This study aimed to describe the trajectories in the burden of working family caregivers of patients with a life-threatening illness, and to describe which factors in paid work and care were related to changes in the burden over time. In line with prior research [21, 22], almost all family caregivers in this study experienced some level of caregiver burden. Two groups of family caregivers were identified according to their trajectories in the level of burden: those with a persistent level of burden and those with an increasing burden over time. Family caregivers with a persistent level of burden seemed to be at risk for burnout throughout the illness trajectory of their care recipient. When an issue arose, they were often able to cope with the situation by making arrangements in the care or at work. However, they were pushing their limits and often felt exhausted.

Family caregivers with an increasing burden were unable to make sufficient adjustments to alleviate the pressure. For these caregivers, the combination of paid work and family care became too much, resulting in sick leave from work. This is in line with the ‘wear-and-tear’ perspective, which predicts an increase in the burden over time for family caregivers due to the cumulative effects of stressors that exhaust their mental and physical reserves. This might not be surprising, however, since providing care to a loved one nearing the end of life was often emotionally burdensome and intensive. In line with our findings, support for the wear-and-tear perspective has been found in partner caregivers and/or caregivers who live with patients [23, 24].

This study was among the first to describe trajectories of burden among family caregivers based on longitudinal qualitative data. Contrary to previous longitudinal quantitative studies [22, 25], this made it possible to identify events related to changes in burden over time, as well as to describe why these events were of importance. Moreover, opposed to prior research [25], this study did not only focus on the care situation, but also considered the impact of work on care and vice versa. Also, discrepancies between family caregivers’ perceptions during and after the care situation could be identified. The results indicated that the experiences of family caregivers did not only differ between caregivers, but individual experiences also differed over time.

In addition, the framework that has been identified based on the first round of interviews [7] was confirmed by the longitudinal data of the current study. Over time, experiences, feelings and needs regarding family care and the combination with paid work were influenced by four domains (caregiver characteristics, the care situation, the work situation, and the context). These experiences, feelings and needs sometimes had an impact on their health and well-being, or prompted caregivers to take action to improve the situation (e.g. changing jobs or arranging more help with care). Also, changes in health and well-being sometimes had an impact on the situation in multiple domains, for instance, when caregivers experienced an increasing burden and eventually went on sick leave from work.

Our findings correspond with a recent meta-analysis that showed a significant increase in emotional exhaustion when proving care for a relative, which was a risk factor for burnout [26]. Similarly, almost none of the family caregivers in our study experienced a decrease in the burden during the illness trajectory. The burden often only decreased substantially after the care situation had ended. Family caregivers emphasized that they felt sad but relieved after the death of the care recipient, and that there was space for normal things again. Our findings are consistent with other research that has shown that depressive symptoms in bereaved caregivers reduce significantly within three months after death, despite intensive caregiving situations [27]. Prior research has found that, although the burden increased towards the death of the care recipient, about three-quarters of the family caregivers did not perceive this burden as a problem. For these caregivers, providing family care did not feel like a problem because the loved one had previously also cared for them, and providing care felt rewarding [28]. This could possibly help with acceptance and finding closure after the loss of a loved one.

In line with other research [4], family caregivers shared the illness experience of the care recipient, and this often influenced their physical, psychological, social and spiritual well-being. Their needs were dynamic, and varied according to the trajectory [2, 4]. However, the caregiver burden trajectories did not directly reflect the three distinct disease trajectories (namely aggressive cancers, organ failure, frailty or dementia) that have been described by Murray and colleagues [2]. There was no clear difference between the caregiver groups with a persistent or increasing burden in the illness types of their care recipients. In both groups, caregiver burden was mostly related to aspects of the care situation, such as high care intensity, the patient’s increasing dependency, not being able to share care tasks, and acute care situations. Thus, although different illness types impose different requirements on the nature and intensity of care, the degree of dependency of the patient and care availability seemed to be more important in determining the burden and specific support needs. Similarly, fatigue and the level of dependency on the family caregiver have been found in prior research to be associated with caregiver burden and depressive symptoms [29, 30].

Demanding care situations made it difficult to combine care with paid work, especially in jobs with little understanding from colleagues/supervisors or few support options. This corresponds with other studies that have shown that family caregivers who combine work and care experienced the highest burden, and that higher patient dependency in combination with employment was related to caregiver stress [6, 21]. Among the groups of caregivers with a persistent or increasing burden, there were two cases of caregivers whose course of burden was slightly different compared to the rest in the group. Their cases showed that burnout symptoms and long-term sick leave from work could be prevented if there is understanding, flexibility, freedom to take time off from work without loss of salary or formal arrangements, or by making substantial changes in work (e.g. changing to a job with more flexibility and less demanding tasks). In agreement with a scoping review that showed that flexible work arrangements were essential for caregivers [14], our findings also suggest that work adjustments and support at work could make a difference in preventing burnout and sick leave among working family caregivers. Similarly, the results showed that caregivers who received support and had flexibility in their job also returned to work earlier after the care recipient’s death. Returning to a supportive work environment could help employees pick up their work again [31]. This emphasizes that giving family caregivers help in combining paid work with care tasks could benefit both the family caregiver and the employer in the long term.

Most of the family caregivers who persistently experienced a moderate burden had a job in professional healthcare. Having professional experience with caregiving and being familiar with the care sector might make family caregiving easier for these caregivers. This is in line with the earlier finding that self-efficacy for coping with the patient’s illness was negatively related to caregiver burden [22]. However, prior studies have also indicated that “double-duty caregivers” (i.e. professional caregivers with family caregiving tasks) are at risk of developing symptoms of overload and often experience mental and physical pressure [32, 33]. Concomitantly, double-duty caregivers often have high expectations for themselves and want to solve problems themselves because they know how to perform care tasks [32]. Flexibility and understanding from the workplace, as well as discussing the care situation with their supervisor, are important in identifying issues and allowing customized support to be offered [32].

