We extracted 110 separate primary descriptive codes from the Results sections of the included studies (Multimedia Appendix 5). We used a combination of grouping and reflective review to categorize these themes, which resulted in 30 secondary codes (Multimedia Appendix 6) and 13 core concepts. Secondary codes and core concepts were synthesized to give 5 themes that seek to encapsulate all aspects of user engagement: Convenience, Ease of Use, Education, Clinical Care, and Communication (Figure 2). Although themes are applicable across patients, carers, and health care professionals, they were associated with different secondary codes and core concepts.

Convenience is any aspect of the intervention that makes the user’s life easier than the preintervention stage, by either time-saving or giving them the freedom to do more, or enhances their comfort in the current environment.

Positive patient experiences included those where the intervention had saved the patient travel time, or opened up new options for the patient to manage their care at home, and be comfortable in their home environment [20]. Negative experiences were associated with extra work, for example, in a complex self-monitoring system that takes time away from daily activities [35]. Loss of control was also commonly reported as a negative experience [35]:

For carers, positive aspects related to increasing options for managing care and increased freedom. Most carer experiences fell into negative codes regarding losing control of their tasks and increasing their workload and hence their stress. Throughout this theme, the concepts of workload, responsibility, and stress correlated with each other with regard to trialing a new intervention.

Positive experiences for health care professionals related to time-saving through faster and more efficient decision making. An example from Taylor and colleagues [37] showed how remote care technologies can help in understaffed situations:

Negative experiences related to the workload associated with implementing the intervention. After using the intervention, some health care professionals also reported that patients had become too dependent on the staff via the increased monitoring they provided [25]. Many clinical professionals were also reluctant to change as new methods were seen to restrict their preferred working pattern [39].

This refers to the technical design elements of the intervention, and how user-friendly and accessible it was.

Positive patient experiences were associated with an easy-to-use intervention, with simple instructions and very few technical difficulties. Patient negative experiences involved devices that were intrusive, difficult to use, and had many technical difficulties. High costs and unreliability of the intervention made some technologies “less accessible.” For example, Agrell and colleagues [18] described a remote care home hub with integrated stethoscope and blood pressure cuff, which proved difficult for sick patients to use.

Positive carer experiences centered around technology that was easy to use (eg, had a clear user interface) and reliable. Much of the negative feedback from carers was related to the stress caused by technical malfunctions of the device [54].

Health care professionals were positive about interventions that were easy to administrate and explain to patients and staff, or provided information in an intuitive manner. They favored interventions that took less time and effort to install and establish into current practice, such as a system that integrates remote monitoring into existing electronic patient records [52]. Negative feedback from health care professionals was associated with high cost, lack of security, poor user interface, lack of training provision, and unreliable results. An added issue for health care professionals was the lack of support and technical expertise to deal with problems [44].

This relates to the ability of the intervention to provide appropriate, user-tailored information.

Positive patient experiences were reported in interventions that delivered self-care information in a simple, structured format, which was easy for patients to make use of. This helped increase their confidence in managing their condition, which in turn increased their engagement [55,64]. Negative patient experiences occurred where either useless or irrelevant information was given. This included remote care devices that give no immediate feedback, and websites or electronic health records that used medical jargon [55].

For carers, it was important to learn about the condition of the patient and understand disease progression. It was also important to access up-to-date information on the status of the patient. Remote interventions that made this easy for them generated positive feedback. When this information was difficult to access or mired in inaccessible jargon, it led to disengagement and discontentment [54].

For health care professionals, education centered on informative benefit that an intervention added to current practice. Positive experience involved interventions that provided guidelines or clinical suggestions to aid in patient management. Clinicians also identified the educational benefit on self-care habits of patients [27]. Education provided through a remote care intervention raises the standard of clinical treatment and involves the patient in their own management. This continuous benefit is valued by health care professionals. By contrast, the negative experiences of education involved insufficient, irrelevant, or poor-quality clinical information [22]. An intervention that did not provide feedback or allow patients to be updated on their health status provides little motivation for using the device or improving self-care, and this had a knock-on effect on health care professionals’ perceptions of the usefulness of the device.

This was a clearly defined theme that involved either improving or hindering the current clinical care provided for heart failure management. In most cases, the effects of this theme could be measured and observed in terms of patient outcome over the long term.

Patients valued interventions that supported better symptom control and improved confidence in their treatment management plan. They often commented that certain remote care interventions made them feel more “looked after” once they were aware that their measurements were being monitored from afar [22]. Negative experiences involved difficulty integrating the intervention into their daily lifestyle. Some patients felt that the intervention had no effect and that the technology was ultimately not needed [28].

Positive aspects for carers focused on them feeling more supported while using the intervention. The reassurance of extra clinical support helped them perform their task so that they felt they need not struggle alone [20,54]. Negative aspects of clinical care from carers were often found where patients already had sufficient care to meet their needs. Additionally, there were cases where the patients were too sick to be under an automated monitoring system and required personal supervision. Where carers mistrusted technology, its introduction was an extra source of stress [35].

For health care professionals, positive aspects involved being able to administer guideline-recommended management reliably and with less error. The ability to provide proactive treatment was also considered to be of great benefit. Often, increased monitoring frequency allowed the staff to identify deterioration more quickly and intervene earlier to prevent worsening outcomes. This improved the safety and quality of management and also fulfilled a “safety-netting” criteria that were found to be very valuable [22]. Remote care bridges the gap between primary and community care, where patients with heart failure are seen infrequently and may have periods of long stability, but may also be at risk of sudden deterioration. Negative aspects of clinical care for staff were equally impactful. They involved interventions that disrupted management, often by giving unreliable, incomplete, or false information, or simply made no difference to management decisions, or patient outcome. In the example by Sharma and colleagues [44], the remote care offered features such as home blood pressure monitoring, but did not provide enough information to assess for an infection. Health care providers perceived that some technology may hinder usual care in this way, and disrupt the current “face-to-face” care.

