Caring beyond hospital walls: Informal caregivers’ experiences in a virtual and in-person hybrid hospital-at-home

Abstract

Background

Hospital-at-Home (HaH) models deliver hospital-level care in patients’ homes, with some models relying on informal caregivers for support. While HaH might improve patient experience, excessive family involvement may lead to unintentional stress and unmet needs for both patients and their informal caregivers. Understanding user experiences is essential for designing person-centered HaH models. This study explored informal caregivers’ experiences with Influenzer, a hybrid HaH program combining telemedicine and in-person home visits.

Methods

We conducted a qualitative study using semi-structured interviews with 16 participants recruited during the feasibility phase of the Influenzer program at Copenhagen University Hospital – North Zealand in Denmark. Interviews were analyzed inductively using reflexive thematic analysis, informed by phenomenological principles.

Results

Two overarching themes were generated: (1) Being the primary informal caregiver supporting a home hospitalized patient was an experience of ambivalent feelings, (2) Informal caregivers were individuals with different relationships, roles, experiences, and needs – some of which were unmet. Many caregivers reported insufficient involvement in decision making and limited guidance on symptom observation and response.

Conclusions

The Influenzer program was generally preferred over hospital admission, yet informal caregivers experienced insecurity linked to unclear roles and responsibilities. Future HaH models should integrate early involvement, tailored information, and structured support to reduce caregiver burden.

Trial Registration

ClinicalTrials.gov NCT05087082, Date: 2021-10-08.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-026-14600-4.

Background

In 2023, a consensus definition of Hospital-in-Home (HaH) was proposed at World Hospital at Home Congress (WHAHC) as: “(…) an acute clinical service that takes staff, equipment, technologies, medication, and skills usually provided in hospitals and delivers that hospital care to selected people in their homes or in nursing homes. It substitutes for acute inpatient hospital care (.)” [1]. Since 2020, the development of various HaH models have accelerated, largely driven by the COVID-19 pandemic and the need to strengthen health systems’ resilience to future medical crises. This trend reflects efforts to address demographic changes and the persistent constraints on hospital capacity [2, 3].

As a care concept, HaH is constantly evolving and expanding globally. Models with a variability of services targeting different patient groups exist [2–4]. Yet only a few offer technology-enabled 24/7 remote patient monitoring (RPM), despite suggestions to promote technological innovations as part of HaH, to e.g., facilitate real-time data exchange and instant communication between the healthcare team and the patients [4]. As the HaH evolve towards virtual or hybrid (virtual and in-person) care pathways, the tasks for patients and their informal caregivers (ICs) will also change.

In 2019, at the first WHAHC congress, patient and caregiver experiences were identified as one of the most important research topics [5]. According to the Cochrane review by Edgar and colleagues, ICs are expected to play a significant role in HaH programs [6]. This would arguably be the case when RPM is used with additional tasks such as patient self-measurement. The importance of ICs for HaH has been supported across different cultural settings. Within an Asian context, caregivers acted as positive enablers, providing essential physical and emotional support for HaH patients with positive outcomes for both patients and ICs, including improved health, ability to maintain normal lives, and improved quality of life [4]. In a Scandinavian context, ICs were important, although their tasks, level of involvement, and efforts varied across different phases of HaH treatment [7]. A British study highlighted the importance of “patients-ICs collaboration” to adapt to HaH [8].

Although ICs generally have a positive attitude towards HaH and play a crucial role by providing support for the patients, their deep involvement may lead to stress [9–11]. This duality emphasizes the need for a better understanding of ICs’ experiences and needs to provide targeted support and avoid unintended harm to the ICs. While some evidence on ICs’ experiences exists, there is a gap in research concerning the role of ICs during a technology-enabled HaH [6, 12]. This paper addresses this gap.

Our research group was responsible for a large research and innovation project Influenzer, that aimed to develop, evaluate, and implement a hybrid HaH. This initiative delivered hospital-led RPM and in-person treatment for patients with acute illness requiring hospitalisation in their homes [13]. The project included three main phases: (1) the Development of a HaH program prototype; (2) a small-scale clinical trial to stress its feasibility; and (3) the final implementation of the revised program to test its effectiveness in a randomised controlled trial (RCT) [14]. A formative assessment conducted during the feasibility phase included interviews with the participants (HaH patients) and their ICs to identify any acceptability issues and potential implementation barriers. No formal analysis was performed at that time. However, the insights were used to inform immediate improvements and to plan the subsequent RCT. This study aims to explore how ICs experienced their role in supporting HaH patients through an in-depth analysis of the ICs’ interviews.

Methods

Aim, design and setting

This study explored the experiences of ICs participating in a hybrid HaH program, Influenzer. It was a part of a clinical feasibility trial conducted between April 2022 and May 2023 at the Department of Pulmonary and Infectious Diseases (DPID) at Copenhagen University Hospital - North Zealand, in the Capital Region of Denmark [14]. It was designed as a qualitative study comprising semi-structured interviews and guided by an inductive analytic approach. The analysis was inspired by phenomenological principles, as the aim was to explore ICs' lived experience of their role in supporting HaH patients [15].

