Place of living at end-of-life according to cause of death: a comparative analysis of all decedents 70 years or older in 2009–2013 in Finland and Norway

Introduction

Background

Many individuals express a preference for dying at home (1-3), and home death is perceived as quality palliative care. Yet, the number of home deaths remain relatively low in many countries (4-8). For some, the preference for place of death changes as death approaches, most often from home to institution (6). Therefore, focus should be given to the number of days people spend at home at end-of-life, and not merely their place of death (6,9-11).

Rationale and knowledge gap

The preference and ability of individuals to stay at home towards the end of their lives depends on various factors, including the individual’s need for care, as well as predisposing factors like age and gender, and enabling factors such as sociodemographic status and access to formal and informal care (12-15). In a previous study, Forma et al. [2020] compared the place of death in Finland and Norway, and found that among individuals aged 70 years or older, 15% died at home in Finland, compared to 11% in Norway (7). The study also revealed that younger individuals and males had a higher likelihood of dying in hospitals, while older individuals and women were more likely to die in nursing homes. Another study by Cohen et al. [2010] identified a variation (ranging from 13% to 45%) in the number of home deaths across different European countries, highlighting the importance of cultural and social factors, and access to both formal and informal care in achieving a high number of home deaths (4).

Few have previously studied the place of care towards end-of-life, and those that, have primarily focused on cancer patients (12,13). The majority of deaths (>80%) occurs in those aged 70 years and above (16). As the population ages and longevity increases, the composition of causes of death at the population level will change (8,17). To address future requirements for palliative care, it is essential to gain a comprehensive understanding of where people currently receive care and how care depends on peoples cause of death, in addition to their predisposing and enabling factors—particularly for the oldest segment in the population (8).

Objective

In this study, our aim was to utilize national register data encompassing all deceased individuals in Finland and Norway between 2009 and 2013. The objective was to provide a comprehensive account of the number of days that individuals aged 70 years and above spent in various care settings during the last 6 months of their lives, including their own homes, hospitals, short-term care (STC) facilities, and long-term care (LTC) facilities. To explore the influence of factors related to care needs on the choice of care setting, we conducted stratified analyses based on the most common causes of death, including cancers, diseases of the circulatory system, diseases of the respiratory system, and mental and behavioral disorders (primarily dementia). Furthermore, we assessed how predisposing and enabling factors, including age, access to informal care (via marital status), and income, were associated with the duration of time spent in different care settings. Sex stratification was performed in the analyses, as sex interacts with the aforementioned enabling and predisposing factors. For instance, females tend to live longer than males, they die from different causes and are more likely to provide informal care, while being less likely to receive informal care (17,18). We present this article in accordance with the STROBE reporting checklist (available at https://apm.amegroups.com/article/view/10.21037/apm-23-269/rc).

Methods

The Norwegian Health Care System

In Norway, four Regional Health Authorities, which are state-owned and administered by the Norwegian Ministry of Health, are responsible for the organization and funding of specialist health care services, primarily hospitals (19). At the time of our analyses, approximately 425 municipalities were responsible for organizing primary care services, including general practitioners, home-based care, and various forms of institutionalized STC and LTC. LTC encompasses nursing homes, sheltered housing, round-the-clock care, and sheltered housing with 24-hour support.

Hospitals are funded by a combination of block grants, which is a fixed budget component that are not influenced by activity, and activity-based funding determined by diagnosis-related groups (DRGs) (20). Between 2009 and 2013, both funding sources contributed to approximately 50% of the total funding (21). The general practitioners (GPs) are funded by a combination of a per capita-component based on the number of patients on the GP’s list, a fee-for-services component and user charges. Home- and community-based health care services are funded by block grants and patient co-payments, with the amount depending on the type of service. Nursing homes receive funding through a block grant from the municipalities. Nursing home residents are required to pay an income-dependent user charge to the municipality, with the maximum charge set at 80% of their income. Patient co-payments for other services, such as outpatient treatment and visits to general practitioners, are modest and subject to an annual upper limit per patient. The health care system is primarily funded through taxes, and opting out of the public system is not possible. Approximately 8% of the population has private insurance, mainly provided by their employers (22).

The Finnish Health and Social Care System

The responsibility for organizing health and social services for citizens in Finland was transferred from municipalities to wellbeing services counties at the beginning of 2023. During the period covered by this article [2009–2013], approximately 320 municipalities were responsible for organizing health and social care for their residents. Municipalities provided services independently, in collaboration with other municipalities, or they could procure services from other municipalities or the private sector, including non-profit organizations and for-profit enterprises. Specialized health care was organized by 20 hospital districts, which were federations of municipalities (23). Hospital inpatient services in Finland are divided into primary health care and specialized health care. Primary health care is offered in primary care hospitals known as health centers, while specialized hospital care is provided in general hospitals, including central and district hospitals, as well as in university hospitals. Round-the-clock LTC is offered in two types of settings: institutional care and in round-the-clock housing services. Institutional LTC is available in nursing homes and in primary care hospitals, but the amount of this type of care has decreased rapidly. Currently, the primary form of LTC is sheltered housing with 24-hour assistance.

