Network Types for End-Of-Life Care and Home Deaths Among Older Adults in Europe During COVID-19 Pandemic

Abstract

The study identified care network types comprising informal and formal care providers during the end-of-life period, and examined their relationship with home deaths. End-of-life interviews were conducted with proxies during the two waves of the COVID-19 pandemic in the Survey of Health, Ageing and Retirement in Europe (SHARE). The sample included 486 participants who passed away during the pandemic and received care during their final year. Latent Class Analysis identified three care network types: “mixed-care”, mostly consisting of adult children in an informal network; “spouse”, dominated by spouses with adult children and formal caregivers; and “professional”, mainly with formal caregivers and some informal support. Patients with mixed-care were more likely to die at home compared to those with professional networks. Informal care, particularly from adult children, played a crucial role in end-of-life care. Policymakers should support adult children in end-of-life care and promote effective models that integrate formal care services.

What this paper adds

• • Identifies different types of care networks that include both formal and informal care for end-of-life patients during a global health crisis.

• • Emphasizes the crucial role of informal care, especially from adult children, in enabling terminally ill individuals to pass away at home.

Application of study findings

• • Inform the development of policies aimed at offering better support for adult children in their caregiving responsibilities.

• • Advocate for the integration of formal care services with informal networks to improve the quality of end-of-life care.

Introduction

The rapid growth of the aging population has generated heightened attention to the importance of implementing appropriate end-of-life care (Nolasco et al., 2020). As policies shift towards the deinstitutionalization of care and the establishment of home-based palliative care practices, there is a growing emphasis on home deaths in both research and clinical fields (Horsfall et al., 2017). Home deaths are considered a quality indicator of end-of-life experience (Motamedi et al., 2021). Home has been reported as the preferred place of death in a population-based study (Wilson et al., 2013), as well as among care-receivers in their end-of-life period, typically defined as up to one year (Bijnsdorp et al., 2019; Nysaeter et al., 2022).

Home deaths have also been highlighted as enabling the preservation of comfort, autonomy, dignity, and well-being of care-receivers. This stands in contrast to medical settings where older adults are often exposed to invasive procedures to prolong life but can adversely affect their quality of life (Quinn et al., 2022; Motemedi et al., 2021). It’s essential to understand that home may not always be the best place to die. This can depend on several factors, including the progression and nature of the illness, the availability of resources for managing pain and other distressing symptoms, and the level of professional and social support accessible (Kenny et al., 2021; Pollock, 2015). During the pandemic, hospital deaths, particularly those resulting from COVID-19, have been identified as indicators of poor end-of-life care. This is primarily due to the restricted access to family and relatives at the bedside during the final moments before death (Feder et al., 2021; Selman et al., 2022).

Yet, home deaths in Europe account for about a third of all deaths (Nolasco et al., 2020). The prevalence varies substantially across countries, with higher rates in Eastern European countries and lower rates in Western and Northern European countries (Srdelić & Smolić, 2022). During the pandemic, some studies reported on a rise in non-COVID-19 home deaths; however, it remains uncertain whether this trend will persist in the post-pandemic era (Aldea-Eamos et al., 2023; Odonnell et al., 2021).

In order to gain a comprehensive understanding of the factors influencing the likelihood of home deaths, various studies have relied on Gomes and Higginson (2006) conceptual model, which encompasses individual characteristics, health conditions, and environmental factors. Empirical research has shown that individual background characteristics, such as gender, age, and education; and health conditions including hospital admissions, cause of death, and pain management, are associated with the place of death (Cross et al., 2020; Jennings et al., 2020; Quinn et al., 2022).

Among the most significant environmental factors influencing the location of death are the accessibility and utilization of formal and informal care resources (Cohen et al., 2015). Informal care includes unpaid support for daily activities, emotional and financial assistance, as well as care management (Rabow et al., 2004). Within informal caregivers, adult children and spouses, play a crucial role in the care of older adults living in the community during the end-of-life period (Milligan et al., 2016). Findings from a few cross-sectional studies based on nationally representative surveys indicate that informal care was associated with higher odds of home deaths compared to formal care. Additionally, care provided by spouses and adult children exhibited a stronger association with home deaths compared to other informal caregivers (Ailshire et al., 2021; Lei et al., 2021; Xiong et al., 2022).

