferring to die in a hospital, palliative care unit, or inpatient hospice compared to living in an urban area (OR = 0.40, P < 0.001). Participants living alone were significantly more likely to prefer dying in a hospital, palliative care unit, or inpatient hospice compared to those living with family (OR = 1.90, P < 0.001). Those with an educational level of high school or above were more likely to prefer dying in a hospital, palliative care unit, or inpatient hospice than those with less than high school education (OR = 1.69, P = 0.002).

Several socio-demographic factors were also associated with a preference not to state where one would prefer to die. Individuals aged ≥ 85 years were less likely to express a preference not to say than those aged 75–84 years (OR = 0.71, P = 0.015). Italian-speaking older adults were more likely to prefer not to state their place of death than those who spoke German (OR = 2.03, P < 0.001). Living in a rural community was associated with a lower likelihood of preferring not to state a place of death than living in an urban area (OR = 0.67, 95% CI: 0.50–0.88, p = 0.005). Participants living alone were also more likely to prefer not to state a preference compared to those living with their family (OR = 1.50, 95% CI: 1.15–1.96, p = 0.002). None of the health- or lifestyle-related factors, including health status, frailty, or physical activity, were significantly associated with preferences for place of death in this analysis.

In our cross-sectional survey study investigating older adults’ preferences for the place of death in South Tyrol, Italy, we uncovered several notable findings. Most respondents (55.3%) expressed a preference for spending their final moments at their own home and 12.7% indicated a desire for specialized end-of-life care in a healthcare facility. However, it is crucial to recognize that a significant proportion (27.9%) refrained from sharing their preferences for this deeply personal subject. Multinomial logistic analyses revealed that socio-demographic factors such as age, native language, educational level, living situation, and community were associated with expressed preferences concerning place of death.

The preference of respondents in South Tyrol desiring to die at home mirrors international literature. The percentage of patients preferring to die at home varies between countries, ranging from 35.3% in Belgium to 50.6% in the Netherlands in one study [19]. In another study in Belgium, the Netherlands, Italy, and Spain, home death preferences ranged from 65.5–90.9% [20], whereby a high preference for home deaths was strongly associated with palliative care provision by general practitioners [21]. In Canada, 70.8% of the general public preferred to spend their last day at home [22]. In Japan, approximately 50% of the general population preferred home as their location of care during pain scenarios, whereas 40% had the same preference in dependent-without-pain scenarios. Cancer patients in the same context indicated around 40% preference for home care in both scenarios [23]. Among terminally ill cancer patients less preference for home death was reported from patients on palliative care units (58%) than among outpatient palliative care (72%) home as preferred place of death [24].

While many individuals express a preference for dying at home, valuing the comfort and familiarity of their environment, little is known about the extent to which patient preferences align with actual end-of-life care provisions. The actual percentage of those dying at home can vary significantly and is often influenced by factors such as palliative care provisions, regional variations, and disease type. Despite a prevalent preference for dying at home in the context of Ireland, the majority of deaths occur in hospitals [25]. Among Italian cancer patients, 57.9% died at home, with a striking regional variation in home deaths. As 93.5% of these patients expressed a preference for home death, there was a significant gap between preference and reality [26]. In Brussels, only 12% of terminally ill patients die at home. However, when a preference to die at home was known, 66% did so, and the involvement of palliative care specialists significantly improved end-of-life circumstances [27]. A shift away from home deaths, as observed by Hurley et al., has implications for the healthcare system, persons, and their families [28]. Divergence from home deaths also raises concerns about the potential increased healthcare costs associated with institutional deaths and the possible overburdening of healthcare facilities [29]. A trend away from home deaths may underscore the need for improved communication between healthcare providers, older adults, and their relatives [19,20,21]. However, a decrease in home deaths may not necessarily indicate that patients’ preferences have not been met. An increase in the number of patients dying in inpatient care could also be attributed to changing patient preferences [10], with more patients expressing a preference for inpatient settings over the years [28]. The availability of specialized care such as hospice or palliative care facilities might attract people and their families because of the comprehensive support they provide, given the escalating complexity of care during the end-of-life phase, which may be difficult to deliver at home. Thus, addressing older adults’ preferences and establishing early conversations about advance care planning, end-of-life care, and preferences regarding place of death are essential to aligning care delivery with individual wishes [30]. Healthcare systems must be adaptable and prepared to address both the logistical and emotional aspects of end-of-life care, ensuring that patients’ preferences and best interests remain at the forefront of care decisions [2].