Services are predominantly funded through taxes, including municipal taxes and state transfers in the form of block grants. Service users are required to contribute a portion of the cost out of their pockets, with the specific share varying across different services. In Finland, clients of LTC services pay a larger share (about 20%) compared to other Nordic or European Union (EU) countries (24).

Data

Patient population

We identified all individuals 70 years or above, who died in Finland or Norway between 2009 and 2013 from the Norwegian Causes of Death Registry and the Causes of Death Register in Finland, which both cover 100% of the populations (25,26). From the registers, we got access to individual’s date and underlying cause of death, the latter noted as ICD-10 codes. We classified individuals according to the underlying causes of death, focusing on the largest groups: cancer (C00–C99), diseases of the circulatory system (I00–I99), diseases of the respiratory system (J00–J99), or mental and behavioral disorders (primarily dementia) (F00–F99). In addition, we reported on all deaths in the period [2009–2013], herein referred to as ‘all causes of death’.

Place of living during the last 6 months of life

In the paper, we divided place of living into days at home, days in STC facilities, days in LTC facilities, and days in hospital. Below, we describe how this information was identified in the registers in Norway and Finland separately.

Norway

In Norway, data regarding the patients place of living during their last 6 months of life was collected from the Norwegian patient registry (NPR) and the individual-based nursing and care statistics registry (IPLOS). Days in hospital was estimated using NPR, which includes information on all hospital treatments that individuals receive in Norway. Each time a treatment is provided to a patient, a claim is sent to NPR which, in addition to other information, includes the number of days that patients have spent in hospital.

Days in STC and LTC were estimated using IPLOS. Since home- and community-based services are not funded based on their activity in Norway, no claims are gathered and reported from these care providers. However, for the purpose of research, quality assurance, future planning and control, all municipalities are required to gather information on the number of patients who have applied for and/or received home and community-based care. The information is gathered in IPLOS. From IPLOS, we got information on how many days the decedents lived in institutions, dividing between STC facilities and LTC facilities. In most analyses (with except the descriptive statistics), LTC was grouped into one category.

Days at home were estimated by subtracting the length of stay (LOS) individuals had in hospital, in LTC, and in STC from the total number of days during 6 months (180 days). Since a patient’s place in a long-term facility is not used by others if she or he is absent (due to for example hospitalization), we allowed for an overlap between hospital stays and stays in LTC facilities.

Finland

Data on the place of living during the last 6 months of life were collected from the Care Register for Health Care and the Care Register for Social Welfare, maintained by the Finnish Institute for Health and Welfare. Service providers collect information on the use of health and social services (i.e., the use of LTC), for example admission and discharge days, and deliver it to the administrator of the registers, who in turn delivers the instructions and framework for data collection.

The days in hospital were identified from the Care Register for Health Care, which includes information on all inpatient care use of university, central, district and private hospitals. In addition, use of STC (LOS <90 days) in primary care hospitals were collected from the register. The use of LTC, i.e., nursing homes and sheltered housing (with and without 24-hour assistance) were collected from the Care Register for Social Welfare, and LTC (LOS ≥90 days) in primary care hospitals from the Care Register for Health Care.

Days at home were estimated by subtracting the number of days the individuals spent in the health care facilities (i.e., hospitals) and in LTC from the total number of days during the 6 months (180 days). The calculation of care days and days at home was based on the admission and discharge days in the Care Registers, and in addition, the number of days when the client is on leave or absent from LTC.

Individual characteristics

Information on individual characteristics was gathered from Statistics Norway and Statistics Finland. We used information on the individuals age at death, sex, marital status, and income. Age was grouped into five age groups: 70–74, 75–79, 80–84, 85–89, or ≥90 years old. Marital status 6 months prior to death was divided into three groups: (I) never married; (II) those currently married/registered as partner; and (III) those previously married, meaning, divorcees, widows/widowers or previously registered as partner. Marital status was used as an indicator for access to informal care; compared to those who were never married, we assume that the married were more likely to both have a spouse and children who could potentially offer informal care, while those who were previously married were more likely to have children who could potentially offer informal care. Income was grouped by quartiles for the entire cohort of patients dying; this means that all individuals were divided by whether they were in the quartile 0–24%, 25–49%, 50–74%, or 75–100%. The quartiles were estimated separately by sex and country, i.e., in Norway, females were divided by quartiles when compared to other females in Norway and so forth.

Statistics

Data were analyzed in parallel, meaning similarly but separately for Finland and Norway. Before analyses were performed, we made a protocol for how to run the analyses to ensure that all analyses were applied similarly across the datasets.

