Combining paid work and family care for a patient at the end of life at home: insights from a qualitative study among caregivers in the Netherlands

Background Population ageing, an emphasis on home-based care of palliative patients and policies aimed at prolonging participation in the labour market are placing a growing demand on working family caregivers. This study aimed to provide insight into experiences with combining paid work and family care for patients at the end of life, factors facilitating and hindering this combination, and support needs. Method Semi-structured interviews were held between July 2018 and July 2019 with 18 working family caregivers of patients with a life-threatening illness who were living at home. Transcripts were analysed following the principles of thematic analysis. Results Some family caregivers could combine paid work and family care successfully, while this combination was burdensome for others. Family caregivers generally experienced a similar process in which four domains — caregiver characteristics, the care situation, the work situation and the context — influenced their experiences, feelings and needs regarding either the combination of paid work and care or the care situation in itself. In turn, experiences, feelings and needs sometimes affected health and wellbeing, or prompted caregivers to take actions or strategies to improve the situation. Changes in health and wellbeing could affect the situation in the four domains. Good health, flexibility and support at work, support from healthcare professionals and sharing care tasks were important in helping balance work and care responsibilities. Some caregivers felt ‘sandwiched’ between work and care and reported physical or mental health complaints. Conclusions Experiences with combining paid work and family care at the end of life are diverse and depend on several factors. If too many factors are out of balance, family caregivers experience stress and this impacts their health and wellbeing. Family caregivers could be better supported in this by healthcare professionals, employers and local authorities. Supplementary Information The online version contains supplementary material available at 10.1186/s12904-021-00780-9.


Background
The care systems of many Western European countries, including the Netherlands, are under pressure due to population ageing. There is currently a shift to homebased care with the aim of curbing public expenditure and overcoming shortages in professional healthcare, and in response to people's preference for living at home until the end of life [1,2]. Care for patients living at home with palliative care needs often requires a substantial contribution from family caregivers [3]. Family caregivers provide unpaid care to relatives or friends, which can include nursing care, coordination of care, assisting with doctors' visits, transportation, administrative support, household care or emotional support [4]. This care is partly provided by family caregivers who have paid work as well, particularly by adult children who provide care to their parents and by partners of people who are confronted with a life-threatening illness when relatively young [5]. Changing policies in care provision and policies aimed at prolonging participation in the labour market, such as an increase in the statutory retirement age, place a growing demand on working family caregivers. Due to these shifts, it is expected that more people will have to combine paid work and family care in the near future [6,7].
In the Netherlands, formal care leave arrangements have recently been made available for workers that also provide family care. These arrangements include paid and unpaid leave for varying periods. For sudden, short episodes, there is the possibility of an emergency leave, which is a leave for few hours or days when a relative suddenly needs care, including full payment for at least one day. Short term care leave involves a maximum of twice the number of working hours per week over a period of 12 months. For example, workers with a 30-hour working week, are allowed 60 h of leave per 12 months. During this short term care leave, these workers are being paid at least 70 % of their salary. For longer episodes of family care, workers can take leave for a maximum of six times their number of working hours per week over a 12-month period. This leave is often unpaid. Even though there are different care leave arrangements available, these arrangements are hardly used by working family caregivers. The reasons for this remain unclear [8][9][10], while there seems to be a need for work adjustments.
A recent scoping review identified several challenges in combining family care and paid work, which included high or competing demands, psychosocial or emotional stressors, adverse health and financial pressure [11]. Family caregivers of a patient who is approaching the end of life are faced with the additional burden of coping with feelings of grief and loss, while they are often involved in intensive and complex caregiving tasks [12,13]. Also, it has been found that family caregivers of patients near the end of life experienced more negative health outcomes compared to other caregivers [12]. One in five working family caregivers of patients with a terminal illness experienced a heavy burden [8], and a heavy burden was also found among employed caregivers of older cancer patients [14].
In addition, the care that is needed for people with a life-threatening illness can be quite unpredictable and differs between patients and illness progression trajectories [15]. This could result in uncertainty about the care that is needed and might make it more difficult to combine caregiving and work responsibilities [16]. Prior research has indicated that unpredictable care needs and emergencies were a source of competing demands and often led to absenteeism from work as the caregiver unexpectedly had to stay with the care recipient or leave work. Also, family caregivers were sometimes unable to concentrate at work, causing a decline in work productivity, because they were worried about the care recipient [11]. In this sense, providing family care for patients with a life-threatening illness could have consequences for their participation in the labour market.
Previous studies have shown that the majority of working caregivers caring for patients at the end of life adapted their work situation (e.g. reduced their working hours, quit their job, took leave or changed jobs) [8,17]. Also, having to make a lot of work adjustments has been associated with worsening mental health of family caregivers [18]. To prevent this, it is crucial that working family caregivers are supported adequately.
Despite the growing attention paid to family caregiving in both the field of occupational health and the field of palliative care, research on the combination of paid work and family care for patients with a life-threatening illness is limited. Also, employers and occupational physicians often do not take the prognosis of the disease into account (e.g. curability), while this might be important for both the burden and the family caregiver's capacity for work (e.g. they might be more emotional). In the field of palliative care, little attention has been paid to the context of work in understanding caregiver burden, whereas work could interfere with caregiving tasks. Because of this, working caregivers might experience more of a burden and more restrictions due to dual responsibilities. The aim of this study was therefore to provide in-depth insight into the experiences and support needs of working family caregivers of patients with a life-threatening illness, barriers and facilitators in the combination of work and family care, and how this affects family caregivers. A broader understanding about this is of relevance to healthcare professionals, employers and policy makers as it will help them offer timely support for family caregivers who provide end-of-life care to a relative or friend.