Given the increased care requirements in the end-of-life period, formal care may serve as a complementary source of care (Visser et al., 2004). Formal care, which may facilitate home deaths, encompasses publicly and privately funded services, including medical, domestic, social, and hospice in-home support. Additionally, a shortage of long-term care beds has also been associated with home deaths, especially in rural areas, suggesting that the location of death may be influenced by the lack of suitable alternatives and not only by personal preferences (Cohen et al., 2015).

Care Networks during the End-Of-Life Period

Care networks refer to the group of people who provide support in response to long-term health issues or functional limitations experienced by their care-receivers. It is argued that a complex intersection between informal and formal care exists, challenging the notion of considering them distinct and exclusive categories. This approach is important not only in identifying the diverse compositions of care networks but also in recognizing the extent to which informal caregivers share the burden of care tasks with others (Bijnsdorp et al., 2019; Jacobs et al., 2018).

Only a limited number of studies have aimed to identify care network types encompassing both formal and informal care mixtures among community-dwelling older adults in the Netherlands (Broese van Groenou et al., 2016b; Jacobs et al., 2018) and in Canada (Keating et al., 2003). Empirical evidence about the end-of-life period is even scarcer (Bijnsdorp et al., 2019). Findings from these studies reveal that spouses often provided care with minimal assistance from other caregivers, whereas children tended to share the caregiving responsibilities with others, resulting in a more diverse network. In cases where spouses or adult children were not available, formal care services became the primary source of support. Notably, care networks during the end-of-life period exhibited similar traits to those observed in the broader older frail population. Finally, a wide range of care-receivers characteristics within each network were examined, such as socioeconomic and health status, and structural aspects of the caregiving networks. To the best of our knowledge, network types comprising a combination of informal and formal care have not been specifically examined concerning the place of death.

The COVID-19 pandemic presented significant challenges to the delivery of end-of-life care, particularly during its early stages. Government-imposed restrictions aimed at reducing the spread of the virus led to decreased staffing levels and a shortage of formal care providers (Fallon & Kilbride, 2021; Hanna et al., 2021). Also, care-receivers often avoided medical appointments and formal care services due to fears of contracting the virus. Informal care also underwent substantial changes. In Europe, there was an overall reduction in informal care provision, largely attributed to restrictions on physical contact and concerns about transmitting the virus to care-receivers or contracting it themselves. However, adult children have increased their provision of care for older parents (Bergmann & Wagner, 2021). It seems that the composition of end-of-life networks during the pandemic has been overlooked in the existing literature. Therefore, this study aims to identify the different types of care networks present in the last year of life of older adults during the pandemic and examine their association with the place of death.

Study Hypothesis

Our hypothesis relies on the identification of care network types, and therefore is explorative. We propose that care networks characterized by a significant presence of informal caregivers, such as spouses and children, are associated with higher odds of home deaths compared to care networks characterized by a predominant presence of formal care providers.

Methods

Data and Participants

The data used in this study were derived from the Survey of Health, Ageing and Retirement in Europe. This is a longitudinal multidisciplinary panel survey that covers 27 European countries and Israel. The survey interviews community-dwelling respondents who are 50 years old and older every two years about different aspects of their lives. These include health status, socioeconomic status, and social networks (Börsch-Supan et al., 2013). The study involved end-of-life interviews where knowledgeable proxies, such as family members or friends, provided information about the respondent’s final year of life. These proxies were not necessarily the caregivers; for instance, a sibling could report that the spouse was the caregiver. Although professionals may have cared for participants, they were not considered proxies. The data from the interviews were extracted from the first Corona Survey (SCS1), conducted between June and August 2020 (Börsch-Supan et al., 2022a), and the second SCS (SCS2), conducted between June and August 2021 (Börsch-Supan et al., 2022b). Both SCS waves focus on the impact of the pandemic on older adult Europeans’ lives (Scherpenzeel et al., 2020). The information obtained from the end-of-life interviews includes details on care network types, place of death, and all control variables, except for the level of education of the deceased (that was taken from the most recent wave).