We used descriptive statistics to display sociodemographic and disease characteristics for the total population, and for the population according to their causes of death. We also used descriptive statistics to describe individuals living situation during the last 6 months of life, meaning, the total number of days individuals spent either at home, in hospital, in LTC or in STC during their last 6 months of life.

To estimate how individual characteristics (age, marital status, and income) were associated with the individual’s living situation, we ran separate multivariate regression models for all outcome variables (days at home, in hospital, in STC and in LTC). Regression model selection was done using the Norwegian data, by first assessing the characteristics of the outcome variables, and further, by assessing the best model fit among different appropriate models using the Akaike information criterion (AIC) and the Bayesian information criterion (BIC) (27-29). Since all outcome variables were count data, we tested the models ordinary least square regression, Poisson regression, and negative binomial. Since the distributions in the variables days at home and days in LTC contained a large number of non-users (many zero observations), we also tested whether a two-part model with a logistic regression in the first part and a generalized linear model in the second part could be appropriate to model these outcomes (29). After the above-mentioned procedure, we selected negative binomial regression as appropriate to model the number of days in hospital and in STC and two-part models with a logistic regression in the first part and a generalized linear model with a Gaussian family and an identity link in the second part to model the number of days at home and in LTC. These models were used both in the Norwegian and in the Finnish data. We used robust standard errors in all analyses.

The regression analyses were performed for all those dying, in addition, separately for patients depending on their cause of death: cancer, diseases of the circulatory system, diseases of the respiratory system and mental and behavioral disorders. All regression analyses were stratified by sex, since we expect the association between the individual characteristics of interest (age, marital status, and income) to differ between females and males. In the regression models we included the covariates age (70–74 years of age as reference category), marital status (never married as reference category) and income [low income (0–24%) as reference category].

Results from the regression models are presented as the average marginal effects (AMEs), thus variables are shown on their original scale (number of days) (30). We show the magnitude of the association for age, marital status, and income as the difference in the number of days at home, in hospital in LTC, and in STC between the reference category (the youngest age group, those never married, and those with the lowest income) to the categories: the oldest age group (90 years or older), currently or previously married, and those with the highest income (75–100%). Results from the full regression models can be seen in Tables S1-S8.

Ethics

The Norwegian Ethics Committee and the Norwegian Data Protection Authority, in addition to all the registry owners, approved this study. Registry owners gave administrative permission to access and use the data. Registry owners include the Norwegian Directorate of Health, the Cancer Registry of Norway, the National Institute of Public Health, and Statistics Norway. The Ethics Committee of Pirkanmaa Hospital District has approved the COCTEL study. Registry administrators (Statistics Finland and Finnish Institute for Health and Welfare) have given permission to access and use the data. Since this was a registry study, we received exemptions from consent. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Results

Descriptive statistics

In 2009–2013, a total of 186,017 and 159,756 individuals 70 years or older died in Finland and Norway, respectively (Table 1). Of these, in Finland (Norway), 20% (19%) died from cancer, 45% (30%) died from diseases of the circulatory system, 4% (9%) died from diseases of the respiratory system, and 6% (5%) died from mental or behavioral disorders (primarily constituting individuals dying from dementia). The majority of individuals (25%) were 85–89 years old when they died. In both countries, individuals dying from mental and behavioral disorders were older than individuals dying from cancers, and diseases of the circulatory and respiratory system. In both countries, most individuals were previously married (60%), while 30% were currently married and 10% were never married. Among cancer patients, more individuals were currently (45%), while fewer were previously (50%), married—while among those dying from mental and behavioral disorders, fewer were currently married (20%) and more were previously married (70%). In Finland, there were more individuals in the group never married (16%) compared to in the Norwegian patients dying from mental disorders (9%).

Description of individuals living situation

On average, decedents in Finland and Norway spent 96 and 84 days at home, 8 and 11 days in hospital, 64 and 80 days in LTC facilities, and 15 and 13 days in STC facilities, respectively. Individuals living arrangement differed according to cause of death; while individuals dying from cancer spent an average of 122 [113] days at home, 14 [20] days in hospital, 24 [35] days in LTC, and 23 [19] days in STC in Finland [Norway], individuals dying from mental and behavioral disorders spent an average of 29 [21] days at home, 4 [2] days in hospital, 142 [157] days, in LTC and 9 [6] days in STC, Figure 1 and Table 2.

Figure 1 The number of days individuals spent at home, in hospital, in LTC, and in STC during their last 6 months of life. Findings are shown for the entire population and when separating between cause of death. LTC, long-term care; STC, short-term care.

Age, marital status, and income

There was an association between individual’s age, marital status and income, and the number of days spent at home, in hospital, in STC and in LTC. Tables 3,4 show results from the regression models for all decedents. (For similar tables for the decedents by cause of death, see Tables S1-S8).