The Influenzer program

The Influenzer program was an early supported discharge hybrid HaH [14]. The full details of the program and the settings are provided in an additional file [Additional file 1].

The program was offered to patients admitted to DPID who were clinically stable adults diagnosed with lower respiratory tract infections (LRTI), had low care dependency (e.g., were able to manage personal hygiene independently), and were able to communicate effectively in Danish. If patients accepted, they were transferred to their homes to continue the hospitalization (HaH patients). HaH patients monitored their vital signs (e.g., oxygen saturation) and entered the data, supported by a symptom checklist, into a telemedicine app. RPM was required at least twice daily and was tailored to individual risk profiles. An alert system notified hospital staff in case of abnormal readings or new symptoms, prompting follow-up via telephone or video. Medical oversight was maintained through scheduled video consultations. A mobile team provided in-home services such as blood sampling and intravenous therapy. In the event of clinical deterioration, patients were transferred back to the Brick and Mortar (BAM) for further evaluation.

There were no requirements to have an IC, and no program tasks were assigned to ICs. However, cohabiting ICs could request that the HaH patient be transferred back to the BAM for any reason (e.g., if feeling overwhelmed or unsafe).

Participants inclusion criteria and recruitment

Immediately upon their recruitment in the feasibility trial, HaH patients were offered to designate a family member as primary IC for this study. ICs were included in this study if they were 18 years or older and provided informed consent. They were unpaid, typically without formal medical/nursing training, and might have assisted HaH patients with practical tasks and care before and/or during the HaH program.

ICs were recruited through consecutive sampling. Immediately after being appointed, the ICs were approached face-to-face if they were present at the hospital or by phone. ICs were informed and recruited and written informed consent was obtained between April 2022 and May 2023 by the clinical project team that was responsible for feasibility trial procedures.

Data collection

Individual telephone interviews were conducted within 14 days of HaH patient discharge. All interviews were conducted by author TSD via telephone from the Department of Clinical Research at Copenhagen University Hospital – North Zealand. No patients were present during the interviews. Prior to each interview, ICs were advised to be alone to ensure that they could speak freely without the influence of the patient. They lasted approximately 30 min, with four exceptions lasting 10, 13, 16, and 50 min respectively. No repeat interviews were conducted. The ICs were aware of TSD’s professional background as a medical doctor and her role in the feasibility trial. The interview guide was developed by TSD and informed by prior experiences and a rapid literature review. It included open-ended questions about participants’ overall experiences and perspectives, supplemented with questions related to tasks during HaH, feelings of security, and improvements for future HaH models. The full interview guide is available in the additional file [Additional file 2]. The interviews were audio-recorded and transcribed verbatim by a project nurse and a public health consultant.

Data saturation

Data saturation in qualitative research is a context-dependent and subjective process that requires detailed, systematic analysis [16]. In the present study, data saturation was pragmatically determined within the constraints of the feasibility trial design, which included 19 HaH patients. As such, the saturation point was assessed in relation to the predefined sample size. Although saturation was determined pragmatically, we argue that the study demonstrated substantial information power [16]. While the interviewer had limited experience with qualitative research methods, her strong communicative skills and ability to establish rapport supported the depth and quality of the interviews. Moreover, although the analysis was conducted as a cross-case exploration of multiple IC experiences, the study’s narrowly defined aim - to explore IC’s lived experience of their role in supporting HaH patients, specifically within the feasibility trial of the Influenzer program - ensured dense specificity. This focus, combined with a phenomenological theoretical positioning and analytical approach, provided a coherent and rigorous framework for generating meaningful insights.

Analysis

The analysis was conducted by authors KU and MLO using an inductive analytic approach. Interviews were analyzed following Braun and Clarke’s framework for reflexive thematic analysis (RTA) [17]. Although the authors followed the six phases of RTA, the process was recursive and iterative rather than strictly linear, involving continuous movement between phases [18]: In phase one, KU and MLO familiarized themselves with the data by reading each interview transcript and noting preliminary ideas. In phase two, the initial coding was then carried out by KU in collaboration with MLO and SGN. Comprehensive spreadsheets (using Microsoft Word and Excel) were generated where codes were organized using color-coding and corresponding names for easy identification and categorization. Through ongoing discussions, the codes were refined by examining their interrelations, similarities, and differences. In phase three, based on these discussions, relevant codes were grouped into preliminary themes and subthemes. In phase four, the themes and subthemes were reviewed against the data to ensure validity. This involved revisiting the coded data extracts, during which some overlap between themes was identified. As a result, two themes were merged. During this refinement process, the authors also identified and defined the two main themes to highlight distinct nuances in the data: One capturing the experience of the ICs, and another reflecting the needs that were associated with their experiences. The process of generating theme 1 is illustrated in anadditional file [Additional file 3]. Participants did not provide feedback on the findings. Finally, KU and MLO discussed whether the themes adequately captured the intended essence and narrative of the data. After considering potential revisions, the authors proceeded to the final phase – writing the article.

