Lessons learned from piloting a heart failure-based remote management care system in a general internal medicine program: A qualitative evaluation

Abstract

Objective

Medly is a remote patient management (RPM) tool that allows patients with heart failure (HF) to record daily measures to inform health status and receive personalized care instructions by a validated algorithm. Medly has been standard-of-care for patients with HF in the Cardiology Division at University Health Network since 2016. While patients with HF are often admitted to general internal medicine (GIM), there are few HF RPM programs used in the GIM setting. We piloted Medly for patients discharged from GIM in the summer of 2023. The purpose of this study was to understand staff factors that enabled or inhibited spread and scale of Medly into a GIM setting.

Methods

Semistructured interviews were conducted with 13 staff involved in the planning, delivery, and implementation of Medly in the GIM setting. Interviews were recorded and transcribed verbatim. Data were analyzed using deductive thematic analysis informed by the Consolidated Framework for Implementation Research.

Results

Implementation factors included: (1) communicative staff, (2) Medly adaptability and relative advantage, (3) need for long-term funding, (4) leadership engagement and learning-centered culture, and (5) sociodemographic characteristics of patients.

Conclusion

This study offers valuable insight into how an RPM can be adopted from the Cardiology to GIM context. Key enablers include providing appropriate staff time and compensation, engaging leadership early and often, modifying the application for the setting, and considering equity with respect to language barriers. A larger pilot with more than 10 patients actively embedded in the GIM pathway is required for further scale of the program.

Introduction

Heart failure (HF) is a chronic condition that affects approximately 827,000 Canadians. ¹ It occurs when the heart does not pump adequately for the needs of the body, causing a myriad of symptoms including reduced functional capacity and impaired quality of life.^1,2 HF is a costly, complex, and debilitating condition that requires resources and support from several levels of the healthcare system. ³ General internists manage most HF hospitalizations in Canada.^4,5 Despite an increasing number of cardiologists in Canada between 2009 and 2018, there was a decreasing number of patients with HF admitted to cardiology wards during this period, with 58% of patients with HF admitted to a general internal medicine (GIM) specialist. ⁵ Patients with HF are at high risk of deterioration requiring readmission following discharge from hospital. ⁶ As such, dedicated transitions of care from inpatient to outpatient care are recommended to reduce avoidable readmissions and improve patient satisfaction. ⁶

Transition of care programs involve activities designed to facilitate safe and smooth shifts from one setting of care (inpatient) to another (outpatient). ⁷ These programs can reduce avoidable rehospitalizations and promote patient satisfaction. ⁷ One such transition of care program that can improve quality of life and other outcomes for individuals living with HF includes virtual care ³ such as remote patient monitoring (RPM). Using an RPM to aid transitions of care for patients with HF has become increasingly popular since the COVID-19 pandemic. ⁸ An RPM involves patients entering data, which is transferred to healthcare providers using remote access technology (e.g. external, wearable, or implantable devices). ⁹ This allows for rapid and continuous assessments of HF symptoms, providing information on HF progression and remote management of HF. ⁹ A meta-analysis involving over 6000 patients revealed that RPM was associated with a significantly lower number of deaths and hospitalizations for patients with HF. ⁹

Medly is a Health Canada approved RPM technology that has been standard-of-care for 8 years in an quaternary care heart function clinic in Toronto, Canada. ¹⁰ Medly is a mobile application where input daily measures of HF data (blood pressure, weight, heart rate, and symptoms) and receive personalized care instructions by a validated algorithm based on their clinical status (e.g. take an additional dose of a diuretic and present to the emergency department). Medly concurrently provides alerts to the patients’ care team when measures fall outside of predetermined, personalized thresholds. A pragmatic study involving 315 patients using Medly between 2016 and 2019 showed that there was a 50% decrease in heart failure-related hospitalizations, clinically meaningful improvements in heart failure-related quality of life, and significant increases in patient self-care. ¹⁰

While patients with HF are often admitted to GIM physicians, ⁵ there are few HF transition of care programs that are tailored for the elderly and multimorbid patient population typically cared for by general internists. Further, there are no HF transition of care programs to our knowledge that incorporate RPM for patients from GIM. Studies that previously assessed the impact of RPM for patients with HF occurred mostly in cardiology. ⁹ These studies also often excluded patients with comorbidities that are commonly seen in GIM inpatients that have HF, including cognitive impairment, chronic kidney disease, cirrhosis, or substance use disorders. ¹¹