Age

Age was highly associated with days at home and days in LTC: when compared to younger individuals (70–74 years), older individuals (90 years or older) spent fewer days at home and more days in LTC. Figures 2,3 report this association by cause of death for females and males, displayed as the difference in the number of days spent in the different settings between the oldest age group compared to the reference category (70–74 years), which is at zero in the Y-axis.

Figure 2 The additional number of days that the oldest females (aged ≥90 years) stay at home, in hospital, in LTC, and in STC during the last 6 months of life, compared to the youngest females (70–74 years). The Y-axis represents the difference (AME) in the number of days females aged ≥90 years have at home, in hospital, in LTC, and in STC compared to females in those aged 70–74 years (who are at 0 on the Y-axis). LTC, long-term care; STC, short-term care; AME, average marginal effect.

Figure 3 The additional number of days that the oldest males (age ≥90 years) stay at home, in hospital, in LTC, and in STC during the last 6 months of life, compared to the youngest males (70–74 years). The Y-axis represents the difference (AME) in the number of days males aged ≥90 years have at home, in hospital, in LTC, and in STC compared to males in those aged 70–74 years (who are at 0 on the Y-axis). LTC, long-term care; STC, short-term care; AME, average marginal effect.

The pattern was consistent between the different causes of death although age was more strongly associated with place of living in Norway than in Finland. For example, in Norway, the difference in the number of days at home for the oldest females (90 years or older) compared to the youngest females (70–74 years) ranged from 50 to 71 fewer days between the different causes of death. In Finland, the difference in the number of days at home for the oldest females (90 years or older) compared to the youngest females (70–74 years) ranged from 28 to 60 fewer days. Corresponding numbers for the oldest males compared to the youngest males were 39 to 59 fewer days in Norway and 28 to 48 fewer days in Finland. Age also influenced the number of days individuals spent in hospital: for example, older females dying from cancer spent 16 and 12 fewer days in hospital compared to the youngest persons dying from cancer, in Norway and Finland respectively.

Marital status

When evaluating the total population, those who were married spent 29 and 14 days (for females) and 31 and 24 days (for males) more at home, in Finland and Norway, respectively, compared to the never married (Tables 3,4). If they were previously married, the number of days individuals spent at home increased by 6 days for females (in both countries) and by 8 and 12 days for males in Norway and Finland, respectively. Figures 4,5 show this association by cause of death for females and males, displayed as the difference in the number of days spent in the different settings between those currently or previously married compared to the reference category (never married), which is at zero in the Y-axis. Across the different causes of death, marital status (i.e., access to informal care) influenced the number of days at home and in LTC more than the number of days spent in hospital or in STC. In both countries, marital status was less associated with place of living for those dying from mental or behavioral disorders.

Figure 4 The additional number of days that the currently and previously married/partnered females stay at home, in hospital, in LTC, and in STC during the last 6 months of life, compared to the never married females. The Y-axis represents the difference (AME) in the number of days currently and previously married/partnered females have at home, in hospital, in LTC, and in STC compared to never married/partnered females (who are at 0 on the Y-axis). C, currently; P, previously; LTC, long-term care; STC, short-term care; AME, average marginal effect.

Figure 5 The additional number of days that the currently and previously married/partnered males stay at home, in hospital, in LTC, and in STC during the last 6 months of life, compared to the never married males. The Y-axis represents the difference (AME) in the number of days currently and previously married/partnered males have at home, in hospital, in LTC, and in STC compared to never married/partnered males (who are at 0 on the Y-axis). C, currently; P, previously; LTC, long-term care; STC, short-term care; AME, average marginal effect.

Income

Increasing income was associated with more days at home and fewer days in LTC in Norway. In Finland, income was, in contrast, associated with less days at home and more days in LTC. Figures 6,7 show the association between income and place of living by cause of death, displayed as the difference in the number of days spent in the different settings between those in the highest income quartile (75–100%) compared to the reference category: those in the lowest income quartile (0–24%), which is at zero in the Y-axis.

Figure 6 The additional number of days that the females in the highest income group (75–100%) stay at home, in hospital, in LTC, and in STC during the last 6 months of life, compared to females in the lowest income group (0–25%). The Y-axis represents the difference (AME) in the number of days females in the highest income group have at home, in hospital, in LTC, and in STC compared to females in the lowest income group (who are at 0 on the Y-axis). LTC, long-term care; STC, short-term care; AME, average marginal effect.

Figure 7 The additional number of days that the males in the highest income group (75–100%) stay at home, in hospital, in LTC, and in STC during the last 6 months of life, compared to males in the lowest income group (0–25%). The Y-axis represents the difference (AME) in the number of days females in the highest income group have at home, in hospital, in LTC, and in STC compared to females in the lowest income group (who are at 0 on the Y-axis). LTC, long-term care; STC, short-term care; AME, average marginal effect.