Our study revealed several sociodemographic factors, notably age, native language, living situation and community, associated with older adults’ preferred place of death. Older adults’ preferences regarding end-of-life care and place of death are not merely clinical considerations; they are also profound reflections of their personal values, cultural and religious influences, and lived experiences. In our study, adults aged 85 years or older were less likely to prefer dying at hospital, a palliative care unit or inpatient hospice and also less likely to not to state their preference than those aged 75–84 years. Thus, age plays a pivotal role in shaping end-of-life care preference. As observed by Fegg et al., older individuals (≥ 65 years) exhibit a greater inclination towards advance care planning [31]. This mirrors the results of Penders et al., where individuals aged 75 or older with cancer had their end-of-life care preferences more frequently recognized and communicated by their general practitioners than those without cancer [32]. Our findings indicate that Italian-speaking older adults were more likely to prefer dying at hospital, a palliative care unit or inpatient hospice and also more likely to not to state their preference compared with those who primarily speak German. The aspect of native language, though not highlighted in the scientific literature, can implicitly play a role in these decisions. Language is deeply entangled with cultural norms and beliefs. Thus, native language might act as a proxy for cultural influences on end-of-life decisions. As Italian speakers may have different cultural norms and attitudes towards end-of-life care and place of death further qualitative investigation is required. The finding that older adults living in rural communities were less likely to prefer dying at hospital, palliative care unit or inpatient hospice might rather related to differences in the living arrangement and family structure between older adults living in rural and urban communities than differences between available health and social care services. From the literature it is known, that rural informal caregivers offer higher levels of care than urban counterparts [33]. However, rural areas might have varying degrees of healthcare infrastructure and accessibility compared to urban settings and could influence individuals’ preferences related to end-of-life care. Because South Tyrol grapples to ensure equitable healthcare across regions, understanding these geographical nuances is paramount. Thus, the divergence in preferences between individuals residing in rural and urban communities warrants further investigation.

This study revealed that individuals with an education level above high school have distinct end-of-life care preferences, i.e., more likely to prefer dying at hospital, palliative care unit or inpatient hospice. This could be indicative of a broader dialogue about health literacy and the potential influence of education on health-related decision making. Unfortunately, we did not assess health literacy in this study. In the constantly evolving healthcare landscape, the role of education in shaping an individual’s understanding and preferences for end-of-life care and place of death is not yet well understood. The difference in preferences between older adults living alone and those cohabiting with their family offers a glimpse into the emotional and social considerations. Although the majority of older adults participating in the study lived with their partner or family members, 30% reported to live alone in their apartment or house. Our study revealed that those living alone were less likely to prefer dying at home than those living with their families. This finding might indicate that those living alone are more open to die in an institutional environment, e.g., as they do not want becoming a burden to relatives and family members they are not already living together or due to difficult family relationships. As living alone may present unique challenges, both logistically and emotionally, that influence end-of-life care desires, it is crucial that the healthcare system remains sensitive and adaptable to various living arrangements when end-of-live care and preferences for death.

A significant proportion of respondents in our study did not report their end-of-life care preferences. A study in the Netherlands revealed that almost half of hospitalized palliative patients had no preferred place of death [30], whereby previous hospital admission, having more chronic diseases and living alone explained the preferred place of death and in 40% was their home. While patients are willing to consider their preferences for place of death, someone has to ask them. Indeed, we found that only 10.1% reporting to have a last will, which is a poignant concern and highlights a critical gap in elderly care in Italy, despite recent legislative advancements. The 2017 legislation on advance directives, affirming rights to refuse treatment and designate medical proxies, marks progress yet doesn’t directly address the specific needs about advance care planning for the elderly. Health policies should therefore emphasize creating an environment where people and their families feel comfortable and safe to talk about their wishes and preferences regarding end-of-life care and place of death. This may involve facilitating open dialogue, developing culturally sensitive communication tools, or incorporating community leaders to bridge the gaps. Comprehensive training programs are crucial to ensure that healthcare professionals such as general practitioners and community nurses are equipped to navigate the nuances of end-of-life care in this setting. These should focus on (i) helping healthcare professionals understand and respect the varied cultural norms and values associated with end-of-life decisions; (ii) enhancing professionals’ ability to communicate effectively across the three primary languages of South Tyrol; (iii) encouraging dialogue between healthcare providers, social workers, and spiritual leaders to provide holistic care; and (iv) building awareness of the importance of initiating early conversations about end-of-life preferences and the legal frameworks supporting them, potentially through incentives for healthcare providers. South Tyrol’s characteristics necessitate a tailored approach to healthcare policies and practices to promote advance care planning and provide effective, patient-centered end-of-life care when combined with the region’s linguistic and cultural nuances.