Researcher reflexivity and positionality

The research team was interdisciplinary, bringing together anthropological, medical, and nursing perspectives. This composition contributed to diversity in expertise and introduced multiple lenses for data collection and analysis. The team also spanned a wide age range, from early-career to senior researchers. Some team members had prior insight into the development of HaH services, the feasibility study, or other exploratory investigations of participant experiences, while others had experience with home-based care in different clinical settings. Collectively, these varied backgrounds enriched the study and supported a nuanced and multifaceted interpretation of the findings.

A medical doctor conducted all interviews and led the feasibility study, and no other researcher had prior relationships with participants. The research group contributed in various ways throughout the different stages of the study, bringing together complementary competencies. Our study is theoretically positioned within a phenomenological epistemological framework. This approach supported the analysis by acknowledging lived experience as a factor that actively shapes the treatment context [19]. It therefore rests on the premise that the lived experience of ICs is both valid and necessary for understanding the phenomenon under study.

Results

Participant characteristics

In total, 16 ICs were interviewed out of 19 possible. Three HaH patients could not appoint any IC, while one patient appointed two ICs, who were both included. A single IC did not provide written consent. Most (12 out of 16) ICs were partners residing with the patient. One (IC3 - daughter) moved into the patient’s home during HaH, while three (IC1 – daughter, IC6 – mother, and IC7 – daughter) lived apart. Full IC characteristics are provided in an additional file [Additional file 4].

Themes

Two themes covering five subthemes were generated to describe the experiences of the ICs (Table 1).

Table 1.

Overview of themes and subthemes describing informal caregiver experience with Hospital-at-Home (HaH)

Theme 1: Being the primary informal caregiver supporting a home hospitalized patient was an experience of ambivalent feelings	Theme 2: Informal caregivers were individuals with different relationships, roles, experiences, and needs – some of which were unmet
Subtheme 1.1: The home as a familiar and joy-filled place	Subtheme 2.1: Relational dynamics affected understanding, support and the overall HaH experiences
Subtheme 1.2: Insecurity, new tasks, and changed roles	Subtheme 2.2: Individual resources and life circumstances influenced caregiver support
	Subtheme 2.3: Caregiver needs across HaH phases

Theme 1: Being the primary informal caregiver supporting a home hospitalized patient was an experience of ambivalent feelings

HaH often presented an ambivalent experience for ICs, regardless of their role as parent, child, or partner. While HaH was valuable and brought joy and relief, it simultaneously induced feelings of insecurity. One IC, the daughter of a patient, described these mixed feelings during her father’s HaH:

The feelings you have are very paradoxical. There was a huge sense of relief in having him home, because I knew how it affected him (…) But there was definitely [also] an insecurity about this and whether I, or we [referring to her mother], took on a responsibility, for his sake, he actually wasn’t fit for yet. (IC1 – daughter).

Subtheme 1.1: The home as a familiar and joy-filled place

The ICs described HaH as joy-filled and with an immediate relief, which contrasted to the BAM hospitalization. HaH enabled spending more time together, which was appreciated by the patient, the IC, and the extended family. One patient’s wife elaborated:

The kids, the grandkids, and their dog were able to visit us here in our home. And the kids could play (…) instead of them having to come to the hospital. It can seem a bit scary for the kids, you know, to see their grandfather lying there [in the hospital bed]. (IC16 – wife).

The husband of another patient described:

(…) it was pleasant having her at home instead of us being apart in those 14 days” (IC14 – husband).

Although these aspects might seem modest, the ICs found them invaluable. They appreciated simple pleasures such as seeing the patient sit outside in the sun, laughing together, or serving meals tailored to the patient's preferences.

One daughter valued the ability to closely monitor her mother’s health condition. She moved into her mother’s home, to better be able to support her:

I think it has been pleasant to have her home because it enabled me to see and hear how she was. You cannot do that in the hospital the same way. During the hour there, you just get a snapshot. But here at home, I have been able to follow her progress all the time. (IC3 – daughter).

Some ICs experienced faster patient recovery at home:

I would almost say that [the patient] was really well, more well [at home] than when she was up there [at the hospital] (…). She was a bit livelier. (IC2 – husband).

Another IC noted that HaH was less stressful, as it reduced the need for transportation and enabled a continuous emotional support:

It has been fantastic! Actually, much less stressful, because I did not have to drive up to the hospital twice a day and sit with him because he was so upset and feeling so unwell. Yeah so, for me it has been a great advantage to see how much faster he started to recover [at home]. (IC11 – wife).

Subtheme 1.2: Insecurity, new tasks, and changed roles

Alongside feelings of joy, immediate relief, and the familiarity of the home context, the ICs also described a sense of insecurity related to the HaH. This insecurity was primarily associated with their perceived shift in roles and responsibilities during HaH. Many ICs described how they performed several practical tasks to support the patient. These tasks included both familiar responsibilities performed to a greater extend, and new practical tasks. Familiar responsibilities for which ICs had the sole responsibility included cooking and grocery shopping. New practical tasks could involve reorganizing the physical space at home to accommodate HaH:

Usually we sleep in the basement, but I had to move our bed upstairs in the kids’ room (…). Then he [her husband] doesn’t have to walk up the stairs. (…). It is a bit weird to sleep three meters from each other, but its fine for now [laughs]. (IC16 – wife).