This pattern (with contrasting findings as income increased) was consistent across individuals dying from the different diseases, except for in individuals dying from mental disorders, where there was a positive association between income and the number of days that females spent at home (in both countries).

Discussion

Key findings

In summary, we examined the residency of individuals aged 70 years or older during the last 6 months of their lives in Finland and Norway between 2009 and 2013. Both countries had similar durations of hospital stays (8 and 11 days) and stays in STC facilities (15 and 13 days). However, Finnish decedents spent more time at home (96 days compared to 84 days) and less time in LTC facilities (64 days compared to 80 days) than their Norwegian counterparts. The living arrangements of individuals based on cause of death followed the same pattern in both countries. In both Finland and Norway, a lower age and the availability of informal care were associated with a greater number of days spent at home, for both males and females. In Norway, a higher income was associated with more days at home and fewer days in LTC facilities, whereas the opposite was observed in Finland. As the population ages, a larger number of individuals will likely need LTC at the end of their lives. Therefore, future healthcare services should be developed in accordance with these needs.

Strengths and limitations

The main strength of our paper was our access to rich data on service utilization from reliable national registries, which made it possible to model complete care pathways for all individuals, 70 years and above, dying in Norway and Finland. Based on information on composition of treatment and care, we were able to identify the number of days individuals spent at home during the last 6 months of life. In addition, we linked individual level data based on personal identification numbers which gave us a unique analytical opportunity allowing us to estimate the associations between the number of days individuals stayed at home, in hospital, in LTC or in STC at end of life, and socio-demographic variables (in terms of age), informal care (in terms of marital status), and socio-economic variables (in terms of income).

A weakness was, however, the requirement that the data in Norway had to be anonymous, because of requirements in the IPLOS. Consequently, we had to restrict the type of variables we could receive, and the level of accuracy for some variables. For instance, income had to be grouped into quartiles by sex. Also, we could not receive information on individual’s municipality of residence. If we had access to the latter, we could have understood more of the underlying mechanisms in our findings. For example, knowing an individual’s municipality of residence would allow linkage to municipal characteristics, such as demography, budgets, total population, and coverage rate of home and institutionalized care services, which could provide more information about the supply side of health care services. There is open access to information on municipal level characteristics through Statistics Norway’s web pages, and future analyses should exploit these variables (31).

Furthermore, the underlying cause of death of individuals can be difficult to determine, and particularly so due to multimorbidity. Information about individual’s cause of death in the Cause of Death registries is based on notes from the physician who completes the death certificate. In the certificate, both underlying and immediate cause of death can be reported (25,26,32). As people age, the level of multimorbidity increases (33). Stating the exact cause of death might therefore be difficult, both because of the difficulty in distinguishing between the underlying and the immediate cause of death, and because individuals might have two or more chronic conditions simultaneously (multimorbidity), which can both be noted as their underlying cause of death (32). For example, if an individual has cancer and chronic obstructive pulmonary disease (COPD), but dies from influenza—what would be the correct cause of death? In the future we can linked data on the patients’ diagnosis (from the hospital registers) to data on patients underlying cause of death, to be able to classify individuals more carefully according to disease and cause of death. The consequences of more information of the true underlying cause of death on place of residence according to cause of death is, however, not clear.

Also, the data used in our analyses are 10–14 years old. Changes that have occurred during the last decade might have influenced the number of days that people spend in different institutions and at home, at their end-of-life. For example, policymakers in both Norway and Finland have aimed at shifting care from the secondary care level to the primary care level, thus possibly influencing the number of days people spend in hospital and in short- and long-term nursing homes (34). Simultaneously, time at home and home deaths are increasingly encouraged, which might have shifted care from formal institutions to home-based care (9,11). Previous research on the place of death indicates that fewer individuals in Norway died in hospital (34% vs. 46%) or at home (14% vs. 18%) in 2011 compared to 1987 (8). Whether there has been a shift in place of care during the last decade is an empirical question that should be explored in future analyses.

Comparison with similar research

To the best of our knowledge, few studies have previously analyzed the time that individuals spend in different care settings at end-of-life, as the focus has been given to place of death rather than place of residence during end-of-life (4,5,7,8). However, place of death seems to vary greatly between countries, for example, Cohen et al. [2010] found that home death among cancer decedents occurred in 12.8% in Norway—while home death was more common in England (22%), Wales (23%), and Belgium (28%), and even more so in Italy (36%) and in the Netherlands (45%) (4). Furthermore, Bekelman et al., found that the percentage of individuals with cancer who died in acute care hospitals was substantial across the countries included in their analyses: 51% in Belgium, 52% in Canada, 45% in Norway, 42% in England, and 38% in Germany (5). Neither of these studies reported place of residence at end-of-life.