Design
Data was used from a longitudinal qualitative study, in which in-depth interviews were held to better understand the experiences and support needs of working family caregivers of patients with a life-threatening illness over time. The interviews are still ongoing and are being held in one to four rounds (situation as at October 2020). This study draws on the first round of interviews. The consolidated criteria guidelines for reporting qualitative studies (COREQ) were followed for the reporting on the qualitative data (see Additional file 1) [19].

Recruitment and sampling
Participants were recruited by the researcher (FB) using two strategies: (1) Purposive sampling via general practitioners from different regions in the Netherlands. The general practitioners were asked to inform family caregivers about the study if they provided family care at home to a patient with a lifethreatening illness and had paid work for at least twelve hours per week. Life-threatening illness was defined as a serious or advanced illness in which the patient cannot be cured and is approaching death (e.g. incurable cancer, chronic lung disease, heart failure, dementia, progressive neurodegenerative disorder). The general practitioners provided the family caregivers with a participant information letter. Maximum variation in the sample was sought in terms of gender, working hours, sector, illness type and intensity of care. (2) Convenience sampling via posters in several departments of a Dutch academic hospital and an item in the corporate newsletter of this hospital with information about the study and contact details of the researcher. Potential participants could contact the researcher and the researcher gave them the participant information letter.
Family caregivers who were willing to participate first completed an online questionnaire (see Additional file 2), including questions about their gender, age, education, work characteristics (e.g. employed or self-employed, working hours per week and sector) and characteristics of the care situation (e.g. relationship to care recipient, type of illness, type of caregiving tasks, intensity of care and whether the care recipient lived at home or in a residential home). The questionnaire could also be filled in on paper on request. Participants were included if they were at least 18 years old, provided at least one hour of family care per week to someone with a life-threatening illness who was living at home, and had paid work for at least twelve hours per week. Dropout did not occur because family caregivers volunteered for participation themselves. The characteristics of the participants are presented in Table 1.