Given Medly's positive impact, coupled with the fact that there is a paucity of GIM-centric RPM, we sought to expand Medly to patients discharged from GIM with a diagnosis of HF. With the input of key GIM and Cardiology collaborators, the Medly GIM pathway was co-designed and implemented with 10 patients enrolled. ¹² While the Medly app and clinician dashboard were unchanged, the workflows differed from the Cardiology model; notably the defined enrollment period was limited to 30 days, the GIM integrated care coordinator supported enrollment and care coordination, and GIM physicians had dedicated clinic time to see patients when urgent. Patients were enrolled for 30 days as GIM has limited capacity to follow patients beyond 30 days following discharge from acute care, and because patients discharged with heart failure should be seen by a healthcare professional within 7 to 14 days of discharge as per Health Quality Ontario and the Canadian Cardiovascular Society, respectively.^13,14 A short paper on patient outcomes related to this pilot was recently published, demonstrating high patient fidelity to Medly, no heart-failure related deaths, re-hospitalizations, or emergency department visits within 90 days of discharge. ¹²

The aim of this study was to examine healthcare provider and staff's perspective on implementation of Medly in the GIM context, to understand factors that will facilitate or inhibit spread and scale in the future. This research was guided by implementation science, which is the study of methods to promote the uptake of innovations into practice. ¹⁵ To guide this research, the updated Consolidated Framework for Implementation Research (CFIR);, ¹⁶ the most highly cited implementation science framework, ¹⁷ was employed. CFIR is a determinant framework that aims to categorize factors related to implementation effectiveness. ¹⁸ The framework can be used flexibly to explain implementation retrospectively or predict implementation prospectively. The CFIR framework was supplemented by the non-adoption, abandonment, scale-up, spread, and sustainability (NASSS) framework. ¹⁹ The NASSS framework was specifically designed to identify implementation challenges for digital health interventions, and is well suited for adaptive interventions while capturing system complexity. ¹⁹ Accordingly, the purpose of this study was to understand staff factors that enabled or inhibited implementation of Medly in GIM, in order to help promote spread and scale of Medly across diverse contexts.

Methods

Design and research paradigm

This study followed a qualitative research design. The research paradigm guiding our study is pragmatism. Pragmatism rejects the dualities of objectivity and subjectivity (and thus dichotomies of postpositivism and constructionism) and allows for intersubjectivity (i.e. there is a single real world, and individuals have their own interpretations of that world). ²⁰ In research, pragmatists are concerned with solving the practical problems of the world by focusing on how the knowledge will be used. ²¹ For example, this study will help to inform the translation of Medly into other areas of medicine and other settings. As pragmatists, we recognize that both theory (i.e. the CFIR and NASSS) and individual experiences matter and are necessary for understanding the implementation of technology in healthcare settings.

Setting and participants

Participants (n = 13) were purposively sampled based on their direct experience with the Medly GIM implementation, including every clinician and staff member involved with the pilot. Participants were emailed by the interviewer to invite them to participate in the study. We included staff in diverse roles related to direct patient care, healthcare planning, care coordination, and departmental leadership. No participants dropped out or refused to participate in the study. We did not interview patients because this study focused on the role of healthcare providers and staff in implementing Medly, rather than patient experiences with Medly, to focus on establishing the pathway in the health system. Further work is now in progress to build on this pilot where GIM patient perspectives will be included.

Participants included staff and clinicians that were involved in the planning, delivery, and implementation of Medly in GIM at a tertiary care academic hospital in Toronto, ON. Ten patients were enrolled in Medly in GIM between 27 May and 24 July 2023. A previous study has been published which includes the description of these patients ¹² as this was not the focus of the current paper. Interviews with staff were conducted between August and December 2023. This hospital is the same hospital where Medly is standard of care in Cardiology. GIM is the largest admitting service at this hospital and the majority of HF acute care hospitalizations are to GIM (428 of 827 admissions; 52% yearly average from 2017/2018 to 2019/2020). No RPM has been integrated into the GIM context prior to this pilot. The only program offered to patients with HF upon discharge is enrollment into the hospital's integrated care program, which provides homecare supports (e.g. nursing and personal support workers) and care coordination (e.g. postdischarge specialist appointment confirmation) postdischarge. ²²

Data collection

This study was approved by the University Health Network's Quality Improvement Committee (QIRC 23-0606). Participants provided written consent for participating in the interview when responding to the study invitation, which provided information on the interview process and research goals. Participants provided verbal consent during the interview and before recording. All interviews were conducted by a female author (SVCL), who was conducting a Health System Impact Postdoctoral Fellowship at the time of the study. The interviewer had 8 years of experience conducting interviews and qualitative research. The interviewer had an established relationship with one of the participants before the study. The other participants worked in the same setting as the interviewer and often interacted professionally through email and attending meetings/trainings together. Participants knew that the interviewer was conducting a postdoctoral fellowship and was interested in supporting digital health interventions across healthcare settings.