The analytical sample included respondents who passed away during the pandemic in 17 European countries and Israel. The goal was to examine care networks during the end-of-life stage in the pandemic context, therefore we selected June 2020 as the starting point, marking three months after the pandemic onset (WHO, 2020). The study concluded with the end date set at August 2021, coinciding with the end of data collection for SCS2. Our final sample consisted of individuals who received assistance with activities of daily living (ADL) and instrumental activities of daily living (IADL) in their final year. On average, participants had 4.30 ADL limitations (SD = 2.70; range = 0–7) and 6.59 IADL limitations (SD = 2.84; range = 0–9), indicating a high level of dependency and a need for continuous support. The total number of participants in this sample was 486 (Figure 1).

Figure 1.

Flowchart for participants’ selection. SCS1 = First Corona Survey. SCS2 = Second Corona Survey.

Study Variables

Dependent Variable: Place of Death

Proxies were asked about the location of death of the deceased. We computed a binary variable, assigning a value of 1 if the deceased died at home, and 0 if elsewhere (i.e., hospital, nursing home, and residential home).

Independent Variables

To identify the Care network types during the last year of life, proxies were asked to list up to three persons (including themselves) who provided care in ADL or IADL during the last year of the deceased. The care providers were grouped into categories: (a) adult children, (b) spouses, (c) other informal caregivers including relatives (grandchildren, siblings, and children-in-law) and non-kin (neighbors, friends, and unpaid volunteers); and (d) professionals (e.g., nurse). Based on these observed categories, we conducted Latent Class Analysis (LCA) to identify distinct care network types, which were then used as the independent variable in our analysis.

Control Variables

We controlled for individual background characteristics of the deceased, including age at the time of death (in years), gender (0 for men, 1 for women), and marital status at the time of death, dichotomized to living with a spouse or partner (1) or not (0). Level of education was based on the Standard Classification of Education (ISCED) 1997, including three categories: low level (ISCED 0, 1, and 2), middle level (ISCED 3 and 4), and high level (ISCED 5 and 6).

Health conditions in the last year of life were reported by proxies, including the number of hospitalizations. Poor mental health in the last month was categorized as feeling anxiety or sadness (1) or not (0). Pain management was recorded as binary, with 1 indicating the deceased experienced pain or took pain medication in their last month, and 0 if not. Causes of death were classified into categories including cancer, cardiovascular, respiratory, nervous system diseases, and other causes, with a separate category for COVID-19 deaths.

We took into account three macro-level covariates concerning the place of death: the differences between countries, the area of living, and the strictness of pandemic-related policy measures during the caregiving period. To assess the level of stringency, we utilized the Oxford COVID-19 Government Response Tracker data, which produces a single number per country per date ranging from 0–100 (Hale et al., 2022). Higher values signify more stringent policies. Caregiving duration, as reported by proxies, fell into several categories: Less than one month; One month or more but less than 3 months; 3 months or more but less than 6 months; 6 months or more but less than a year; and a full year. We aligned the Oxford data with the upper limit of the caregiving period of each deceased participant. This way, the final score reflects the average stringency score during the caregiving period. Area of living contained three categories: 1 = urban areas serving as a reference category; 2 = town areas; and 3 = rural areas.

Analytic Plan

Identification of Care Network Types during the Last Year of the Deceased

We conducted LCA using the Mplus 8.6 program (Muthén & Muthén, 2017) to identify distinct care network types during the last year of life. LCA allowed us to uncover hidden patterns in caregiving arrangements by grouping individuals based on observed caregiver categories: adult children, spouses, other informal caregivers, and professionals. We estimated models with one to three classes, as three is the maximum number of classes that can be identified with four binary dependent variables. We followed the recommendations of Nylund and colleagues (2007) for choosing the number of classes, which included interpretability considerations, a low log-likelihood chi-squared value, a high entropy index (indicating more reliable classification), and the smallest Bayesian Information Criterion (BIT) value (indicating better model fit). Additionally, we used the bootstrapped likelihood ratio test (BLRT) to evaluate the statistical significance of improvements in model fit with additional classes. After assigning each participant to a class, we used class membership as an independent variable in logistic regression analyses.

Data Analysis

First, we present the types of care network identified through Latent Class Analysis. Following that, we provide descriptive statistics for both the main and control study variables. The multivariate analysis included a binary logistic regression, examined the relationship between care network types and the place of death. All analyses, except for the Latent Class Analyses, were conducted using STATA 14.

Results

As presented in Table 1, approximately 40% of the respondents passed away at home, while the remaining individuals died in institutional settings such as hospitals and nursing homes. Among the study sample, around 40% were women, the average age at the time of death exceeded 80 years (M = 81.48, SD = 9.29), and above 60% were married at time of death. Among the participants, 57.82% had low education, 31.69% held middle level education, while 10.49% had attained a high level of education (Table 2).