Setting and data collection
Semi-structured interviews were held with 18 family caregivers between July 2018 and July 2019, and were guided by a topic list (see Additional file 3). The interpretations of the researcher were checked by the participant after summarizing the most important themes during and at the end of the interview. The interviews were held at the participant's location of choice, which was often their own home, their workplace or the premises of VU University in Amsterdam. All interviews were audio-recorded and conducted by one female researcher who was trained in in-depth interviewing (FB). The duration varied between 45 and 120 min. After the interview, ad hoc field notes and a summary of the conversation were made to provide a first impression of the content. A summary of the results was returned to the participants for comment and/or correction. Data was collected until no new themes emerged from the data.

Data analysis
All interviews were transcribed verbatim and were analysed in Atlas.ti 8 following the principles of thematic analysis [20,21]. Data from the first five interviews was coded inductively and discussed by FB, HRP and HK. Discrepancies in the coding were discussed until consensus was reached. Some topics related to communication needed to be highlighted more in the following interviews and were added to the topic list. The remaining transcripts were analysed and coded by FB. Also, an across-case analysis was conducted, in which all text fragments associated with the code were explored to provide a broader understanding of the content of the codes [22]. The emerging codes were grouped into overarching themes and integrated into a framework to illustrate the relation between the themes. Interpretations, themes and the framework were discussed regularly with all team members. The framework was validated deductively by FB by rereading all transcripts and constantly comparing the framework with the transcripts. This led to additional changes to the framework, including the adaption of strategies and the impact on emotional and social wellbeing.

Results
The family caregivers who were interviewed generally experienced a similar process in which four domains, namely characteristics of the caregiver, the care situation, the situation at work and the context, influenced their experiences, feelings and needs regarding either the combination of paid work and care or the care situation in itself (see Fig. 1). Examples of caregivers' characteristics were the caregiver's own health, their attitude towards caregiving, their perceived caregiving skills, their feeling of responsibility, their wish for control and the ability to set boundaries. The care situation included the relationship with the care recipient, the illness trajectory,

-12 months
Family Daily prognosis and care intensity. The situation at work included working hours, support and understanding from the supervisor and colleagues, autonomy and flexibility. The broader context covered aspects such as the ability to share care tasks with others, and communication with healthcare professionals, municipalities or other organizations. In turn, experiences, feelings and needs sometimes affected the caregivers' health and wellbeing, or they initiated certain actions or strategies to improve the situation. For some caregivers, the adopted strategies could prevent deterioration in their health and wellbeing, and could influence the situation in multiple domains such as work or care provision. Likewise, changes in health and wellbeing could affect the situation (or the perceived situation) in these multiple domains. Table 2 illustrates the framework by relating two stories of working family caregivers looking after patients with a life-threatening illness. To ensure privacy, names and some characteristics have been changed. In the following section, the themes within the framework are explained in detail.
Impact of multiple domains on experiences, feelings and worries about combining work and care Keeping a balance between paid work and family care Some family caregivers felt they could combine work and care successfully. Key facilitators for keeping a balance between work and care demands were being able to set boundaries, sharing care tasks with family or friends, support and understanding from supervisors and colleagues at work, having autonomy at work, flexibility in working hours and being able to work remotely ( Table 3, Q1). Supervisors and colleagues in particular played an important role in how family caregivers experienced the combination of work and family care ( Table 3, Q2). Having short lines of communication with healthcare professionals helped family caregivers to make decisions effectively and plan ahead, which made the combination easier. Also, it was less complicated to combine work and care if there were few unpredictable, acute situations with the care recipient.  Other caregivers struggled with the combination of paid work and family care, and experienced this as burdensome. In general, a lack of the aforementioned facilitating factors hindered the combination of work and care, and increased the burden. Some caregivers found the combination of work and care difficult because they had to juggle so many responsibilities and hardly had time to rest (Table 3, Q3).