Interviews were conducted between September and December 2024. Interviews occurred over video conferencing software (Zoom, San Jose, CA) between the interviewer and participant and ranged in length from 20 min 16 s to 30 min 16 s (M = 26.8 min, SD = 2.9 min). No repeat interviews were conducted. The interviews were audio recorded and then transcribed verbatim using Microsoft Word (Microsoft, Redmond, WA). The interviewer took notes during the interview and reflected on these notes during analysis. The first author (SVCL) reviewed and corrected the transcripts for accuracy. The transcripts were offered to participants to review, but no participants provided additional feedback. The interview guide (see Supplemental file) was developed with the goal of using data to support future implementations of Medly in other healthcare settings. The interview guide was informed by the CFIR. ¹⁶ The interview guide was co-developed by SVCL and WKS, and pilot tested between the two authors (see Supplemental file). The interviews were semistructured and flowed based on participants’ interests and experiences. Questions were structured to include introductory, key, and closing questions with probes. Key questions included “Tell me about some factors that enabled the success of this pilot; and What changes are needed to facilitate this pilot into routine workflow?”

Data analysis

Using NVivo (Lumivero) to manage the data, a thematic analysis was employed to generate themes underpinned by core concepts in the interviews. ²³ Data saturation was not a focus of this study, as we aimed to interview all staff involved with the Medly pilot in GIM. We focused on collecting diverse experiences and the depth of responses within each experience, rather than identifying common patterns across multiple participants. Since we embraced a pragmatist approach, the focus was on practical considerations and the utility of knowledge rather than on generalizability. The justification for the sample size—interviewing all staff who participated in the Medly pilot—emerges from the study's intent to capture the full range of perspectives among those directly involved in the implementation. Since the group of staff engaging with Medly was finite and central to the pilot, including all of them ensures that the data reflects the diversity and depth of relevant experiences, rather than aiming for saturation or broad generalization.

Data were analyzed deductively using the CFIR ¹⁶ and NASSS ¹⁹ to identify themes relating to factors what would enable or inhibit implementation of Medly into new settings. Specifically, the analysis first relied on the CFIR to identify barriers and facilitators. The CFIR is useful in providing structured constructs using a broad, determinant-based framework, whereas the NASSS framework clarifies overall complexity in adopting technology in the health system. Therefore, the NASSS domains were examined after the CFIR coding to layer in themes that were not captured in the CFIR. This was especially important for themes around patient characteristics to catch the complexity of the implementation context. The themes were generated primarily from the CFIR and NASSS frameworks but modified and adapted as needed to ensure the experiences of participants were not changed to fit the framework. Discussion about the application of findings to GIM settings was central to the analytic process.

The first author listened to interviews and re-read transcripts to familiarize herself with the data. Transcripts were initially annotated using semantic codes to capture the surface-level meaning of the data. The first author made refinements to initial codes by grouping codes together to begin generating themes. The CFIR and NASSS frameworks were used to apply codes to initial themes, which were then reviewed and further refined. The final step was to name themes and further group themes together where necessary. Participants were offered to provide feedback on the themes and two participants engaged in this process. Reflexivity was central to creating themes and these themes were discussed in depth with a critical friend (WKS).

Results

Five participants identified as men and nine participants as women. See Table 1 to understand the roles of participants in implementation: implementation leads and innovation deliverers (ILIDs), innovation deliverers and implementation team members (IDIMs), other implementation support (OISs), and mid-level leaders (MLLs). These roles were assigned according to the construct definitions in the CFIR provided by. ¹⁶

Table 1.

Implementation roles of participants.