Implications for practice and further research

The evident preference for home as the desired place of death underscores the necessity of bolstering home palliative care services in South Tyrol. Enhanced home palliative care not only aligns with older adults’ preferences to die at home but may also contribute to better utilization of healthcare resources by potentially reducing the strain on institutional facilities. Linguistic variance may indicate deeper cultural nuances and beliefs. This indicates the importance of culturally sensitive palliative care and the need to comprehend the underlying values and beliefs that patients from various linguistic backgrounds bring to the healthcare setting. South Tyrol should consider these disparities, prompting further investigation into how the cultural norms associated with each language influence preferences and decisions regarding end-of-life care. Finally, the findings of this study can form the basis for developing, implementing, and evaluating targeted interventions to promote advance care planning and provide effective, patient-centered end-of-life care that allows older adults to die in their preferred place.

While this study has revealed an initial understanding of end-of-life care preferences in South Tyrol, there are a vast expanse of research opportunities that can deepen this understanding and translate it into meaningful patient-centered interventions. First, there is a conspicuous need to explore the reasons for the non-disclosure of end-of-life care preferences more deeply. While this study provides preliminary insights, a dedicated qualitative investigation employing methods such as in-depth interviews or focus groups can shed light on the intricate cultural, personal, or systemic factors that deter individuals and families from expressing their preferences. Second, given the cross-sectional nature of our study, conducting longitudinal research in this domain would be beneficial for monitoring older adults’ preferences for place of death over time, as in many cases, end-of-life decision making does not conclude with a stable choice [30].

Strenghts and limitations

One of the salient strengths of this study was its robust design. Stratified sampling ensures comprehensive and systematic representation of various subgroups within a population. By adopting a population-based approach, this study mitigates the risk of selection bias and offers insights that can be generalized to a broader population. Despite a participation rate of 47%, one potential limitation could be the bias introduced by non-participation. Individuals who chose not to participate or opted not to disclose their preferences may inherently differ from those who did, potentially leading to non-response bias. Additionally, reliance on self-reported measures presents a set of challenges. We did not assess older adults’ (health) literacy, which might have influenced the way they expressed their preferences and responded to the questions, thus contributing to response bias. Participants might have either inadvertently or intentionally misreported information due to poor recall or a desire to provide socially acceptable answers, introducing recall or social desirability bias. Although we assessed several health-related factors in this study, we did not explore older adults’ pathologies and comorbidities, which might also influence end-of-life preferences and should be considered in future studies. Finally, the cross-sectional nature of the study provides a singular snapshot in time; however, the design does not allow for observation of changes in end-of-life preferences or other related factors over time. This temporally static view may not capture the nuances and evolution of attitudes and preferences that can occur over extended periods [10].

Conclusions

The present study revealed distinct end-of-life care preferences among older adults in South Tyrol, highlighting the intertwined influence of socio-demographic factors. The majority expressed a preference for spending their final moments at their home, with a significant proportion (27.9%) refraining from sharing their preferences. The findings presented here reiterate the significance of age, language/cultural aspects, educational level, living situation, and community in end-of-life decisions, and more precisely, preferences regarding the place of death. Recognizing, respecting, and accommodating the end-of-life care preferences of older adults is an ethical imperative, and achieving a preference regarding the place of death should be considered the gold standard in palliative care. Greater efforts from healthcare policies and care providers are needed to foster open communication among older adults, their families, and healthcare professionals about advance care planning and end-of-life care, including the place of death. Further qualitative research on the place of death is needed to better understand the cultural, personal, and systemic factors that deter individuals and families from expressing their end-of-life care preferences.