Relation between experiences with combination of work and care and (support) needs
Experiences, feelings and worries were often related to needs in care, at work or for their own health. For instance, caregivers who experienced barriers in their work often wanted more appreciation, support and guidance at work to help them find out what arrangements were available for them (Table 3, Q4). Some caregivers who experienced barriers in the care situation wanted to share care tasks with others more often, but were sometimes unable to arrange this. Others wanted more flexibility from healthcare professionals in scheduling care appointments (Table 3, Q5).

Impact of domains on experiences, feelings and worries about the care situation and illness trajectory Challenges in arranging care and support
In some cases, caregivers experienced consequences of a staff shortage in professional healthcare for the care recipient (e.g. long waiting lists), which put more pressure on them. Difficult communication with municipalities or other organizations proved to be an important theme. Problems included bureaucracy, lots of paperwork, unclear application procedures and files that went missing, which made it hard for caregivers to arrange the care that was needed (Table 3, Q6). This was timeconsuming and strenuous for them, on top of all the other responsibilities in work and care. For many family caregivers who experienced this 'struggle' with organizations, the care they were entitled to and the support options were unclear. They had trouble arranging support Integration into the framework James is 57 years old, works part-time in the health and social care sector and provides care for his wife, who suffers from a progressive neurological disorder and dementia. James finds the combination of paid work and family care burdensome. Because of the progressive neurological disorder and dementia, his wife's behaviour has changed and she is sometimes aggressive and unpredictable. James cannot share the care tasks with others because his wife does not accept help from others. When James is at work, he often receives phone calls from his wife, which distracts him sometimes. James has a senior position, but he wants less demanding tasks and does not want to work on weekends anymore because it is too much to combine this with the care for his wife. James discussed this with his supervisor, but there was no understanding for his situation and it was not possible to adapt his work situation. He feels pressured, he experiences a psychological burden and the situation costs him a lot of energy. When James starts experiencing physical complaints because of the stress, he decides that changes have to be made. He cannot change the situation at home because his wife does not accept help from others. Therefore he feels that his only option is to find another job.
Situation: For this family caregiver, the barriers are mainly at work (e.g. demanding tasks, working in weekends, no understanding from his supervisor, no arrangements to adapt his work) and in the care (e.g. the care recipient claims the family caregiver and does not accept help from others). Experiences and needs: Because of these barriers the family caregiver experiences the combination of work and care as burdensome.
Strategies and impact on health and wellbeing: His first strategy (discussing his situation with his supervisor) did not have the desired result as no arrangements could be made. This impacted his mental and physical health, which made the situation even more difficult. Then he adopted another strategy (searching for another job), which will impact his work situation if he changes jobs and will probably improve his health and wellbeing.
Emma is 52 years old, works fulltime in the financial sector and provides care to her father, who was diagnosed with incurable cancer three months ago. The prognosis is bad and they have started chemotherapy to prolong his life and ease symptoms. Emma expects that her father will not be around for much longer. She is determined to provide the care he needs until the end, together with her sister. Emma and her father have a close relationship and can talk about everything. Her father indicated that if the care becomes too intensive for her, he wants to go to a hospice. Knowing this gives Emma confidence that she will be able to cope with the situation. Emma has informed her supervisor and colleagues about her father's situation. They support her by listening and they help find solutions if necessary. Emma can be flexible in scheduling work appointments and can easily work remotely. Sometimes she works on her laptop in the hospital when her father is receiving chemotherapy. Emma likes her job and working provides respite from care. Because of the social support she receives at work, Emma does not experience barriers in the combination of work and care and therefore experiences no impact on her own health. However, knowing that she will lose her father soon makes her feel very sad and frustrated. For her the grieving process has already begun.
Situation: This family caregiver experiences few barriers in the combination of care and work since she has resources from work (e.g. social support, flexibility at work, ability to work remotely and respite from care