This theme encompasses the quality and frequency of interpersonal contact involved with the intervention.

When discussing communication, patients often referred to the frequency of contact with their nurse or doctor. In general, patients favored human contact to being monitored by a noninteractive device. Therefore, interventions that facilitated human contact appealed to patients [18,29]. Patients commented that greater human contact helped with their feelings of isolation and encouraged them to self-care. Remote care interventions that facilitated interaction with health care professionals were seen as valuable and worth the investment. Patients also liked interventions that could connect with their family and friends for further support. Some social interventions connected patients with each other and these showed widespread approval within this cohort of patients with heart failure. A chronic illness that reduces mobility often impacts the social activities of the patient and it is not surprising that this isolation can often be overlooked when considering these interventions [65].

Conversely, many of the remote care interventions led to reduced contact with physicians. Patients often missed the human contact element of regular clinic visits. Rather than having a device to provide information for them, they preferred being able to ask questions from staff members directly [18]. Without a human contact element, some felt more distant and perceived that something was lacking from the consultation. While remote care may obtain information efficiently, sometimes human contact is more reassuring and has a bigger impact on patient engagement.

For carers, “isolation” was predominantly related to the fear of leaving the patient on their own. The ability of remote care to mediate communication between both the patient and carer with health care professionals was seen as valuable and reassuring [26].

Health care professionals are often concerned with patient contact and with connection to specialists. Positive communication with other staff members is found in interventions that seamlessly connect multidisciplinary teams [37]. This in turn encouraged teamwork. Regarding staff to patient communication, health care professionals felt that frequent contact led to better awareness of the patient’s tendencies and hence earlier and better decisions on their care [27,31]. The increased availability of communication with patients via remote care was felt to increase trust between the staff and their patients and staff felt their patients were more open to them. This fosters a good clinical environment, especially in the management of chronic disease where familiarity with the patient is vital to detecting early deterioration. Increased communication opportunities also motivate both parties to continue to use the device, for reassurance and safety-netting. Negative staff experiences with communication included systems that did not connect with other team members or caused a disruption in teamwork. An example was a remote monitoring system that provided the general practitioner with extra information but was unable to send this information on to heart failure hospital consultants, and thus made referrals harder [22]. Staff also found that interventions that reduced interaction with patients received much less support from both parties. Reduced face-to-face time due to remote intervention created a sense of distance with the patient, which was concerning to some staff [22].

In our review of interventions and the populations they were used on, various health inequality issues became apparent: usability, patient selection criteria, demographic distribution, and the potential to increase inequality gap through implementation of remote care.

In a heart failure cohort, interventions need to be usable by patients who have visual impairments or problems with dexterity [43]. If the intervention is not designed to take this into account, it can exclude the patients that might have benefitted the most. Riley and colleagues [34] recount an example of one such patient who struggled to use a remote care device, leading to undue stress:

In addition to co-ordination and visual problems, some patients were unable to use a weighing device because it required them to remain standing for several minutes [26].

In any interventional trial, inequality may arise between patients who do and do not have access to remote interventions. Clinical decisions on which patients are most suited to the intervention can be subjective. The weight monitoring intervention in the study by Johnston and Weatherburn [26] demonstrated this selective service based on clinician opinion, creating a defined cohort of highly monitored patients above those of current care. With limited resources, clinicians may set up their own criteria for prioritizing access to remote care based on need, as they do with other treatments [22]. Over a short intervention trial period, some clinicians felt that monitoring was not useful in patients they deemed stable. However, this may lead to situations where patients are only provided with remote monitoring once deterioration has occurred.

In intervention trials, the criteria for proving the new remote care intervention are often decided a priori by the research group. Studies considered here show a distinct preponderance to White ethnicity and male gender. In addition, there is generally poor reporting on religion, socioeconomic status, social capital, disability, and vulnerable groups. It is therefore difficult to assess the impact of remote care integration on more disparate groups within populations.

Some clinicians showed anxiety that technology could widen the health inequality gap. The study by Earnest and colleagues [55] highlighted concerns around how remote care would translate in areas with fewer resources. Many remote care interventions take the form of highly technologically advanced, expensive devices, which, while rich with clinical data, are prohibitive for widespread use in terms of costs and resource drain. It is worth noting the impact of these interventions on the rest of the clinical care of the population, whether it creates a wider benefit to the whole or simply allows greater monitoring for those who have the most care in the first place.

We have attempted to mitigate bias by double-coding, cross checking, and using grounded theory processes to construct a “blank-slate” perspective. However, in line-by-line coding some lines may have a double-meaning, fitting into multiple codes, and decisions on coding are subjective.

The studies included were of a wide range of different methodological and analytical qualities. First-order codes from poor-quality studies may be subject to greater variability than high-quality studies. However, we included studies of all quality to avoid missing useful information and to achieve saturation. The wide range of study sizes may skew the experiences toward those that are mentioned in small sample sizes, as it increases the frequency of appearance of the related codes. We have listed the sample size of each study in the characteristics of studies table (Multimedia Appendix 1).

The researcher’s reflexivity plays a role in the overall generation of the themes and can lead to first-degree bias. This process happens both inductively and intuitively and it cannot truly be distanced from preconceptions and beliefs of the reviewers themselves.