New practical tasks, which temporarily altered ICs’ roles in their relationship with the patient, also occurred. These included helping their relative taking a shower, even though the healthcare professionals had assessed the patients to be able to perform this activity independently, or supporting eating and drinking:

He thought he wanted orange juice, so I bought orange juice, and then he didn’t feel like drinking orange juice after all. Then he wanted apple juice, so I bought apple juice… but then he did not want apple juice… It was like that all the time. And yes, I could have said ‘well if nothing works for you then you have to drink what we have’. But he had lost so much weight, you know, so you do everything you can just to give him some energy so he can start feeling better. (IC5 – wife).

Many ICs experienced a change in their roles in relation to the patient, transforming into helpers performing various practical tasks at home, that they had not done prior to HaH. This temporary shift also involved an increased sense of responsibility. As the closest and most present IC, they felt responsible for the patients. This sense of responsibility stemmed from an immediate human concern and devotion to the patient’s well-being, rather than instructions from the healthcare professionals. In other words, an unspoken inclination to take on tasks and provide support when a family member became ill. It led to heightened vigilance and manifested as a new practical task: Continuous observation of the patient to identify potential symptoms of disease progression. Paradoxically, while ICs reported that the patients slept better at home, several of them felt like they were on duty day and night. One described:

We slept with open doors between our rooms so that I would be able to hear him if there was even the slightest thing, right? (IC11 – wife).

ICs experienced that they would be the first to notice, and therefore additionally act, if anything unexpected happened, making them more vigilant and uneasy. They worried about not being able to respond adequately in case of acute deterioration in the patient’s health. While HaH minimized stress for the ICs living far from the BAM, eliminating the need for travel, distance could also cause stress and insecurity, due to being far from help in a potential emergency. This was especially important if the ICs assessed the patient to be in an unstable condition. Vigilance and a sense of responsibility were reflected in the following quotes by IC1 and IC8, who were related to the same patient known with cancer, home hospitalized for 48 h. IC1 explained:

(…) You have a responsibility for something I do not really know anything about. I would not know what to do if suddenly he passed out or something. The insecurity that it entails, would have been less if he lived closer to the hospital, but he lives on the countryside, so when we came home and I saw how ill he was, that was something that really fuelled this insecurity. (IC1 - daughter).

Meanwhile, IC8 (IC1’s mother) experienced stress and lack of sleep:

I would have shouted for help if it had been like that for a week. I did not sleep at all. I was afraid to (…). The way he was breathing and… he was very tired right? (…). And then of course it [technology from the hospital] was new, and it [his measurements in the home] had to be done again four hours later, so before my adrenalin went down again, the phone rang, even though it shouldn’t [the hospital staff should only ring in case of out-of-range self-measurements]. You see… I had adrenaline in my body all the time, right? (IC8 - wife).

Notably, some ICs reported initial worries regarding their roles and responsibilities during HaH. However, upon retrospective reflection of their experience, their worries were mostly unfounded as they had felt well-supported by the hospital. As one noted:

My initial thought was: What type of responsibility will I have? Because I had seen her when she was really ill. But I was ascertained by your setup, and I truly think it worked well. You [the hospital staff] called her [his wife] within two minutes of her reaching out about her dyspnea, to assist her. So, I have had no problems. (IC9 – husband).

Overall, most ICs (11 out of 16) expressed a preference for HaH over BAM, due to the joyful aspects of being in the home, if they experienced the patient’s wellbeing as not too severe to manage.

Theme 2: Informal caregivers were individuals with different relationships, roles, experiences, and needs – some of which were unmet

Each IC in this study had distinct characteristics, life circumstances, resources, and relationships with the patients. Their experiences and needs during HaH were influenced by these factors, along with their perceptions of the severity and stability of the patient’s condition.

Subtheme 2.1: Relational dynamics affected understanding, support and the overall HaH experiences

One relational factor affecting the HaH process was the degrees of mutual understanding and emotional support in the IC-patient relationship. As described in Theme 1, many ICs experienced a temporary change in their roles in relations to the patient. However, ICs married to or partnered with the patient often viewed HaH as teamwork process, characterized by an intuitive unspoken understanding. These ICs described how they did most things together in their daily lives, and this continued during HaH. For example, IC14 explained how he helped his wife put on the heart rate monitor and gather items for her measurements to send to the BAM. He also expressed how taking care of each other had become a habit after many years together:

If, at night for instance, we don’t hear the other one snoring, we might just take the other’s hand. (IC14 – husband).

These relationships were characterized by a high degree of mutual understanding and support, which was then naturally carried into the HaH, and influencing the process. As IC14 expressed:

It is simply a habit after so many years, to look out for one another, so it didn’t require us to adjust.