Experiences and needs
The frustrating thing is that I'm standing there next to him and I see it but I can't do anything. He's in pain, but I can't make the pain any less, I can't help him, I can't do anything, all I can do is watch that person deteriorating. There's no point in being sad because then you... I think it's a taboo for a family caregiver, you're not allowed to complain about how hard it is for you, or to say that you feel sad or need to cry; you don't do that because you're not the one dying. Right? (#4)

Q10 Grieving process Experiences and needs
Yes, that was a real 'bad news' talk. They were very clear in [the hospital], very honest, very good. And we looked one another in the eye, kind of this is it, this is the beginning of the and would have liked proper information and guidance in finding help (Table 3, Q7). Several caregivers were not satisfied with the communication between them and healthcare professionals, since they had to make a lot of phone calls themselves to find out medical results or treatment options (Table 3, Q8). In some cases, family caregivers felt there was a lack of aftercare from healthcare professionals when curative treatments stopped, and they wanted to be guided to the next step. In addition, when a relative needed care, daughters and sisters were more inclined to take up care responsibilities than sons or brothers. This sometimes caused tension between family members, since the women wanted to share their caregiving tasks more with their siblings. This was, however, difficult to accomplish, since (according to the caregiver) their siblings often did not have time to help, felt less responsible to provide care, or had other ideas about what would be suitable care (e.g. moving care recipient to a nursing home rather than providing intensive care at home).

Challenges related to illness progression
An unpredictable illness trajectory with acute exacerbations was identified as challenging, and family caregivers were unsure about what to expect in the future. This made it difficult to determine when to involve more (professional) care or when to take compassionate leave at work. Some caregivers felt frustrated, because they could not really help the care recipient, and could only watch their health decline (Table 3, Q9). Anticipatory grief was an important theme, as many caregivers said that they were already thinking about the impending death of their relative, and the corresponding grieving process had already begun (Table 3, Q10). For them, the situation was still emotionally demanding, even if they could combine work and care successfully. However, some caregivers indicated that taking care of their relative also gave a sense of fulfilment (Table 3, Q11). Caregivers found it difficult sometimes to cope with the fact that the care recipient's behaviour had changed. This was identified as a problem in the case of patients with dementia or neurodegenerative disorders. In some cases, caregivers stated that it was important for them or for the care recipient for care to be provided at home until death. This sometimes put more pressure on the family caregiver, since a nursing home or hospice was then not an option.

Impact of experiences and needs on strategies and action
Negative experiences and unmet needs caused the caregiver to initiate certain actions or strategies to improve the situation. Broadly, these strategies focused on three domains: changes in the care situation, changes at work, or improving their own ability to cope with the situation.

Strategies to improve the care situation
Strategies aimed at the care situation included arranging respite care, reaching out to other family members or friends to share care tasks, involving volunteers or privately paid care, or upscaling professional healthcare provision. For some caregivers it was not easy to involve other people in the care situation since the care recipient did not accept this. Several caregivers said that they did not want their paid work to suffer from the care situation and they tried to schedule or reschedule care appointments around their work schedule, or share care tasks with others (Table 3, Q12).

Strategies to improve work situation
Strategies aimed at work were adjusting working hours, taking compassionate care leave or taking unpaid leave. Notably, some family caregivers were unaware of the compassionate care leave options. Others did not take compassionate leave because they found it difficult to determine when they should take it, since working caregivers are only entitled to a limited number of days per year. They did not want to 'waste' the compassionate leave, because the care demands were unpredictable and the future was uncertain, or the financial impact would be too large (Table 3, Q13). Several family caregivers, mostly caregivers looking after a child or partner, expressed the view that the current compassionate leave scheme was not sufficient for their situation. Caregivers sometimes used other leave schemes to provide care, such as parental leave or vacation leave on a structural basis. Others saved vacation leave hours for acute care situations. Many caregivers made personal arrangements with their supervisor, including alternative or less demanding tasks, working more from home and flexible working hours, or they were referred to occupational health professionals. Less often, caregivers made arrangements with colleagues, such as trading shifts or taking over shifts. For some caregivers it was important to remain active at work, and work provided respite from the care situation. Several caregivers expressed the view that the care situation was more important to them than work and they tried to schedule or reschedule meetings at work to fit around care demands. In some cases, caregivers were searching for another job because they were unable to reconcile the current work and care demands. Sometimes caregivers explored whether early retirement was a solution, but often this was not feasible because of the financial impact.