Explanations of findings

The current study is a descriptive study, and results should be interpreted as such. Still, our findings give new insight to the literature. The place of residence at end-of-life was similar in Finland and Norway giving an indication of the duration individuals 70 years or older live at home and in institutions during their last 6 months of life in countries where individuals are covered by national health insurance through taxes. These estimates can be used to indicate future need for of home and institutionalized care at end-of-life, given current composition of care services. However, it should be noted that we do not know the actual need for care in these individuals, nor whether the current level of care actually responds to individuals need.

We find that, in both Finland and Norway, the cause of death (as an indicator for underlying health condition) was associated with place of residence during the last 6 months of life. In addition, and for all health conditions, age was strongly associated with place of residence. In the future, the distribution of causes of death will change due to, among other explanations, an aging population. For instance, the age standardized rates of death from cancer, cardiovascular diseases, and diseases in the respiratory systems are declining, while death from dementia and mental disorder are increasing. However, because of the aging population, the absolute number of individuals dying from all these causes might still increase, but dementia will likely increase at a faster relative pace (17). This should be carefully considered when organizing health care services in the future to meet the need for LTC which will likely increase.

We also find that marital status, which was used as a proxy for informal care, was associated with place of residence. Those who were currently married or with a partner lived longer at home indicating that a spouse or a partner was the most important informal caregiver. Those who were previously married also lived longer at home, indicating that children were also an important informal caregiver. It is important to point out that we did not observe the amount of informal care provided. When the population is aging, there might be a reduction in the likelihood that individuals will have access to informal care. As age and level of multimorbidity increases, the likelihood that spouses will be able to care for each other might decrease. Also, the aging population is expected to not only increase the demand for health care services, but also reduce the relative supply of people in the workforce; while the number of working-age people per person above the age of 60 years was 4.7 in 2000, it is expected to decrease to 3.5 in 2030 and 2.9 by 2050 (35,36). The pressure of the working-age population to participate in both full-time paid work and in informal care will be challenging.

The associations between income and place of residence vary between Norway and Finland. In Norway, higher income was associated with an increasing number of days at home, in hospital, and STC. In Finland, higher income was associated with fewer days at home and higher number of days in LTC. We cannot explain these findings, however, we can hypothesize why the associations were like this. First, in Norway, as in most other countries, individual incomes are higher in urban than in rural areas. The levels of LTC services are, on the other hand, higher in rural than in urban areas due to national distribution rules inherent in the grant system between the central state and the municipalities (37). Individual income at municipal level is therefore negatively correlated with total municipal revenues. The positive effect of income on number of days at home could therefore be a spurious effect and the ‘real’ effect could be one of low level of municipal services. Another explanation might be that LTC in Norway to a large degree is financed by income-dependent out-of-pocket payments, where residents pay up to 85% of their income (22). Consequently, those with high income have a lower incentive to move to a nursing home since they will have to pay more compared to their less wealthy counterparts. A further explanation for the findings in Norway might be that individuals with higher income are more likely to live in urban areas. Their preference to stay at home might therefore be higher, because of the safety that closeness to acute care hospitals provides. If this was the case, income would be a confounder for urban vs. rural living. Also, the same pattern (of higher income in individuals living in urban areas) is true in Finland, thus, this would not explain the differences identified between Finland and Norway. We could hypothesize with several other explanations; however, the most important point is that these finding should be further explored in future analyses, where more detailed information is available. An important component is that a previous study from Finland found that LTC was more common among people with lower income, thus, why this was not the case in the population under analyses in this paper is puzzling (38).

Implications and actions needed

In the current study, we find that residency towards end-of-life differed similarly among older individuals (age >70 years) by cause of death in the two countries; individuals dying from mental and behavioral disorders (primarily dementia) spent nearly all their time in LTC facilities. In both countries, and for all causes of death, lower age, and availability of informal care (indicated by marital status) were associated with more days at home, for both males and females. As population’s age and the causes of death shift, while the access to informal care is likely to decrease, individuals will likely need more formal care and spend less time at home and more time in LTC facilities towards end-of-life. This will increase the need for institutionalized end-of-life care. The political goals, which has currently been to increased home-based care, might need to shift to align with the populations’ actual need for care during their last months of life.

Conclusions

In summary, we examined the residency of individuals aged 70 years or older during the last 6 months of their lives in Finland and Norway between 2009 and 2013. Both countries had similar durations of hospital stays (8 and 11 days) and stays in STC facilities (15 and 13 days). However, Finnish decedents spent more time at home (96 days compared to 84 days) and less time in LTC facilities (64 days compared to 80 days) than their Norwegian counterparts. The living arrangements of individuals based on cause of death followed the same pattern in both countries. In both Finland and Norway, a lower age and the availability of informal care were associated with a greater number of days spent at home, for both males and females. In Norway, a higher income was associated with more days at home and fewer days in LTC facilities, whereas the opposite was observed in Finland. As the population ages, a larger number of individuals will likely need LTC at the end of their lives. Therefore, future healthcare services should be developed in accordance with these needs.