Table 1.

Descriptive Statistics of Outcome Variable, Control Variables and the Social Network (n = 486).

	Mean (±SD / n (%)	Range
Dependent variable
Place of death
Home	198 (40.74%)
Other place	288 (59.26%)
Individual characteristics
Gender
Women	189 (38.89%)
Men	297 (61.11%)
Age at death	81.48 (±9.29)	56–102
Marital status at the time of death
Not married	185 (38.07%)
Married	301 (61.93%)
Level of education
Low education	281 (57.82%)
Middle education	154 (31.69%)
High education	51 (10.49%)
Health conditions
Number of hospitalizations in the last year	2.17 (±4.25)	0–43
Deceased felt anxiety and sadness in the last month a	259 (53.29%)
Deceased felt pain or took medication for pain in the last month a	333 (68.52%)
Cause of death
Cancer	131 (26.95%)
Cardiovascular diseases	165 (33.95%)
COVID-19	29 (5.97%)
Respiratory diseases	33 (6.79%)
Diseases of the nervous system	64 (13.17%)
Other	64 (13.17%)
Macro-level factors
Stringency index	49.09 (±16.69)	14.25–83.99
Area of living
Urban	91 (18.72%)
Town	207 (42.59%)
Rural	188 (38.68%)

^aReference categories: Poor mental health: did not feel anxiety and sadness in the last month; Pain and pain management: felt no pain/didn’t take medication for pain in the last month.

Table 2.

Comparison of Latent Class Models.

# Of classes	Likelihood ratio Chi-squared	BIC	p BLRT	Entropy	Proportion of the smallest class
1	205.93*** (df = 11)	2682.60
2	109.45*** (df = 6)	2617.46	<.001	0.66	0.46
3	54.03*** (df = 1)	2593.39	<.001	0.85	0.13

***p < .001.

According to the proxies, the average number of hospitalizations in the final year of life was just above two (M = 2.17, SD = 4.25). In the last month of their lives, proxies reported that over half of the respondents felt depressed or anxious, and almost 70% experienced pain. The two primary reasons for death were cardiovascular diseases and cancer. We also analyzed macro-level factors including the area of living, the level of stringency of the policy measures implemented during the pandemic while the participants received care, and the country of residence. The distribution of the area of living was as follows: 18.72% of participants lived in urban areas, 42.59% in towns, and 38.68% in rural areas. The average stringency index during the care period was 49.09 (SD = 16.69). For details on the number of participants in each country, see the Supplemental Appendix.

We conducted Latent Class Analyses to identify different care network types. Three models were estimated, with the best solution identifying three classes: (1) mixed-care network (85% adult children, <15% spouses, 60% other informal caregivers), (2) spouse network (100% spouses, ∼40% adult children, ∼20% professionals), and (3) professional network (100% professionals, ∼50% adult children, ∼25% other informal caregivers). The spouse network was the largest (47.74%), followed by the mixed-care network (38.48%), and the professional network was the smallest (13.79%). See Table 3 for more details. To evaluate measurement invariance, we also performed separate LCA based on demographic subgroups (see Supplemental Appendix).

Table 3.

Care Networks (n = 486).

	Mixed-care networkN (%)	Spouse networkN (%)	Professional networkN (%)
Adult children	158 (84.49%)	96 (41.38%)	33 (49.25%)
Spouses	25 (13.37%)	232 (100%)	0 (0%)
Other informal caregivers	107 (57.22%)	13 (5.60%)	16 (23.88%)
Professionals	0 (0%)	50 (21.55%)	67 (100%)
Total (N)	187 (38.48%)	232 (47.74%)	67 (13.79%)

Note. The other informal caregivers category includes relatives (grandchildren, siblings, children-in-law) and non-kin (neighbors, friends, unpaid volunteers).

^aReference categories: Poor mental health: did not feel anxiety and sadness in the last month; Pain and pain management: felt no pain/didn’t take medication for pain in the last month.