Strategies to improve coping ability and mental health
Family caregivers sometimes sought help from a psychologist (e.g. cognitive therapy) or enrolled in a coaching programme (e.g. corporate welfare training or a lifestyle coach) ( Table 3, Q14). In some cases, family caregivers practiced mindfulness or yoga, or went cycling or on evening walks to clear their mind, which enabled them to cope better with the demands of work and care.
Others found an outlet in choir singing, or talking to family, friends or other caregivers, which could relieve tension. Many family caregivers tried to schedule free time for themselves to relax, but it was often difficult to stick to the time off they had set for themselves.

Impact of experiences and strategies on health and wellbeing and the situation in different domains
As shown in the framework, experiences and needs could impact the health and wellbeing of family caregivers. This was especially the case when family caregivers found the combination of work and care burdensome. They felt listless and tired, had a short temper, were often in a bad mood and the combination cost them a lot of energy. Some caregivers said that their life was no fun anymore and that the situation was overwhelming them. They experienced burn-out symptoms or physical complaints such as shortness of breath, tiredness, a need for antidepressants, panic attacks, headaches or crying episodes (Table 3, Q15). Getting enough sleep and having good health themselves proved to be important in enabling them to cope with the situation, while many caregivers worried a lot and were not able to get sufficient sleep (Table 3, Q16). For the majority of caregivers, the care situation impacted their social life and involvement in the community. They felt lonely because they had no time for their friends, hobbies, volunteer work or relaxation because of the care and work demands. Several caregivers mentioned that the situation also impacted their financial situation (Table 3, Q17). The strategies they adopted to improve the situation could directly impact work or care provision, for instance if caregivers reduced working hours or shared care tasks with others. If the strategies were successful, caregivers experienced less or no impact on their health and wellbeing. If the strategies did not have the desired results, their health and wellbeing were sometimes negatively affected. Declining health and wellbeing, in turn, often influenced the situation in the different domains as well. For example, if caregivers experienced burn-out symptoms due to the combination of work and care, work was impacted if the caregiver then searched for another job, or the care situation was affected as the caregiver was no longer able to perform all the care tasks.

Discussion
This study provides in-depth insight into the experiences and needs of working family caregivers of patients with a life-threatening illness, the barriers and facilitators in the combination of work and care, and how this combination affects family caregivers. The experiences, needs and impact on health and wellbeing were found to be very context-specific. Barriers in one domain (e.g. a lack of flexibility or insufficient arrangements at work) could be compensated by resources from other domains (e.g. sharing care tasks with others). However, when there were too many barriers and insufficient resources to overcome them, caregivers experienced the situation as burdensome. This was often problematic for their health and wellbeing if they could not adopt strategies to improve the situation. In turn, this could affect work and care, and subsequently further impact experiences and support needs.

Caregiver' experiences, challenges and facilitators
In line with earlier findings [8,23], experiences with the combination of paid work and family care for a relative with a life-threatening illness were not solely positive or negative, and did not depend on characteristics of work and care only. Even when caregivers could combine work and care successfully, the situation could still be burdensome or demanding. Anticipatory grief and emotional distress seemed to play an important role here, and in line with other research [24], often came with uncertainty about the illness, feelings of disruption of one's own life, and a changing relationship with the care recipient. Aside from this anticipatory grief, challenges and facilitators in combining work and care were quite similar to what has been identified among a broader population of working family caregivers. This included for instance high and competing demands from work and care, unpredictable care needs, work interruptions, travel distance and inflexible schedules, which could be compensated by sharing care tasks, making work adjustments or receiving support at work [11,25,26]. When the care recipient required intensive caregiving, the illness progression was unpredictable with acute exacerbations, and/or when the prognosis of the care recipient was bad, family caregivers often experienced a negative impact on their mental health. This was especially found when other circumstances were also unfavorable, such as few support options at work. If these unfavorable conditions were present for a longer period of time, it was more difficult for caregivers to cope with the situation.