An exception was noted in the case of IC12, where strained dynamics between the IC and patient during HaH, lead to conflict. IC12 described how her husband became “unruly”. A conflict occurred between the couple, where the patient, had the experience of his wife making mistakes when handling his oral medications – a task she had taken full responsibility for during HaH. Being accused of this, caused her frustration and led to her withdrawal from this task. This prompted her husband to request re-transfer to the BAM, as he wanted support for managing his medications. It remains unclear whether the strained dynamics in their relationship existed prior to or were a result of HaH. However, it was evident that this relational tension influenced mutual understanding and support, thereby affecting the HaH process.

Although family relationships varied in character and quality, a consistent pattern in the data was the ICs' sense of responsibility. This stemmed from an unspoken inclination to take over tasks and provide support when a family member became ill, as described in Theme 1.

Subtheme 2.2: Individual resources and life circumstances influenced caregiver support

Support and collaboration during HaH was shaped by the ICs’ living conditions, work obligations, and daily life routines. Additionally, individual resources – such as technological competencies, available time, and personal capacity – played a significant role in their ability to provide support.

Influence of occupational status

Many ICs were elderly and retired, providing them more time to support the patient physically and emotionally compared to those who were still working. However, even among retired ICs, some adjusted their daily routines to accommodate the needs of the patients during HaH. As one IC described:

I play golf on Tuesdays, and I don’t remember if I went down there [to play]. (…). It was in my interest to make this [HaH] work. I think I just told myself: ‘Okay well, right now this is how it is, so this is what we do’. No, I stayed home Tuesday. (IC4 – girlfriend).

Some ICs reflected on the potential burden of working during HaH, which was closely linked to the level of support they felt their relative required. IC5, who was retired, shared the following reflection:

If I had been working, then maybe… he would have had to do a lot of things by himself. And I’m unsure how well that would have worked out. (…). He didn’t have the energy to cook. He barely had the energy to sit up and eat. Stuff like that. (IC5 – wife).

However, for ICs who were employed, being absent during parts of the day was not necessarily perceived as problematic if they felt the patient could manage HaH independently and had sufficient professional support:

She wasn’t dying or anything so I was very calm, and I knew that she had a number she could call if anything came up. I was at work, so she was home alone. (…). It was fine for me because she had that number. She wasn’t left alone in that sense. (IC13 – husband).

Influence of living circumstances

ICs’ living circumstances varied significantly. In one IC-patient relationship, the living arrangement was temporarily altered during HaH, while in other cases, ICs and patients continued living apart. However, most ICs and patients lived together, making the ICs more available for the patient and more knowledgeable about HaH.

Technological experience

ICs had varied experiences and understandings of the technological equipment used during HaH. While some found the devices easy to use and felt adequately instructed, others encountered difficulties. IC13 explained:

I don’t think the app worked very well. It asked for something…what is it called… I don’t know if it is blood pressure or whatever it is, but it says something like YX, something… we don’t understand that! Like what kind of reading are we supposed to do on that oxygen thing? There are several different numbers, so which one are we supposed to type? It’s not clear enough [in the instructions]. (IC13 – husband).

In contrast, another IC stated:

“It was fine. No problem at all. It was well thought out, and well executed and it was easy and simple and there were no problems at all”. (IC8 – wife).

Subtheme 2.3: Caregiver needs across HaH phases

Needs in the initial phase – involvement and communication

ICs expressed a need for greater involvement before the start of HaH. Many described the process as having ‘come out of the blue’ with little or no preparation. Some ICs noted that even a short period for planning – such as adjusting remote work schedules – would have been beneficial. A key factor in promoting a safe and coherent HaH experience was early and clear communication between ICs, who were most often at home, and BAM healthcare professionals. This collaboration was especially important from the beginning of the initial phase of HaH. In one case involving IC8 and IC1, both related to the same patient, IC1 was registered as the primary IC, while IC8 was the one living with the patient and providing most of the day-to-day support during HaH. In this case, the BAM-home communication had skipped IC8, who was then left out of the decision making, training and planning of HaH. This case highlights the importance of considering and questioning whether the patient’s initial designation of primary IC is also the most meaningful or relevant choice. IC8 felt insecure and worried during HaH due to a lack of communication and involvement from the beginning of her husband’s HaH, this combined with her perception of her husband’s illness being severe. Additionally, some insecurity may have stemmed from a lack of communication between mother and daughter, highlighting the importance of an awareness that such communication between relationally close ICs, is not given.

Needs during HaH – guidance on observation and actions

The degree of insecurity and worry experienced by ICs, as described in Theme 1, was related to how confident they felt in observing the patient’s symptoms. ICs had varying prerequisites for this task. Some had a background in healthcare, which provided them with useful knowledge and understanding - but also heightened their concerns due to increased awareness of potential complications. As IC10 described:

I don’t know a lot about legionella [LRTI] so I could have used more knowledge on this, because it also affects the liver and kidney. And due to my education [as nurse], I might be thinking more about that than a regular informal caregiver, right? I was thinking: ‘When would I have to react on something? (IC10 - wife).