Number of participants	Name of role	Description of role
N = 3	Implementation leads and innovation deliverers	Physicians directly involved in the delivery and care of patients with heart failure
N = 3	Innovation deliverers and implementation team members	Nurses involved in care
N = 4	Other implementation support	Employees providing project management
N = 3	Mid-level leaders	Physicians involved in a leadership capacity

Five major themes related to the CFIR (themes 1–4) and NASSS (theme 5—Patient Characteristics) that enabled successful implementation of this work were generated through analysis: (1) Characteristics of Staff: Supportive and Communicative, (2) Innovation Domain, (3) Outer Setting: Financing and Policies, (4) Inner Setting, and (5) Patient Characteristics (see Figure 1). Themes (and subthemes) will be described below.

Figure 1.

Themes and subthemes of Medly implementation in general internal medicine settings according to the consolidated framework for implementation research. ⁴

Note: *This theme was informed by the nonadoption, abandonment, scale-up, spread, and sustainability framework. ¹²

Characteristics of staff: Supportive and communicative

The first theme relates to the characteristics of staff that supported Medly implementation in GIM. This theme is modified from the “Individuals” and “Characteristics” domains. ¹⁶ Participants of all roles described how valuable it was to work with team members who were supportive of each other and patients. A physician discussed his perspective on the nurses’ roles: “…the nurse practitioner support. I think that was really invaluable having people not only with expertise, but who could also dedicate their time to actually give these patients the time of day” (P1). A nurse in a separate interview stated similar qualities about those physicians: “Are they going to be receptive to that communication, that patient review and [P1] and [P6] were amazing and they were always available to us and like communication was really, really good” (P5).

Innovation domain

The second theme relates to the Medly product and its applicability to the GIM setting. There are three subthemes for the Innovation Domain, including adaptability, relative advantage, and trialability.

Adaptability

Adaptability refers to the ability of Medly to be modified, tailored, or refined to fit the local context or needs of GIM. A nurse practitioner discussed Medly adaptability:

I think that [Medly] needs to evolve for these patients. They are complex, they have multiple layers of comorbidities, it's almost like a necessity to have in their care to help them navigate through the system … in order for [Medly] to be successful, it needs to evolve with the patients. And there needs to be kind of built-in feedback from the patients or providers … It has to be Medly GIM, not Medly Heart Failure, trying to fit into Medly GIM. (P10)

The participant highlights the importance of allowing for continuous feedback to be built into the Medly system for new settings to allow for iterative adaptability. The needs and contexts of patients and settings are so diverse that Medly cannot be copied directly from a heart failure setting and inputted into a GIM setting. For Medly GIM, this included changing the defined enrollment period to 30 days, having the GIM integrated care coordinator provide enrollment and care coordination, and allowing GIM physicians to have time to see patients when urgent.

Relative advantage

Relative advantage refers to the idea that Medly is better than other innovations or options available in GIM for heart failure patients. A physician discussed the importance of having a resource for heart failure patients:

…embedding it into integrated care ensures that patients have follow up after cardiology… I think just the way general medicine is practiced in Toronto is that we don't have a lot of capacity. We don't really follow patients on a longitudinal basis for things like heart failure and so at the outset, sort of making sure that patients had some follow-up. And I thought it was really helpful. (P1)

Medly demonstrated a relative advantage for heart failure patients, and a nurse discussed how Medly facilitated their role as a staff member supporting patients upon discharge:

I have to say I was working so much overtime the whole year. And then the Medly pilot came on and I was like, wow, I can breathe, you know? Like for once I can have this little breather right now, it was perfect. It was so good. (P9)

Thus, Medly showed relative advantage for both patients and staff.

Trialability

Trialability refers to the fact that Medly was piloted on a small scale with 10 patients. Participants liked that they were able to test out Medly first but recognized that a larger scale implementation was required to test the applicability in GIM and features for success in order for Medly to be permanently embedded into GIM:

I think what I would see the next phase would be a larger research grant. Some more support for a physician and another fellow to pilot 100 patients for a year… taking that 100 patient pilot and presenting it to the Ministry of Health for permanent funding, so I think I'm hoping we that can be the next phase, a one year 100 patient pilot and then again presenting that data. (P7)

This perspective from a physician leader demonstrates that Medly should be trialed again on a larger scale to demonstrate effectiveness to secure funding for full implementation.

