Between Home and Hospital: Is There an Optimal Location of Death for the Dying Child?

Philippe Sylvestre

Marc‐Antoine Marquis

Nathalie Gaucher

Clinicians, researchers and healthcare administrators often consider location of death (LOD) as a key outcome to assess end‐of‐life care in paediatric palliative care [1]. Studies have sought to determine which LOD would be “best” for patients by asking patients, families and healthcare professionals about their preferences, and by exploring where the best care may be provided. Overall, the medical literature suggests that children, parents and healthcare providers favour home death when possible. However, a significant minority opt for a different setting when given the choice [2]. On a healthcare administration level, many policies and official guidelines (primarily for adults) now view home death as an indicator of “good quality” end‐of‐life care [3, 4]. However, in practice, discussing and planning LOD with children and their families is far more complex than any one‐size‐fits‐all approach. In this article, we explore the challenges inherent to these discussions and call for an individualised, goal‐concordant approach to discuss location of care, not only at the time of death but throughout the end of life.

1. Understanding Preferences Regarding LOD

A recent umbrella review examined LOD preferences from 15 systematic reviews, including over 110 000 patients (mostly adults) and 30 000 family members [5]. Home was the preferred LOD for patients (51%–55%) and family (17%–100%). Hospitals and hospice were preferred by 17% and 10% of patients, respectively. Studies exploring the preferences of children are very limited, and these suggest a home preference in 65%–85% of children [6, 7]. These data indicate that while the majority of people prefer home death, a significant proportion—up to 49% of patients, 83% of family members, and 35% of children—do not prefer to die at home.

For many adult patients and family members, home death is considered ideal: perceived as more peaceful, personal, surrounded by loved ones, with greater autonomy and dignity [3, 5]. However, concerns about symptom distress and family burden at the end of life are often cited as reasons against home death [5]. In paediatrics, very few studies have explored the reasons underlying child and parental preferences regarding LOD. For some parents, important considerations may support preferences for a hospital death: ensuring better symptom management and professional support, avoiding painful memories if their child were to die at home, and presence of other siblings at home [8, 9]. Johnston et al. reviewed the effects of child LOD on parental bereavement outcomes, suggesting that these may be better when their child dies at home [2]. However, they highlight numerous caveats to their conclusions: there are very few recent studies, many are small and prone to selection bias, and outcomes are hard to interpret and require nuance.

Many authors have cautioned against policies promoting an increase in home deaths in children, advocating instead for discussions that help identify and enable patients' preferred LOD [5]. Some have also suggested that the opportunity to plan LOD may be a better indicator of good‐quality end‐of‐life care than LOD per se [10]. Altogether, the current evidence does not support systematically aiming for home when a child's death can be planned. Preferences regarding LOD vary considerably; to ensure goal‐concordant care, it is critical to respect the wishes of those who prefer a location other than home.

2. Working Towards Goal‐Concordant Locations at the End of Life

Although enabling a LOD in line with each patient's preferred LOD may seem most appropriate, identifying and supporting this is complex. In practice, asking patients where they want to die does not suffice. How then, should preferences regarding LOD be discussed with children and their families?

2.1. From Preferred LOD to an Ethics of Care

A rhetoric of choice has reinforced the idea that individual autonomy at the end of life allows patients and families access to a death that aligns with their preferences [11]. However, asking patients where they want to die has its pitfalls. First, eliciting parental preferences regarding their child's LOD may be particularly challenging, because many parents remain undecided when confronted with these discussions [8]. Reasons for this indecision include: uncertainty of their child's prognosis; difficulty balancing diverging values (the comfort of home death versus the security of hospital death); diverging opinions (parents may have different preferences); and lack of recognition of the possibility of their child's death (and reluctance to participate openly in advance care planning surrounding death) [8]. Advocating for a child's preferred LOD is challenging when clear preferences are lacking.

Second, organisational challenges may undermine the ability to support goal‐concordant locations of death—even when preferences seem to be clear. The end of life of children—many having complex medical conditions—often requires considerable resources (technical, human, financial) which are not always available at the preferred LOD, notably at home [2]. These challenges shape preferences regarding LOD as they are inherent to balancing the pros and cons of different options [11]. Preferences are embedded within socioeconomic and family contexts (presence of siblings, family dynamics, provision of necessary medical care) oftentimes hard to capture when simply asking about preferred LOD. In particular, the availability of local hospice resources equipped to care for children (or lack thereof) is key to understanding the kind of support children may receive would they die out of the hospital [2]. Exploring these organisational contexts and challenges with families is crucial when advocating for goal‐concordant LOD in children.

Promoting goal‐concordant LOD calls for an ethics of care that goes beyond asking patients where they want to die. This shift in paradigm echoes Mol's distinction between a “logic of care” and a “logic of choice” [12]. It moves the focus away from explicit individual choices (“Where do you want your child to die?”), and instead towards a partnered understanding of the child's and family's experiences (“What is important for you and your child at the end of life? How does this inform preferences regarding location?”).

2.2. From Locality to Location Characteristics

Research exploring preferred LOD to date has focused on comparing different locations, mostly opposing “home” to “hospital”. This approach has created an apparent dilemma that may hinder discussions exploring goal‐concordant LOD. How can we rethink this dilemma? Rather than discussing locality (home, hospital, hospice), Sathiananthan et al. invite us to focus on LOD characteristics [13]. They found that “homely” characteristics—such as a place close to loved ones, that feels safe and familiar, that enables the exercise of autonomy and meets the needs of both the patient and the family—more often determine LOD preferences than the location itself. Shifting the focus from locality to characteristics allows for explorations of what truly matters to the patient and their family and highlights that there may be more than one “best” LOD for any given patient. A preference for a “homely” environment can be met in many settings, including hospice or hospital, provided these settings can cultivate the desired characteristics.