Other ICs, without a medical background, also expressed insecurity about identifying symptoms and acting accordingly. Most ICs described spending time and energy interpreting signals from the patients, as IC9 illustrated:

She has started walking up and down the stairs and doing gymnastics. These things comfort me (…) whereas, if she had deteriorated, then I would also feel more uncomfortable about having to handle this myself. (IC9 - husband).

ICs described a vigilant attention to the patients’ wellbeing, often relying on all their senses. One IC focused on the patient’s skin colour, another on whether they could feel a change in the patient’s muscle size, and a third IC paid close attention to the sound of the patient’s voice and what they said, interpreting jokes as a sign of improvement. IC3 explained:

I am the one seeing her all the time, so I will be the one who notices if she suddenly feels unwell. Or if she suddenly changes skin colour or if she says: ‘now I’m hurting somewhere,’ then I’m the first witness and first caregiver to notice right? (IC3 – daughter).

When asked if she felt more worried than when the patient was hospitalized at the BAM, one IC stated:

Well, I guess I have been a bit more alert in terms of trying to read him, right? (IC11- wife).

Another IC expressed a need for more information on what to look for and how to act on her observations during the HaH:

I could have used some more guidance; Should I make sure that he drinks more? Should I keep an eye on him like that? Was there something I should make sure that he got to eat? I know that during such a period of illness your appetite is low. But I can also see now that he has lost quite some weight. (IC4 – girlfriend).

Needs during discharge

Some ICs expressed uncertainty about the point at which the patient was officially discharged from HaH, and no longer affiliated with the hospital - particularly when further assistance was needed, one IC Stated:

She didn’t realize that she had been discharged. Well… they had talked about the possibility of her being discharged if everything went well, but then they must say: ’Now we will push that button’. (IC15 – boyfriend).

Discussion

This study adds a nuanced understanding of IC experiences in technology enabled HaH models. Our main findings indicate that hybrid HaH was an ambivalent experience for ICs. On one hand, it felt familiar and joy-filled; on the other hand, it felt insecure. ICs felt a strong sense of responsibility for the patient, as they were likely to be physically present and the closest to them in case of an emergency. This led to heightened vigilance regarding the patient’s well-being. Despite these concerns, most ICs preferred HaH over traditional BAM hospitalization. We also found that ICs had varying competencies across multiple domains, and the quality of their relationship with patients differed, resulting in diverse challenges and needs. Some of these needs were not adequately met, particularly those related to involvement in decision making and planning for the HaH trajectory, as well as guidance on identifying potential signs of disease progression and responding appropriately to their observations.

Our findings on unmet needs amongst ICs related to involvement and guidance are in line with other literature addressing IC experiences. Sanders find that IC burden is common in the care of patients with dementia due to a lack of specific training or resources to manage the continuous and often exhausting responsibilities involved [20]. The coexistence of both positive and negative feelings is also documented in other HaH studies [7, 10, 21]. Especially, the first days of HaH were reported as stressful for ICs experiencing a total of 14–49 HaH days [7]. Our findings did not support this, which might be due to the shorter duration of HaH, with a median length of 2 HaH days [13]. We found that ICs did not feel sufficiently involved in or informed about the decision regarding HaH and how it was going to proceed. This is supported by other studies reporting e.g., that ICs were excluded from the decision-making process of choosing HaH over BAM [7] or lacked precise information about the HaH practicalities, resulting in some ICs feeling misled [10] or even experienced HaH as emerging ‘out of the blue’ with no involvement in the comprehensive assessment preceding the decision [9]. However, improved preparations and information to ICs could reduce these frustrations [10]. These findings highlight the importance of involving relatives as early in the process as possible and ensuring that HaH is based on a common informed basis, while securing patient autonomy and consent.

The reason ICs found it important to be involved may be caused by a sense of responsibility during HaH. This was noticed in a recent study by Bertelsen & Loft [21], which explored Danish IC experiences with acute home-based care, very like this study, however, without 24/7 RPM and the associated tasks. The authors found that while ICs did not perceive themselves as having a greater responsibility, showing trust in the health care professionals’ decisions, they felt, however, responsible for notifying the professionals if they noticed a change in the patient’s condition. In this way, they bore a responsibility, which they could not handle if it required professional health competencies [21]. Ko and colleagues also reported that within a Singaporean setting, ICs felt responsible for supervising the patient, despite lacking medical training [11]. This aligns with our findings, where ICs were vigilant and observed the patients – a challenging effort, as they felt insecure about how to interpret or act on signs. This was previously described in a study examining the needs of ICs during care transitions between BAM and home during the discharge [22]. Although the setting were not within a HaH program, we believe that the aspect of BAM-to-home-transition is an inherent part of HaH and makes it a valid comparison in terms of ICs’ needs and worries during and after this process.

Supporting Ko and colleagues’ findings [11] we found that ICs took on a new role as practical and emotional helpers. The ICs were uncertain about the meaning of this role and the expectations and responsibilities it entailed. We argue that more attention to the role and experiences of ICs during HaH is important - not only for the ICs themselves, but also for the patients and healthcare professionals. Better definitions of the IC role can be valuable in addressing needs and improving the sense of security.

