3.1 Characteristics
Table 1 overviews the 19 HCPs and five other stakeholders who participated in this qualitative study. The interviews included 9 GPs and 10 PNs from 10 primary care practice centres, with a mean age of 41 years and a range of 31–65 years. The focus group discussions included three GPs, three PNs, two project managers, two IT-domain experts and one health insurer. The three GPs and three PNs were also involved in the interviews.
3.2 The adoption of a panel management approach
Most participants experienced the current CVRM workflow functioning sufficiently well. Nevertheless, they emphasised the potential for enhancing the personalisation of care for patients with varying levels of CVD risk. 'In itself it does work, but I am concerned about the frequency of visits of patients. It may not be necessary for them to come in so often, and at times it might be more beneficial to have a new person scheduled for that time slot instead.' [Professional ID: 13] Participants also mentioned the challenge of identifying everyone at risk of CVD. They perceived the current CVRM approach as mainly reactive rather than proactive, underscoring the importance of tailored healthcare. ‘There is a significant demand for customised interventions, incorporating individualised adjustments, which could be effectively achieved by implementing intelligent algorithms.’ [Professional ID: 4]. Panel management has the potential to enhance personalised care and proactive measures.
3.2 Identification of potential barriers and facilitators
Regarding the prespecified CFIR domains, the most important barriers and facilitators of each panel management step were summarised in Fig. 2. An extensive overview of all the barriers and facilitators can be found in Supplementary Tables S1, S2, and S3.
3.2.1 Panel management steps
Step i. Empanelment
Intervention Characteristics
Different perspectives were expressed regarding identifying individuals with a similar risk of adverse care events and their allocation to administrative subgroups. GPs suggested that patients with well-controlled blood pressure or those at low risk of complications would benefit the most from RPM. According to them, such patients did not necessarily need in-person visits but could be monitored remotely. GPs also pointed out that patients starting or changing medications might require a significant number of consultations in a short period, making RPM beneficial for them as well. In contrast, PNs recommended including individuals with unhealthy lifestyle habits, comorbidities (such as diabetes and poor kidney function), and young adults at high risk in the RPM panel. Four participants expressed the view that digital health should not be limited to panels but should be widely accessible and customisable based on individual needs rather than solely their level of risk. Additionally, some professionals suggested that stratifying patients into groups and linking them to different interventions might distribute their workload more effectively. 'Indeed, it generates additional work as different approaches are required for each group, thereby necessitating filtering and subsequent actions to be taken.' [Participant ID: 3]
Outer Setting
All stakeholders considered collaboration with external partners essential for establishing the digital infrastructure. Primary care practices in the region had formed close partnerships with collaborators responsible for the technical infrastructure, which was crucial for implementing the panel management approach alongside the RPM intervention. This collaboration will additionally simplify the utilisation of risk stratification tools, which leverage risk algorithms developed by private technology developers.
Characteristics of individuals
Most participants valued an empanelment approach, leading to a more tailored consultation frequency and potentially reducing the time needed, thus lessening the burden on patients and professionals. However, some individuals expressed concerns about the effectiveness of this empanelment approach, stating that not all risk factors necessary for stratification could be accurately captured through quantifiable metrics alone. Factors like social determinants, including socioeconomic status, language, and literacy skills, might not be fully reflected in the available routine data and, therefore, might not always be recorded in the EMR.
Process
Utilising risk algorithms to stratify the primary care practice population into patient panels is a relatively new field and operating method. Consequently, participants advised that clear communication with all stakeholders engaged in the implementation regarding the purpose and advantages of the risk stratification approach is vital for its adoption. Practical methods such as training during kick-off meetings and information sessions at the primary care practice centers were identified as effective ways to promote this communication. 'For effective engagement, it is essential for individuals to be well-informed about the process and benefits of stratification. Therefore, the system should be designed to be easy to use, requiring minimal time and effort to navigate, and approachable to enhance user adoption.' [Participant ID: 18]
Step ii. Appropriate intervention
Intervention characteristics
GPs and PNs suggested that the proposed RPM intervention could enhance lifestyle, self-management, and shared decision-making, especially when considering reimbursement, education, and technical support. Furthermore, the RPM intervention was anticipated to provide a population-level perspective without incurring direct labor costs. 'By enabling the reallocation of time spent on low-risk patients, digital health technologies can potentially improve care for high-risk patients. Thus, where feasible, removing this burden of low-risk patients represents a vital aspect of such technologies.' [Participant ID: 6 & 24] Anticipated adverse outcomes included excessive data to be processed from home measurements, which could burden PNs due to extra data processing, additional questions and extra consultations, whether in-person or over the phone. Additional potential obstacles to adoption included introducing a supplementary digital dashboard alongside the current EMR and the inability to use devices that patients had purchased independently.
Outer setting
Despite the potential of the interventions to improve self-management skills among patients, concerns were raised about their limited effectiveness among patients with low literacy levels, as these interventions might primarily benefit the 'worried well'. Participants emphasised the significance of collaborating with patient organisations. Regarding the implementation process, HCPs and project managers stressed the importance of having an implementation and/or project manager (distinct from the PN or practice manager) responsible for overseeing the project's processes and coordinating with external partners.
Inner setting
When asked about integrating interventions into existing workflows, GPs and PNs indicated that having a pre-existing technical infrastructure (including dataflow and integration of wearables) in partnership with the private entities responsible for that infrastructure is beneficial. To improve the integration of workflows in a new practice, a project manager recommended providing a clear and universally understandable manual at the beginning of implementation.
Step iii. Surveillance of care gaps
Intervention characteristics
The participants, especially GPs and PNs, extensively discussed the costs associated with the intervention and the responsibility for financing it. They were aware that funding presented a significant barrier to implementing such interventions, primarily due to reliance on one-time subsidies rather than structural reimbursement and the lack of compensation for the additional time required for end-users to implement the intervention. A healthcare insurance provider emphasised that the focus should not only be on cost reduction but also prioritise the efficient allocation of resources for patient care and promote patient and healthcare professional satisfaction. HCPs acknowledged the importance of qualitative data, including patient and professional satisfaction, in addition to clinical outcomes and cost-effectiveness. Furthermore, GPs and PNs revealed that patients often purchase their own blood pressure monitors, especially if they can connect the device to the EPD. "It would be motivating for patients to receive reimbursement for a trial period, and if they find the device effective, they can then pay a portion of the device cost," suggested a GP [Participant ID: 9].
Process
The significance of intervention ambassadors was highlighted to ensure the active involvement of patients and HCPs. These ambassadors, who can be GPs or PNs, serve as essential opinion leaders and can significantly influence patient attendance at information sessions and compliance with the interventions. Furthermore, it is crucial to clearly explain the purpose and benefits of the panel management approach during contextual activities such as introductions, training sessions, and e-learning courses. Additionally, providing a clear protocol for managing specific patient panels and specifying whom to contact for technical support is essential.