Outer setting: Financing and policies

The third theme relates to the outer setting, which refers to the larger hospital system and the provincial ministries and agencies that fund and run Ontario hospitals. This domain was modified from the CFIR to combine “financing” and “policies.” Financing and policies related to health professional regulation and funding directly impacted the success of Medly in GIM. Participants described that Medly in heart function settings has been implemented outside of the scope of a heart failure physician's financial compensation because they view Medly as being so beneficial. Participants foresee this as an issue for GIM long term:

There's not really any billing codes or billing methods in the Ministry of Health and the schedule of benefits when we provide that support. The emails, the messages, the contact with the Nurse Practitioner's, the discussions with the patients even I was phone calling, giving [patients] advice over the phone, arranging their appointments in the clinic, all of that additional administrative work…No, the ministry doesn't support that currently. And there's not really even a hospital role, where the hospital itself or the university aren't supporting that. So, it had to be through research, right? (P6)

This physician describes that current standard funding practices from the province do not allow for billing when using digital healthcare or innovations like Medly. Therefore, there needs to be either considerable, sustainable funding from research or advocating for new billing practices at the funder's level.

Inner setting

The fourth theme relates to the inner setting, which refers to the University Health Network GIM setting. Given the several subthemes related to the inner setting, Table 2 was used to display relevant descriptions and quotes. The subtheme of “leadership engagement” was adopted from the first version of the CFIR.

Table 2.

Subthemes related to the inner setting domain for the implementation of Medly in GIM.

Inner setting subtheme	Description	Quote
Work Infrastructure	How the GIM team roles and responsibilities may need to be changed to allow for Medly to be scaled appropriately	“[P1] and [P6] offered up their availability and their time for those initial 10 patients. But I think if we were really to kind of sustain this and embed this and scale this within GIM, I think a different resource support model would be required again for that MRP support and for the Medly Coordinator support because that was another challenge. [P9] was able to identify the patients that she had on the integrated care pathway that were heart failure patients and … would be a good candidate for Medly. But that's not within the capacity of her role … So again, … like an all teams coming together to be able to do this for the 10 patients, but I think when you talk about scale and spread, I think a more sustainable plan around resourcing would probably be required” (P2, Other implementation support)
Relational Connections	The GIM team had high quality formal and informal information sharing practices across all levels of implementation	“It was a really positive experience for me. The line of communication was amazing between the MRP's and [P9] and I found out we had a lot of support with their care that the patients needed that's outside of Medly. So if patients needed extra support with social work or getting resources available to the patients and their family, we were able to turn to [P9] for a lot of help and assistance.” (P8, implementation team member and innovation deliverer)
Learning-Centeredness Culture	The GIM team had shared values, beliefs and norms around continual improvement and experimenting with new innovations for the betterment of staff and patients	“And everybody felt that we needed to step into a time of renewal and regeneration because the crisis of physician burnout, wellness, the patient volumes crisis, the fact that we have to take a heavily oversubscribed inpatient hospital footprint and somehow move it out to the community urgently—sort of drove all of this thinking as well … I think like it took a crisis to kind of start thinking that, hey, we really can't just do the same old thing. We have to rethink all of our models of care.” (P4, Mid-level leader)
Compatibility	There were challenges fitting Medly within the workflows, systems and processes of the GIM team especially with other innovations and the use of the electronic health record	“I think … just part of the workflow development piece—it's challenging at times to figure out where everything fits. …we have a lot of different programs here at UHN and it's all very exciting, but really trying to figure out how they seamlessly meld together, how it's always been an exciting challenge, I would say.” (P3, other implementation support)
Leadership Engagement	GIM had many leaders who were committed, involved, and supportive of implementing Medly which led to its overall success	“I'm in a leadership position … Having a fellow who could assist me with the work was a huge bonus… having the right leadership group involved in it… So we had a group of people who had a lot of expertise in rolling out new programs and how to how to analyze these programs. And that's really [P1's] area of expertise as well. So we really lucked out with the collection of people who were interested in this. I think other enablers, obviously [lead of Medly for the heart function clinic] and the entire Medly group was very supportive” (P6, Implementation lead and innovation deliverer)

GIM: general internal medicine.

Patient characteristics

The fifth theme relates to characteristics of patients (innovation recipients) who participated in Medly and how these characteristics may be unique to a GIM setting compared to a heart function clinic setting. The subthemes include language, digital health literacy, and comorbidities and sociocultural influences. This theme was developed based on the NASSS framework for technology programs. ¹⁹ The NASSS framework discusses the importance of recognizing the complexity of patients’ conditions in how successful technological innovations will be for spread and scale.