To further understand this argument, we refer to anthropologist Roger Zetter’s work on labels as conceptual metaphors, and how they are defined differently depending on the interpretations of those ascribing and obtaining them [23]. Understanding the term Informal Caregiver as a label, we can see it as something assigned to or obtained by an IC with a particular meaning. Therefore, different expectations for ICs coexist depending on who interprets the label – the IC, the patient, or the healthcare professionals. The perceptions others might have of what it means to be an IC do not necessarily align with the responsibilities that ICs take on themselves or wish to have during HaH. We argue that healthcare professionals can better support ICs and address their concerns if they set their subjective assumptions aside and instead focus on how ICs interpret their role and expectations. Aligning expectations through early communication in the HaH process may alleviate ICs’ insecurity.

Retrospectively, it could have been valuable to explore the multidimensionality of caregiver appraisal through the four key categories described in a recent concept analysis by Wang and colleagues [24]: How caregivers appraise their own situation, how they appraise the care receiver’s situation, how they assess their own abilities to manage caregiving responsibilities and how they evaluate their relationship with the care receiver. The four dimensions could have informed the interview guide and provided a multidimensional view on caregiving appraisal from the outset. This multidimensionality is supported by our findings, which resonate with several of the four categories identified in the concept analysis.

In connection with our findings on ICs’ sense of responsibility, many expressed that supporting the patient was regarded as self-evident. This echoed previous findings describing HaH as a perceived obligation with no alternative to it [10], or how ICs often set aside their own needs to support the patient’s wish for home care, suggesting that caregiving is a natural, ethical response to vulnerability [21]. Drawing on Løgstrup, one study argued that this reflects deep relationships built on trust, where caregiving emerges spontaneously as part of a fundamental human interdependence [21]. Similarly, we argue that ICs experienced a sense of responsibility, which can be understood through their connection to the patient. Drawing on Emmanuel Levinas’ theory of ethical phenomenology, we assert that human decisions and actions as individuals are influenced by our interactions and that we feel morally obliged to others [25, 26]. Thus, ICs’ choices are influenced by their relational connection to the patient, making it challenging to disconnect from the HaH. Adding to this argument, we suggest a nuance centering around the IC’s perception of themselves. Following Levinas, we argue that ICs may feel responsible and vigilant not because of patient expectations, but because they are confronted with the moral consideration of who they aspire to be as ICs, and how they can embody this role in a way that aligns with their personal values. We argue that it can be beneficial for the healthcare professionals to be aware of the complexity of saying no to HaH or expressing individual doubt about one’s ability to be a primary IC. For the ICs, it is not just a question of individual capacity of resources, but also a question of being a good person in relation to the patient and their situation. This complex interplay might lead to a willingness to exceed personal limits to offer support, unintentionally introducing burden and distress. This might be prevented if HaH healthcare professionals engage ICs early in the process, including clear communication, uncovering ICs’ individual expectations and interpretations of the appointed Informal Caregiver role.

Additionally, in the present study 10/16 ICs were women [Additional file 4]. While the question of gender in informal caregiving has not been the focus of the present study, another study found that women were less likely than men to be eligible for HaH, as fewer female patients had male ICs available to provide the necessary care [27].

Future implications

This study shows that ICs have various individual needs and expectations. This emphasizes the need for development and testing of a HaH-IC assessment tool, which can shed light on these implicit perceptions. This would be the first step towards offering necessary and tailored support. The World Health Organization advocates for health systems’ support centered on people rather than diseases, through Person-Centered Care (PCC) [28]. PCC places individuals at the center, considering their individual context, history, family, and resources. Individuals, families, and communities are served by - and can participate in - healthcare systems that meet their needs in holistic ways [29]. While the Influenzer program followed a patient-centered approach, applying PCC as a framework for future HaH design could strengthen the programs, as all relevant persons, including ICs, would be at the center. A family-centered care (FCC) approach could also be justified, as it conceptualizes the family as a unit of care [30]. However, a PCC approach can also accommodate close family involvement when this aligns with the individual’s preferences. In the absence of a validated HaH-IC assessment tool, we suggest using other existing methods to prepare the HaH patients and their ICs for a clarifying dialogue with health care professionals. The patient and IC could reflect on open-ended questions such as: “If I were to be hospitalized at home, my first thought would be…”, inspired by Guided Self-determination [31]. In this way, the HaH dialogue would begin early and hopefully clarify the expectations, concerns, and needs of both the patient and their IC. Such an approach would include ICs on their own terms and help align expectations of their role.

Finally, it is important to communicate clearly the RPM workflows, including the hospital’s responsibilities and workflows in case of signs of clinical deterioration. However, since insecurity may still arise in the ICs, an action card explaining how patients and ICs should respond to specific symptoms during HaH might provide more reassurance. As the suggested methods are not tested as part of this study, we strongly recommend a pilot-testing in the relevant HaH context before implementation.