Acknowledgments

We would like to thank MSc Jani Raitanen for valuable contributions to analyses.

Funding: The research was supported by the Norwegian Cancer Association (No. 208164), the Research Council of Norwegian (Nos. 295989 and 296114), and the Academy of Finland through the funding for the Centre of Excellence in Research on Ageing and Care (Nos. 326567, 336670, and 335870).

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, Annals of Palliative Medicine for the series “Value of Palliative Care”. The article has undergone external peer review.

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://apm.amegroups.com/article/view/10.21037/apm-23-269/rc

Data Sharing Statement: Available at https://apm.amegroups.com/article/view/10.21037/apm-23-269/dss

Peer Review File: Available at https://apm.amegroups.com/article/view/10.21037/apm-23-269/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://apm.amegroups.com/article/view/10.21037/apm-23-269/coif). The series “Value of Palliative Care” was commissioned by the editorial office without any funding or sponsorship. G.W.B. served as the unpaid Guest Editor of the series. E.A. and G.W.B. received funding for the project from Norwegian Cancer Associations—research grant (No. 20816). The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The Norwegian Ethics Committee and the Norwegian Data Protection Authority, in addition to all the registry owners, approved this study. Registry owners gave administrative permission to access and use the data. Registry owners include the Norwegian Directorate of Health, the Cancer Registry of Norway, the National Institute of Public Health, and Statistics Norway. The Ethics Committee of Pirkanmaa Hospital District has approved the COCTEL study. Registry administrators (Statistics Finland and Finnish Institute for Health and Welfare) have given permission to access and use the data. Since this was a registry study, we received exemptions from consent.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Yao CA, Hu WY, Lai YF, et al. Does dying at home influence the good death of terminal cancer patients? J Pain Symptom Manage 2007;34:497-504. [Crossref] [PubMed]
2. Nilsson J, Blomberg C, Holgersson G, et al. End-of-life care: Where do cancer patients want to die? A systematic review. Asia Pac J Clin Oncol 2017;13:356-64. [Crossref] [PubMed]
3. Cai J, Zhang L, Guerriere D, et al. Where Do Cancer Patients in Receipt of Home-Based Palliative Care Prefer to Die and What Are the Determinants of a Preference for a Home Death? Int J Environ Res Public Health 2020;18:235. [Crossref] [PubMed]
4. Cohen J, Houttekier D, Onwuteaka-Philipsen B, et al. Which patients with cancer die at home? A study of six European countries using death certificate data. J Clin Oncol 2010;28:2267-73. [Crossref] [PubMed]
5. Bekelman JE, Halpern SD, Blankart CR, et al. Comparison of Site of Death, Health Care Utilization, and Hospital Expenditures for Patients Dying With Cancer in 7 Developed Countries. JAMA 2016;315:272-83. [Crossref] [PubMed]
6. Agar M, Currow DC, Shelby-James TM, et al. Preference for place of care and place of death in palliative care: are these different questions? Palliat Med 2008;22:787-95. [Crossref] [PubMed]
7. Forma L, Aaltonen M, Raitanen J, et al. Place of death among older people in Finland and Norway. Scand J Public Health 2020;48:817-24. [Crossref] [PubMed]
8. Kalseth J, Theisen OM. Trends in place of death: The role of demographic and epidemiological shifts in end-of-life care policy. Palliat Med 2017;31:964-74. [Crossref] [PubMed]
9. Kaasa S, Payne S, Sjøgren P. Challenges related to palliative care. Lancet Oncol 2011;12:925-7. [Crossref] [PubMed]
10. Peppercorn JM, Smith TJ, Helft PR, et al. American society of clinical oncology statement: toward individualized care for patients with advanced cancer. J Clin Oncol 2011;29:755-60. [Crossref] [PubMed]
11. Helsedirektoratet. Nasjonalt handlingsprogram med retningslinjer for palliasjon i kreftomsorgen. 2019. Available online: https://www.helsedirektoratet.no/retningslinjer/palliasjon-i-kreftomsorgen-handlingsprogram
12. Bjørnelv GMW, Edwin B, Fretland ÅA, et al. Till death do us part: the effect of marital status on health care utilization and costs at end-of-life. A register study on all colorectal cancer decedents in Norway between 2009 and 2013. BMC Health Serv Res 2020;20:115. [Crossref] [PubMed]