The logistic regression results shown in Table 4 indicate that there is a significant association between the type of care network and the place where one dies, after accounting for individual characteristics, health conditions, and the macro-level factors. Participants who received care from a mixed-care network were 2.16 times more likely to pass away at home compared to those who received care from a professional network (95% CI [1.04–4.47], p = .038). Additionally, participants who received care from a spouse network had borderline higher odds of dying at home compared to those receiving care from a professional network (95% CI [0.96–4.63], p = .063).

Table 4.

Odds Ratio and Confidence Intervals for Informal Care Network Types Within the Last Year of Life and Home Deaths: Logistic Regression (n = 486). ^a

	Or (95% CI)	p
Care network types b
Mixed-care network	2.16 (1.04–4.47)	.038
Spouse network	2.11 (0.96–4.63)	.063
Individual characteristics
Women b	1.17 (0.71–1.92)	.528
Age at death	1.02 (0.99–1.05)	.064
Married b	1.42 (078–2.55)	.243
Level of education b
Middle education	0.87 (0.51–1.51)	.742
High education	1.42 (0.67–3.03)	.174
Health conditions
Number of hospitalizations in the last year	0.97 (0.92–1.03)	.300
Deceased felt anxiety and sadness in the last month b	1.64 (1.04–2.58)	.032
Deceased felt pain or took medication for pain in the last month b	0.57 (0.35–0.93)	.023
Cause of death a
Cancer	14.99 (3.13–71.83)	.001
Cardiovascular diseases	10.03 (2.15–46.74)	.003
Respiratory diseases	5.79 (1.05–31.93)	.044
Diseases of the nervous system	35.22 (6.99–177.50)	.000
Other	7.30 (1.47–36.30)	.015
Macro-level factors
Stringency index	1.00 (0.99–1.02)	.337
Area of living a
Urban	0.71 (0.38–1.33)	.292
Town	0.82 (0.49–1.36)	.495
Pseudo R-square	15.04

^aAdjusted for countries.

^bReference categories: Care network types: professional network; Gender: male; Marital status: was not married at time of death; Level of education: Low level; Poor mental health: did not feel anxiety and sadness in the last month; Pain and pain management: felt no pain/didn’t take medication for pain in the last month; Cause of death: COVID-19; Area of living: rural area.

Discussion

The study aimed to examine the types of care networks that were involved in providing end-of-life care to older adults in European countries during the pandemic. The study also investigated whether these care networks were associated with the likelihood of home deaths. The care networks were classified based on the combination of informal and formal sources of care, and their association with home death was examined. The results showed that the two most common types of care networks were the spouse network and the mixed-care network, which mainly consisted of adult children. A smaller proportion was attributed to the professional network. These findings suggest that informal care plays a significant role in end-of-life care, aligning with previous research (e.g., Xiong et al., 2022).

The mixed-care network in our study revealed multiple informal caregivers, consistent with previous research that shows adult children often collaborate with others in caregiving (Broese van Groenou et al., 2016b; Jacobs et al., 2018). Our findings indicate that adult children significantly contribute to care within spouse networks, contrasting with prior studies in the Netherlands (Bijnsdorp et al., 2019; Jacobs et al., 2018). In Jacobs et al. (2018), the average age of care-receivers in a spouse-dominated network was 73.15 years, while Bijnsdorp et al. (2019) reported an average age of 76.6 years. In comparison, our study found the average age to be 77.9 years. Although the age difference is modest, it may indicate that the older spouses in our sample were more likely to experience frailty, resulting in a greater dependence on their adult children for caregiving. Furthermore, the broader European context of our study, which includes 18 countries with diverse caregiving norms and formal care systems, may also help explain these variations in caregiving patterns for older parents.

Upon examining the proportion of formal and informal care within each care network, we have identified different patterns in their collaborative functioning as resources. Our findings indicate that formal caregivers were not present in the mixed-care network. The absence of formal caregivers from the mixed-care network could be attributed to the unique circumstances of the pandemic. Informal caregivers faced challenges in navigating complex healthcare systems and acquiring caregiving information, which may have been exacerbated during the pandemic (Fallon & Kilbride, 2021). Also, during the pandemic, many people were afraid of contracting the virus, which led to them avoiding formal care services or canceling medical appointments (Bergmann & Wagner, 2021). However, pre-pandemic studies conducted in the Netherlands found that formal caregivers were present in care networks dominated by adult children (Bijnsdorp et al., 2019; Jacobs et al., 2018). Our study encompassed a wide variety of European countries, each with different welfare regimes that influence the availability and provision of long-term care services (Brenna & Di Novi, 2016). To understand better, future research should explore the extent to which formal care complements care provided by adult children.