Importance of the workplace
In accordance with prior research [25,26], support from supervisors and colleagues in the workplace was critical in how the caregivers in this study experienced the situation. Negative attitudes or limited understanding of the caregivers' situation contributed to conflict at work [11], sometimes forcing them to quit their job. On the other hand, work provided respite from the care situation for many caregivers. Consistent with earlier findings, this indicates that remaining active in the workforce is meaningful for family caregivers, as it could facilitate psychological respite and contribute to wellbeing [27]. Hence, it is key that family caregivers are supported in the workplace in such a way that they can retain a balance between work and care. Future research could further assess how communication about family caregiving at work takes place, and what supervisors need to do to give caregivers the support they need.

Comparison between framework and existing models
The themes that were integrated into the framework share a number of commonalities with existing models that define important factors for family caregivers' mental health. There are some clear similarities between the factors that proved to be important in our framework and the components of the Informal Care Integrative Model (ICIM) [28], which builds on the Model of Carer Stress and Burden [29] and the Job Demands-Resources model [30]. The ICIM states that caregiver characteristics, caregiving setting and social environment determine the subjective evaluation of balance or imbalance between demands and resources. This evaluation could then mediate between the consequences of the determinants and burn-out [28]. This is comparable to the situation in the multiple domains (i.e. caregiver characteristics, work, care and context) that relate to feelings and experiences in the current framework, which could in turn impact the health and wellbeing of the caregiver. In addition, the ICIM states that burn-out and physical or psychological issues could impact aspects of the caregiving context or social environment. Gérain and Zech [28] pointed out that these feedback loops have often been neglected in prior models, while it is critical to address the development of strain over time. These feedback loops were also incorporated in the current framework, where caregiver health and wellbeing could influence the situation in multiple domains.
To our knowledge, this was one of the few studies that provides in-depth insight into the experiences, challenges and facilitators of family caregivers for patients at the end of life in combination with paid work. The framework adds to academic knowledge about family caregivers at the end of life by showing that work characteristics and occupational demands and resources were important in determining experiences, needs, health and wellbeing. Also, we found that certain actions or strategies could mediate between experiences and health and wellbeing, or could modify the various domains. Future research could examine trajectories of experiences when combining work and care for a relative at the end of life and identify factors that influence this over time.