13. Bjørnelv G, Hagen TP, Forma L, et al. Care pathways at end-of-life for cancer decedents: registry based analyses of the living situation, healthcare utilization and costs for all cancer decedents in Norway in 2009-2013 during their last 6 months of life. BMC Health Serv Res 2022;22:1221. [Crossref] [PubMed]
14. Andersen RM. Families' use of health services: a behavioral model of predisposing, enabling and need components. West Lafayette: Purdue University; 1968.
15. Babitsch B, Gohl D, von Lengerke T. Re-revisiting Andersen's Behavioral Model of Health Services Use: a systematic review of studies from 1998-2011. Psychosoc Med 2012;9:Doc11. [PubMed]
16. Statistics Norway. 12954: Døde per uke, etter kjønn og alder. Foreløpige tall 2000U01 - 2024U14. [cited 2023 Jul 3]. Available online: https://www.ssb.no/statbank/table/12954
17. Folkehelseinstituttet. Tall fra Dødsårsaksregisteret 2021. 2022 [cited 2023 Jul 3]. Available online: https://www.fhi.no/op/dodsarsaksregisteret/tall-fra-dodsarsaksregisteret-2021/
18. Jakobsson N, Kotsadam A, Syse A, et al. Gender bias in public long-term care? A survey experiment among care managers. Journal of Economic Behavior & Organization 2016;131:126-38. [Crossref]
19. Hagen TP, Kaarbøe OM. The Norwegian hospital reform of 2002: central government takes over ownership of public hospitals. Health Policy 2006;76:320-33. [Crossref] [PubMed]
20. Biørn E, Hagen TP, Iversen T, et al. The effect of activity-based financing on hospital efficiency: a panel data analysis of DEA efficiency scores 1992-2000. Health Care Manag Sci 2003;6:271-83. [Crossref] [PubMed]
21. Yin J, Lurås H, Hagen TP, et al. The effect of activity-based financing on hospital length of stay for elderly patients suffering from heart diseases in Norway. BMC Health Serv Res 2013;13:172. [Crossref] [PubMed]
22. Lindahl AK. International Profiles of Health Care Systems. The Norwegian health care system. The Commonwealth Fund. 2017. Available online: http://www.commonwealthfund.org/publications/fund-reports/2017/may/international-profiles
23. Keskimaki I, Tynkkynen LK, Reissell E, et al. Finland: health system review. Health Systems in Transition 2019;21:1-166. [PubMed]
24. European Commission. State of Health in the EU. 2022 [cited 2023 Jul 3]. Available online: https://health.ec.europa.eu/state-health-eu_en
25. Folkehelseinstituttet. Dødsårsaksregisteret. Available online: https://www.fhi.no/hn/helseregistre-og-registre/dodsarsaksregisteret/
26. Statistics Finland. Causes of death. Available online: http://www.stat.fi/til/ksyyt/index_en.html
27. Schwarz G. Estimating the dimension of a model. Ann Stat 1978;6:461-4. [Crossref]
28. Akaike H. Statistical predictor identification. Ann Inst Stat Math 1970;22:203-17. [Crossref]
29. Deb P, Norton EC, Manning WG. Health econometrics using Stata. 1st ed. College Station: Stata Press; 2017.
30. Williams R. Using the margins command to estimate and interpret adjusted predictions and marginal effects. Stata J 2012;12:308-31. [Crossref]
31. Statistics Norway. KOSTRA. 2024. Available online: https://www.ssb.no/offentlig-sektor/kostra/statistikk/kostra-kommune-stat-rapportering
32. Pedersen AG, Ellingsen CL. Data quality in the Causes of Death Registry. Tidsskr Nor Laegeforen 2015;135:768-70. [Crossref] [PubMed]
33. Salive ME. Multimorbidity in older adults. Epidemiol Rev 2013;35:75-83. [Crossref] [PubMed]
34. Norwegian Ministry of Health and Social Affairs. St.meld. nr. 47 (2008-2009). Samhandlingsreformen. 2009. Available online: https://www.regjeringen.no/no/dokumenter/stmeld-nr-47-2008-2009-/id567201/
35. Statistics Norway. Nasjonale befolkningsframskrivinger. 2022 [cited 2023 Jul 3]. Available online: https://www.ssb.no/befolkning/befolkningsframskrivinger/statistikk/nasjonale-befolkningsframskrivinger
36. Norwegian Ministry of Health and Social Affairs. Meld.St.29 (2012-2013). Morgendagens omsorg. 2013. Available online: https://www.regjeringen.no/no/dokumenter/meld-st-29-20122013/id723252/
37. Regjeringen.no. NOU 2022: 10. Inntektssystemet for kommunene. 2022. Available online: https://www.regjeringen.no/no/dokumenter/nou-2022-10/id2925606/
38. Nihtilä E, Martikainen P. Household income and other socio-economic determinants of long-term institutional care among older adults in Finland. Popul Stud (Camb) 2007;61:299-314. [Crossref] [PubMed]