The longitudinal design, in which 53 interviews were carried out with 17 participants over a time period of 2.5 years, provided rich and detailed information. Moreover, this made it possible to further clarify themes regarding the combination of paid work and family care at the end of life that were discussed in prior interviews. This approach also allowed differences to be identified in the views on how the combination of paid work and family care was arranged at various time points in the illness trajectory. Interviewing the same participants multiple times also builds trust between the participant and the interviewer, and this enhanced rapport might have made it easier to discuss difficult topics. Moreover, this helped us to interview family caregivers who were sometimes under extreme pressure or even on sick leave from work, and learn about their experiences. It could be that these people would not have participated in such research at that time if they had not already been familiar with it. In this way, we were able to tap into a wide range of experiences. At the same time, some of these caregivers emphasized that they valued being able to share their stories and be listened to.

In addition, analysing multiple interviews from the same participant as a single unit made it possible to create individual timelines of changes in burden, work and care. Although the course of burden over time was roughly the same for the caregivers within each group, experiences with caregiving were heterogeneous and changed over time [24]. Some family caregivers, however, did not have a particularly clear course of burden. Given that their experiences were dependent on various aspects of the care and work situation within a certain time period, we should refrain from ‘labelling’ individual family caregivers on the basis of the burden trajectories. In addition, not all family caregivers were in the same phase of the illness trajectory during the interviews. It has been found that the burden increases particularly as the patient’s illness progresses and in the terminal stage [28, 34]. Hence, it could be that for some of the caregivers who experienced persistent levels of burden during this study, the burden did eventually increase (or decrease) closer to the death of the care recipient. However, this would likely not have led to different trajectories at the group level, since all burden trajectories included family caregivers who had lost the care recipient during the study period. Future research could take the duration and phase of the illness trajectory into account when investigating burden trajectories of family caregivers of a patient at the end of life.

Family caregivers were recruited and included via general practitioners. This could be considered as a strong point, since the general practitioners had detailed information about patients with a life-threatening illness and were often in contact with the family caregivers as well. However, there is a possibility that general practitioners refrained from asking family caregivers in intensive care situations. Even so, we did not miss out on this group since some family caregivers in intensive care situations applied to take part in the study in response to the posters that were placed in a hospital. Furthermore, although we sought for maximum variation in the sample in terms of gender, working hours, sector, illness type and intensity of care, the views and experiences of family caregivers with certain characteristics might be underrepresented. This could for instance be the case for male caregivers or family caregivers with a non-Western cultural background. Prior research has indicated that male caregivers report less of a burden compared to female caregivers, and that gender differences in the caregiver’s burden increase over time among family caregivers who provide end-of-life care in the home setting [30]. Furthermore, earlier studies have found that cultural norms and values among family caregivers with a non-Western cultural background might also influence their caregiving experiences [35, 36]. Hence, it could be that these caregivers have different experiences with the combination of paid work and family care at the end of life, and, accordingly, have a different burden trajectory. Future research could give more emphasis to gender and cultural norms in distinguishing caregiver burden trajectories in the combination of paid work and family care at the end of life.

The family caregivers’ perceptions of the situation often changed over the course of the study. Some caregivers reported in earlier interviews that they had good arrangements for the combination of paid work and family care, while in later interviews they said that this was not the case (or not anymore) and it had been quite tough when they looked back at the situation at that point. This suggests that some family caregivers may find it difficult to recognize or identify unmet support needs at the time when the need is most urgent. It could also be the case that, for some caregivers, not acknowledging that the situation is too demanding might function as a coping mechanism allowing them to keep going. This could be problematic, especially if they refrain from seeking or accepting help and support while burnout is looming.

In agreement with earlier findings [4], key time points when family caregivers might have unmet support needs are the diagnosis, transitions in care (e.g. hospital admission, returning home after a hospital stay, and transition to a nursing home or hospice), when a disease-modifying or potentially curative treatment is stopped, in the terminal stage and after death. Healthcare professionals, such as nurses and home-care staff, could play an important role in supporting the family caregivers of a patient at the end of life at these times. They could use needs assessment tools to discuss potential unmet support needs. The Carer Needs Assessment Tool (CSNAT), for example, has proved to be an effective tool to do this [37, 38]. Moreover, given that the perceptions of family caregivers about how things were arranged often differed over time, discussing potential unmet support needs early on and at key time points in the illness trajectory might prevent a situation in which caregivers keep on going and providing care themselves, while there could be other support options available that they are unaware of or that did not cross their minds. In addition, the results showed that providing care to multiple persons contributed to burden. Given that in the Netherlands, about one in three family caregivers provide care to two or more persons and that this number is likely to further increase in the near future [39, 40], healthcare professionals could be more alert to unmet support needs among family caregivers with care tasks for multiple care recipients.

A supportive workplace appeared to be important in helping family caregivers combine care with work responsibilities. It is vitally important to discuss the theme of combining paid work with family caregiving by creating a safe environment for employees to share worries or issues that complicate the combination of paid work and family care for them. Supervisors could try to accommodate the needs of the caregiving employee by discussing possible solutions and offering tailored support, for instance in the form of flexible work hours, different tasks or working remotely. In addition, organizations should communicate policies regarding formal arrangements more actively, since not all caregivers were aware what support options were on offer. Social workers and occupational physicians could play an essential role in the provision of support for grieving employees and promoting awareness among supervisors [31].