Language

Participants noted that GIM patients were more likely to speak English as a second language than patients in the heart function clinic. As one project manager said:

I know there's been some great strides to actually develop resources in different languages. But I do think that when we look at the [GIM] population, that does continue to be one of the greatest limiting factors, particularly when we look at Toronto Western. And this isn't just for Medly. I think it's all for any digital type, remote care monitoring application, this an opportunity to make that a bit more accessible to you know the broader population. (P2)

There have been efforts to include other languages in Medly, and this will need to be ramped up for the GIM setting.

Digital health literacy

A general concern in both the heart function clinic and GIM setting is a lack of digital health literacy. These populations tend to be older and not accustomed to using smartphones, especially for health information or self-management purposes. Nurses discussed that they often worked with caregivers to encourage Medly adherence as they may be more digitally literate. A nurse discussed their alternative experience with working with seniors and Medly:

Usually, [seniors] are pretty good at navigating the app. And the older generation are usually more keen on making sure all their numbers are in every day… And you have to kind of know as a nurse, I have to know who I'm dealing with … If I know that they don't even have a cell phone … then … I know where to start…you're an educator. And you kind of explain what Medly is about. You just have to put your numbers in. It's like, no, this is monitoring your health status. This, we hope, will keep you out of the emergency room, will help you not get admitted to the hospital … we're educating them. (P5)

This nurse acknowledged that although some patients lack digital health literacy, it is the role and responsibility of the nurse to properly educate patients and their caregivers on how to use Medly to maximize effectiveness.

Comorbidities and sociocultural influences

Participants also outlined how Medly GIM patients differ in terms of their comorbidities and social determinants of health in comparison to Medly patients in the heart function clinic. A physician leader explained the GIM population:

Because of the medical complexity, because sometimes the patients are more social minorities and the marginalized internal medicine population of patients is so heterogeneous that it's hard to necessarily think that there's going to be a very clear pathway and process which can unilaterally apply to most patients. (P4)

Equity and the social determinants of health (e.g. race and socioeconomic status) need to be at the core of every implementation of digital health innovations.

Discussion

Interviews with key clinicians and supports involved in the Medly GIM pilot revealed important indicators for successful implementation and recommendations for future sustainability of Medly in diverse settings. Namely, staff were supportive and communicative and helped to adapt Medly to fit the patients in the GIM setting. Staff and patients viewed Medly as being advantageous to usual care, but there are clear gaps in funding physicians’ role in RPM programs in Ontario's current health system. Medly was also relatively successful in GIM because staff propagate a learning-centered culture for continual innovation. Future iterations need to better incorporate equity into implementation and sustainability of Medly, such as considering language, sociodemographic characteristics, and comorbidities. Models for physician billing/funding within RPM are also needed.

Characteristics of staff and inner setting

Medly implementation was successful in the GIM setting ¹² in part due to the supportive and communicative nature of staff across roles. A systematic review of hospital-based interventions demonstrated that successful implementation relies on clear communication from motivated staff who feel empowered to take on challenges. ²⁴ Participants highlighted the learning-centeredness culture of GIM that drives innovation. Staff across the department and leadership supported trying new approaches that could improve patient care. Research shows that when people perceive their organization as innovative, they feel rewarded for adopting new innovations. ²⁵ To further spread Medly across other health systems, these characteristics need to be nurtured and developed in healthcare providers. In this pilot, staff across roles were heavily involved in the organization of the study, leading to shared decision making. Shared decision making across staff can contribute to positive communication, local ownership of the implementation of Medly, and more likely to be sustainable and effective. ²⁶

Innovation characteristics

Medly will only continue to be effective if the intervention is adapted to fit its context, rather than adapting the context to fit Medly. Medly was built to be highly adaptable, as patients can receive cuff, scale and/or devices or bring their own. ²⁷ Medly demonstrates two key innovation characteristics: adaptability and compatibility, which are related to each other and the integration of new programming into a hospital. ²⁶ Shared decision making (as discussed above) can influence adaptability and computability because collaborative input involves issues related to intervention integration into existing contexts. ²⁶ Staff from all levels shared in decision making to incorporate Medly into GIM by having defined enrollment periods of 30 days, using an integrated care coordinator to support enrollment and care coordination, and having dedicated clinic time for GIM physicians to see patients when urgent.