Strengths and limitations

This study adheres to the COREQ recommendations, ensuring transparency and rigorous description [Additional file 5]. The analysis is illustrated with a code tree, enhancing clarity and replicability. While a code tree might not be in alignment with conducting RTA as this is a subjective, reflexive and more fluid process than able to be illustrated through the code tree, we have chosen to add it to facilitate transparency in line with the COREQ criteria. It was conducted by a multidisciplinary team who engaged in team coding discussions thereby reducing the influence of a single author’s positioning and preconceptions. Participants were interviewed within 14 days after HaH patients’ discharge, minimizing recall bias. However, a potential risk of desirability- and power bias remains, as participants were familiar with the interviewer - a clinician involved in the feasibility trial. Providing reflections to an authority figure associated with the project may have influenced participants to offer less critical accounts, potentially portraying their experiences in a more favorable light than they actually perceived them. Acknowledging this potential power imbalance, we took deliberate steps to minimize its influence. Specifically, the interviewer did not occupy the role of attending physician during the patients’ HaH, and all interviews were conducted only after the completion of HaH.

While interviews conducted within 14 days of discharge only capture early experiences rather than longitudinal caregiving trajectories, this was deemed appropriate. Given the acute nature of the HaH and its median duration of two days, extended caregiving trajectories were not expected.

Transferability is inherently limited by the specific characteristics of the study context, including the feasibility study setting, the participants’ self-selection into an experimental programme, the focus on a single disease trajectory, and the Danish healthcare context. Nonetheless, the detailed descriptions of participants’ experiences may allow readers to assess the relevance of the findings to other settings organisational structure or patient populations.

Conclusion

This study highlights the dual experience of joy and insecurity among informal caregivers whose loved ones participated in a hybrid virtual and in-person HaH program. Caregivers expressed a range of implicit needs and expectations, many of which remained unmet, contributing to varying levels of strain. Notably, early and meaningful involvement emerged as a key priority. Future HaH models should incorporate structured caregiver engagement and tailored support strategies to mitigate caregiver burden.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

BAM

Brick and Mortar

DPID

Department of Pulmonary and Infectious Diseases

HaH

Hospital-at-Home

IC

Informal Caregiver

LRTI

Lower Respiratory Tract Infection

PCC

Person-centered Care

FCC

Family-centered Care

RPM

Remote Patient Monitoring

RTA

Reflexive Thematic Analysis

Author contributions

Karoline Uldbjerg (KU) has contributed to methodology; literature search; validation, data curation; formal analysis and data interpretation; validation; visualization; writing - original draft and review, and editing. Signe Hellung Schønning (SHS) has contributed to methodology; formal analysis and data interpretation; literature search; validation; writing – review and editing. Tatjana Sandreva Dreisig (TSD) has contributed to conceptualizing the intervention; methodology; investigation; project administration, including management of the clinical feasibility trial (wherein this work was embedded); literature search; data collection; writing - review and editing. Susanne Grøn Nielsen (SGN) has contributed to formal analysis and data interpretation; investigation; writing - review and editing. Maria Normand Larsen (MNL) has contributed to investigation, including participant recruitment, writing – review & editing. Thyge Lynghøj Nielsen (TLN) has contributed to conceptualizing the intervention and the overall research project Influenzer; methodology; project administration; review and editing. Charlotte Demuth von Sydow (CDvS) has contributed to conceptualizing the intervention and the overall research project Influenzer; funding acquisition; project administration; writing - review and editing. Thea Kølsen Fischer (TKF) has contributed to conceptualizing the intervention and the overall research project Influenzer; project Influenzer administration; the main supervision of the clinical feasibility trial (wherein this work was embedded); funding acquisition; writing - review and editing. TKF is the guarantor. Mette Linnet Olesen (MLO) has contributed to literature search; formal analysis and data interpretation; validation; writing - original draft and review, and editing.

Funding

Open access funding provided by Copenhagen University. The Influenzer project is partly funded by Innovation Fund Denmark (grant number 0176-00017B). The funding organization was not involved in the study design, data collection, data analysis, data interpretation, or manuscript writing. The research was conducted without any influence from the funder.

Data availability

The datasets generated and analyzed during the current study will not be made publicly available to preserve the privacy of the study participants under the European General Data Protection Regulation. The datasets are available on request for academic purposes by contacting authors CDvS and TKF.

Declarations

Ethics approval and consent to participate

All ICs received verbal and written information about the project and were given time to consider their participation before providing written informed consent and the ethical principles of the Declaration of Helsinki were followed [32]. The feasibility study of the Influenzer program, was registered with the Danish Data Protection Agency (P 2021 − 862) and has ethical approval (98381) from the Regional Committee of Human Health Research Ethics, Capital Region of Denmark.

Consent for publication

Not applicable.

Use of AI

To improve the clarity and readability of the manuscript, a Large Language Model (Co-pilot) was used to assist with language revision. As none of the authors are native English speakers, this support helped ensure that the phrasing and sentence structure met academic standards. All revisions were carefully reviewed and verified by the authors.

Competing interests

The authors declare no competing interests.