The study found that about 20% of caregivers in the spouse network were formal caregivers, suggesting a complementary care model. Having a spouse as the primary caregiver, who lives with and assists the care receiver, may enhance contact with formal care providers compared to non-cohabiting caregivers (Broese van Groenou et al., 2016b). Formal care services were found to be the predominant source of support only when the spouses were not part of the care network. This may be explained by the prevailing societal perceptions that prioritize spouses as primary or default caregivers. Consequently, in situations where spouses are not available to provide care, formal care services can serve as an alternative caregiving resource (Broese van Groenou et al., 2016a). Overall, the varied combinations of informal and formal care in the end of life period illustrates that care providers operate as part of a complex and dynamic network (Keating et al., 2003), rather than being divided into separate formal and informal networks.

Our main objective was to investigate the associations between different types of care networks and the location of death. We found that about 40% of deaths occurred at home, which is higher than the European average of approximately one-third (Cabañero-Martínez et al., 2019). This higher proportion may reflect a trend of increased home deaths during the COVID-19 pandemic, as reported in other studies (i.e., Aldea-Eamos et al., 2023), as well as the relatively larger representation of Eastern European countries in our sample, where home deaths are more common (Srdelić & Smolić, 2022).

Our research showed that participants who received care from a mixed-care network (dominated by adult children) were more likely to die at home compared to professional networks. The spouse network showed a similar trend, reaching borderline significance. These results align with previous studies highlighting the vital role of informal care from adult children and co-resident caregivers in enabling home deaths among older adults (Ailshire et al., 2021; Xiong et al., 2022).

Our study found that both care arrangements, which were mainly provided by informal caregivers, achieved similar results. However, the lack of formal care services in the mixed-care network is concerning. Existing research emphasizes the crucial role of integrating formal care services to ensure adequate end-of-life palliative care, including pain management, easing anxiety and depression symptoms, and providing physical support (Franchini et al., 2021; Quinn et al., 2022). Limited access to formal care services, in particular during health pandemic, may lead to an increased burden on informal caregivers and potentially have harmful effects on their well-being (Salifu et al., 2021). Our research indicates that adult children play a pivotal role in providing end-of-life care. They were involved in all types of care networks, and when they were the main caregivers, the likelihood of dying at home was significant. This finding is essential to consider in light of current trends that suggest changes in family structures and sociocultural norms, which may lead to a decrease in solidarity within families. This could result in a substitution of family caregivers with non-kin social network members, such as friends (Broese van Groenou et al., 2016a). It is worth noting that the study was conducted during the pandemic when people tended to rely more on their immediate family members for caregiving. This resulted in a larger share of caregiving responsibilities for the closest family members (Lightfoot et al., 2021). Additionally, evidence from Europe has shown that during the pandemic, there was an increase in care provision to parents compared to other types of caregiver-care-receiver relationships (Bergmann & Wagner, 2021). Therefore, it is necessary to conduct further research in the post-pandemic era to determine whether this trend continues.

The study has limitations that should be considered when interpreting its results. The classification of care networks was based on end-of-life interviews with proxies rather than directly from participants. Proxies, however, were able to identify sources of help, including informal and formal caregivers, providing valuable insights. Additionally, the study focused solely on the pandemic period, which may affect the generalizability of the findings to the post-pandemic era. Notably, significant results were found even after controlling for policy restrictions and COVID-19 mortality. The pandemic has underscored the challenges faced by aging societies with high care demands but limited resources (Raiber et al., 2022). Further research is needed to understand care networks during end-of-life periods across different cultural contexts and welfare systems, as our study was limited by sample size.

Conclusions

Our research highlights the vital role of informal caregivers, particularly adult children, in facilitating home deaths for older parents in Europe. Future studies should investigate this group’s caregiving patterns and intergenerational dynamics for a deeper understanding of end-of-life care. Stakeholders and policymakers should recognize the specific needs of adult children caregivers and prioritize formal services that support their roles. Overall, efforts should aim to develop effective integrative care models for the end-of-life period.

ORCID iDs

Liora Cohen https://orcid.org/0000-0003-1458-5137

Sharon Shiovitz-Ezra https://orcid.org/0000-0002-7967-9137