A valuable finding from this study was that GIM staff had better experiences (e.g. work-life balance and communication) than with usual care. This is because there were more staff to support with patient care allowing work to be more evenly distributed. Medly also provided some project and nurse coordination support staff with defined roles that were clear to implement during their workday. Better staff experiences are significant as research demonstrates that nurses who care for patients with heart failure have high levels of personal and work-related burnout. ²⁸ The relative advantage of Medly for both patients and staff is important for applying for funding to scale up implementation moving forward. The relative advantage of Medly expressed by staff in this study was supported by the data from the ten patients in the pilot previously published. Silverstein et al. ¹³ found that there were no heart failure-related deaths, rehospitalizations, or emergency department visits within 90 days of hospital discharge. Medly supports Ontario's quintuple aim of healthcare as it improves population health, provides a positive patient experience, and enhances the provider experience. ²⁹ Further research is needed to understand Medly's influence on improved value and health equity.

Outer setting

This pilot study was able to be implemented because of short-term research funding acquired by the authorship team to protect clinician time to dedicate to this work. However, sustainable financing and policies are needed from funders to support Medly implementation in diverse contexts. Currently, physicians are unable to bill for care provided when using remote programs like Medly. The federal and provincial governments identified healthcare priority areas, one of which is modernizing healthcare information systems and tools for sharing information. ³⁰ A first step in supporting this priority is to fund evidence-based, equity-informed remote and digital health programs, and considering physician billing practices of digital and remote programming. While Medly was developed in-house at UHN in a nonprofit system, other RPMs are at risk. Without economic investment from funders, digital and remote health programs will continue to develop under for-profit investors driven by commercial interests, which can worsen health inequities and limit opportunities for public health impact. ³¹

Patient characteristics

For scaling up the implementation of Medly in noncardiology settings, health equity considerations and patient characteristics must be prioritized. In this pilot, efforts to incorporate equity included providing scales, phones, and data equipment to patients who could not afford these items. Quality improvement efforts without considering equity can worsen health inequities. ³² In GIM, patients with HF often have more comorbidities, experience language barriers, have poorer health literacy, and are more likely to have lower socioeconomic status as compared to those admitted to a Cardiology service. ³³ These factors intersect and are associated with increased incidence of cardiovascular disease even after adjusting for other risk factors (e.g. diabetes). ³⁴ Inadequate health literacy is associated with increased risks of mortality and hospitalization among patients with heart failure. ³⁵ As Medly scales up implementation, it will be important to provide Medly in other languages, consider the central role played by caregivers, offer health resources appropriate for patients’ financial circumstances, create tailored educational materials designed for patients with lower health literacy levels, and have nurses dedicate time to educate patients on Medly and their health.

Limitations

This study is not without its limitations. First, one of the lead authors of this study was also involved as a participant in the interviews. This may have biased the results to be more positive and favorable towards Medly implementation and scale-up. However, this author was essential in reporting on implementation as they were a key staff in the planning and delivery of implementation. The participant also provided several suggestions for improving Medly in the future. Second, only 13 participants were recruited for the study, and this research could have benefitted from more interviews. Although this might seem like a small number of participants, the authors interviewed all staff that were included in the pilot implementation. The interviews achieved acquired diverse perspectives from clinical, support, and leadership roles. Third, this study investigates the implementation of Medly at one care center, which may bias the results. Implementation considerations may be different in different wards, centers, or regions. Importantly, this work may be less generalizable to other contexts, but this research did support the scale of Medly in GIM at this institution. Fourth, only one author was involved in the analysis of the data, which may limit the interpretation of the results. However, the supervising author was involved in results discussions and provided feedback on the analysis. Fifth, interviews were conducted after Medly deployment rather than during deployment, and therefore the implementation process may not have been captured well. Future research could interview staff throughout the pilot process several times to capture the implementation process domain of the CFIR framework.

Conclusion

Overall, the pilot implementation of Medly in GIM was successful as it demonstrated advantages for patients and staff. Key facilitators for success included a learning-centered organizational culture, communicative staff, and innovation adaptability. To successfully scale up Medly, government funding for physicians caring via RPM, appropriate referrals, and reducing language barriers are needed for sustainable and equity-focused implementation. Future studies should focus on applying these lessons to scaling Medly to diverse health system settings. Importantly, a comparison between the Medly pilot implementation in GIM and fully scaled implementation in GIM should be conducted to understand the real-world impact (e.g. financial and sustainability) of Medly in the health system.