2.3. From Location at the Time of Death to Location of Care Throughout the End of Life

Studies define LOD (and preferred LOD) as location at the time of death. This reductionist concept overlooks the temporal dimension of dying [14]. For many patients—especially children with chronic medical conditions—the process of death and dying can span many days, weeks, even years. Supporting families towards goal‐concordant locations should look beyond location at the time of death, and emphasise the importance of location of care (LOC) throughout the end of life [3].

Focusing on LOC throughout the end of life fosters a relationship with location that keeps decisions open and embraces the inherent uncertainty in dying. The process of death and dying is dynamic, oftentimes with unclear prognoses, unpredictable care trajectories, and evolving healthcare needs. As circumstances change, so do patient and family preferences, goals, and needs. When asked about preferred LOD, up to 20% of patients change their preferences at least once throughout their end of life [3, 5]. The direction of changes may vary (hospital to home, home to hospice, home to hospital), as well as their underlying reasons (capacity to care, uncontrolled symptoms, acute medical events) [5]. Working towards goal‐concordant locations at the end of life should also be dynamic. Healthcare professionals' ongoing efforts to ensure a patient's environment is appropriate and meaningful throughout their end of life has recently been coined as “placing work” [14]. This concept highlights the need to continually adjust the patient's surroundings over time. It invites creative efforts to transform the environment—whichever it may be—and an understanding of patients' preferred LOD as an open‐ended manifestation of their goals and values.

3. Conclusions

Although a majority of children, parents and healthcare providers may favour home death when possible, a significant minority choose a different setting when given the choice. Exploring these preferences is complex, as they are embedded within organisational and family contexts, and inherently shaped by resources available. In this article, we have advocated for a shift of focus away from LOD and preferred LOD (PLOD), and instead towards what we have described as “goal‐concordant locations of care” (GCLOC) throughout the end of life (see Table 1 for a comparison of LOD, PLOD and GCLOC frameworks). Working towards GCLOC should look beyond location at the time of death. It should also go beyond asking patients and their families where they want to die. Focusing on location characteristics rather than location itself may make it easier to understand that, for any given child, there may be many “best” locations to die, and that any location where the family feels supported and aligned with their goals can be appropriate. This should inspire ongoing efforts to adapt patients' environments as they evolve throughout their end of life, and to foster a relationship with location that keeps decisions open and helps embrace the inherent uncertainty in dying. On a practical level, Table 2 highlights key considerations for working towards GCLOC in the clinical encounter. Additional training may be necessary to empower clinicians for these sensitive conversations. On an organisational level, working towards GCLOC should also inspire efforts to develop local hospice resources—free‐standing hospice and/or providing palliative care within the home—equipped to care for children, as these are key to ensure the support children may need outside of the hospital throughout their end of life.

TABLE 1.

Comparison of LOD, PLOD and GCLOC frameworks and implications.

Conceptual framework	Location of death (LOD)	Preferred location of death (PLOD)	Goal‐concordant locations of care (GCLOC)
What is advocated	Proxy for evaluating the quality of end‐of‐life care, promoting an increase in home death to match overall population preferences	Identify and enable patient preferences regarding location at the time of death	Explore the child's and family's goals of care throughout the end of life, and continually adapt their environment to ensure it is safe and meaningful
Role of patient preferences	Patient preferences considered on a population level, suggesting an overall population preference for home death	Individual preferences are central, articulated through a rhetoric of choice focused on locality (home, hospital, free‐standing hospice)	Individual preferences are central, explored through an ethics of care cognizant of organisational challenges (available resources, family dynamics) and focused on location characteristics
Temporal dimension	Location at the time of death	Location at the time of death	Locations throughout the end of life (may be multiple, changing over time and according to circumstances)
Approach	Which proportion of patients have died at home, in the hospital, and in a free‐standing hospice? Does this ratio correlate to population‐level preferences?	Where do you want your child to die? How can we make this happen?	What is important for you and your child at the end of life? How does this inform preferences regarding location? How can we continue to adjust and transform the environment accordingly?

TABLE 2.

Key considerations for working towards goal‐concordant locations of care.

	Example clinician interactions with patients and families
Clinical considerations	“Managing worsening symptoms (such as pain or respiratory distress) at home can be challenging for some parents, while others find it feasible. How do you feel about managing these symptoms at home?” “Some families find it difficult to manage oxygen therapy (or other medical technology) at home, while others feel comfortable doing so. What are your thoughts on this?”
Family considerations	“Caring for a sick child at home while also attending to other children can be difficult for some families. Others prefer it because it eliminates the need to travel between locations. What would work best for your family?” “Each family has a unique way of coordinating their child's care at home. Who currently provides medical care in your home? Is your current arrangement working for you?”
Organisational considerations	“Families have diverse organisational needs that may change over time or in response to specific circumstances. Are there any additional resources or forms of support that would be particularly beneficial to you at this stage?” “Are the resources and support systems you already have set up/in place adequate for continuing care in the current [home/hospital/hospice] setting? If not, what additional support would be helpful?”
Goals regarding location of care	“What is most important to you and your child at the end of life?” “At the end of life, different families prioritise different things—such as minimising suffering, being close to loved ones, or ensuring a sense of security. What matters most to you and what might your current priorities be?” “How can we adapt and optimise the [home/hospital/hospice] care environment to align with your family's goals and preferences?”

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.