Methodological considerations
Using semi-structured interviews provided rich and detailed information about the current situation and experiences of working family caregivers of patients with a life-threatening illness. Moreover, the rich data prompted the development of our framework. This framework is partly in line with existing theory, but also provides new insights, which is a strength of this study.
Although we included all male caregivers who responded, the experiences of men combining paid work and family care might be underrepresented. The results did, however, not indicate clear differences in how male and female caregivers experienced the combination of work and care, in the strategies that they used to improve the situation, or the impact of the situation on mental health and wellbeing. In addition, at the population level, women are more often involved in family care than men [5,31]. However, in further research it might be good to give more emphasis to (including) male caregivers in order to perform appropriate gender analysis, since caregiving seems to be a women's unpaid job.
Family caregivers with a non-Western cultural background were also underrepresented, and the role of cultural norms and values in the experiences of these caregivers may have been overlooked as a result [28,32]. Furthermore, the family caregivers in the study generally worked in jobs with considerable autonomy, enjoyed flexibility at work and/or did not work fulltime. Caregivers working fulltime in jobs with little autonomy or with a lack of flexibility at work possibly did not have time to participate in the current study, while they might experience more difficulties with the combination of work and care. The majority of caregivers in this study worked in sectors with a high emotional workload (e.g. health and social care, and education), which might also have contributed to more emotional strain [33]. Caregivers with fulltime jobs and a lack of autonomy or flexibility deserve more attention in future research, since their experiences and challenges might be different. Also, although the educational level and job sector of the family caregivers provided some insights into the socio-economic position, specific information about income was not available. Since some caregivers noted that the financial impact of formal care leave arrangements would be too large, future research could also take the level of income into account.
The inclusion via general practitioners could on the one hand be considered a strong point, as they have information about patients with a life-threatening illness and are often also in contact with the family caregiver. However, on the other hand, this could also have contributed to selection bias, since they might have been more hesitant to invite family caregivers of patients in intensive care situations. The caregivers who provided the most hours of family care had applied to take part in this study themselves after they had seen the poster in the hospital. This suggests that even family caregivers in intensive situations are sometimes willing to participate in research and share their experiences. In addition, although we were interested in the experiences of family caregivers themselves, it should be considered that the results of this study are from a single stakeholder perspective. Future research could triangulate information from general practitioners and employers to increase the study's comprehensiveness and validate the study results.

Implications for healthcare professionals, employers and policy makers
This study indicated that there were some commonalities in support needs that have implications for policy and practice. Many working family caregivers of a relative with a life-threatening illness experienced anticipatory grief and emotional distress, which influenced their mental health in an adverse way. More psychological support might be needed to increase their emotional preparedness, and to guide them through the process until or after bereavement [34]. This fits the palliative care approach which is directed at supporting not only patients, but also their families [35]. This support need could be signalled by the healthcare professionals who are involved in the care trajectory of the care recipient. They could also advise caregivers about respite options, since many caregivers hardly had time to rest. In addition, healthcare professionals could improve guidance and aftercare when there are no more curative treatments by discussing questions and worries [36]. Increasing collaboration between family caregivers and professionals, and more involvement in planning care could help family caregivers to combine their care tasks with work demands.
In the workplace, supervisors could be better informed about the support options for family caregivers in their organization and how to deal with these situations, as understanding and open communication were experienced by caregivers as very supportive. Employers could provide more guidance about the available support options, such as compassionate care leave or other work arrangements, since family caregivers often had to find out themselves. Also, employers could facilitate consultation hours with an informal care expert, who could assess the caregiver's situation in multiple domains, identify solutions for unresolved issues, and arrange respite care or for administrative tasks to be taken over temporarily. This support could help family caregivers to feel better able to combine work and care, and could prevent burn-out and long-term absence from work. Also, compassionate care leave arrangements warrant more attention by policy makers, since family caregivers were hesitant to use this type of leave. Family caregivers of patients with an unpredictable illness trajectory often found it difficult to determine when to take this leave because of its limited nature. Hence, it could be that the people who need this type of leave the most are not able to use it as they don't want to 'waste' it when they might need it for worse times ahead. In addition, for some family caregivers it was financially not feasible to take compassionate care leave, since they would receive less or no financial compensation and the impact on their income would be too large.
Lastly, local authorities and health insurance companies could improve accessibility and application procedures, since many family caregivers experienced communication difficulties. They could increase awareness and advice about support services, as some caregivers were unaware of the available options. Policy makers should re-evaluate and simplify application procedures and provide information that is understandable for the general public. This could reduce feelings of pressure amongst caregivers and a sense that they are in a fight with organizations.

Conclusions
Experiences with combining paid work and family care at the end of life are diverse and depend on several factors. If too many factors are out of balance, caregivers experience stress and this impacts their health and wellbeing. To facilitate a successful combination of paid work and family care, flexibility and understanding at work, timely support from professional caregivers or support services and/or sharing care tasks with others are important. Family caregivers could be better supported in this by healthcare professionals, employers and local authorities.