Bottom-Up Analysis of Telemonitoring Costs: A Case Study in Slovenian Primary Care

The escalating demands on global health systems resulting from the management of chronic diseases have underscored the need for innovative solutions. Arterial hypertension (AH) and type 2 diabetes (T2D) are among the most prevalent chronic conditions worldwide, with projections indicating a surge in their prevalence, particularly among the elderly (1–3).

In response to the mounting burden of chronic diseases in primary care, telemonitoring has emerged as a practical solution. Telemonitoring involves using medical devices to collect real-time physiological data, such as blood pressure (BP) and blood glucose (BG), which is then directly transmitted to a telemonitoring centre, where it triggers a response from a healthcare provider, often complemented by a teleconsultation (3, 4).

Previous studies reported that telemonitoring can effectively lower overall costs by reducing interaction time with healthcare professionals, preventing early health deterioration, reducing hospital admissions, cutting patient travel costs, and transferring specific elements of care from professionals to patients (4–6).

To seamlessly integrate telemonitoring into healthcare systems, it is crucial to pinpoint and optimise factors that influence costs. A successful strategy for identifying these factors involves employing a bottom-up approach that considers both healthcare provider and societal (patient) perspectives. This method entails a thorough examination of individual cost components, starting from specific aspects within the healthcare provider’s realm and extending to the broader societal context. The provider perspective primarily focuses on medical costs, encapsulating telemonitoring technology infrastructure and operational expenses. Simultaneously, the societal perspective encompasses more extensive effects, including indirect and non-health-related costs such as patient and caregiver time, out-of-pocket costs, and productivity loss (7, 8).

Slovenia, a high-income country in central Europe with the Bismarck healthcare model, has made significant efforts to implement an integrated care package for individuals with AH and T2D in primary care settings (9). However, the national-scale implementation of telemonitoring in Slovenia is still pending despite its prior evaluation in pilot studies (10–13). This delay could be attributed to the absence of comprehensive clinical impact and cost analyses, which would enable decision-makers to extend financial support towards telemonitoring initiatives.

To address this gap, we have designed a pilot multicentre randomised controlled trial (3) aimed at assessing the feasibility, acceptability, and clinical effectiveness of telemonitoring for older people with AH and T2D in primary care in Slovenia. The aim of this sub-study is to explore the cost structure of telemonitoring for the elderly with AH and T2D in primary care and identify factors influencing costs for potential future expansions.

In Slovenian primary care, the use of telemonitoring interventions for individuals with AH and T2D yielded notable economic benefits. The annual infrastructure costs per patient were €489.4, coupled with operational costs of €97.3 (95% CI 85.7-109.0). Additionally, patient participation costs were €215.6 (95% CI 190.9-241.1) annually, demonstrating the multifaceted financial benefits of telemonitoring. Importantly, the telemonitoring group exhibited a significant 12-month reduction in out-of-pocket costs compared to the standard care group, showcasing the potential economic benefits of our intervention (€132 vs. €545).

Previous studies have revealed varying telemonitoring costs, influenced by factors such as healthcare systems, hourly rates, and intervention intensity. For instance, a Canadian study in 2019 reported BP telemonitoring costs of €279 for the first three months and €300 annually for the next 20 years (16). In a European context the rates were higher, such as €2,104 in the United Kingdom (5), €4,859 in the Netherlands (21), and €1,962 in Italy (22). However, previous studies often required more frequent interactions between patients and healthcare workers, with operational costs being the primary driver and infrastructure costs accounting for only about one-third of the total costs (9, 16).

In our study, infrastructure costs emerged as the principal cost driver. To optimise these, we propose establishing a national or institutional telemonitoring platform. This platform, complete with the user-friendly mHealth application, has the potential to reduce additional expenses related to mobile phones and service subscriptions (4, 12). However, the feasibility of this approach hinges on factors such as the existing infrastructure, technological readiness, data protection, financial resources, and national healthcare system priorities (4, 12, 25).

The operational costs in our study were predominantly driven by data review and teleconsultation costs led by GPs. Patients exceeded the expected number of BP and BG measurements by 46.5% and 252.8%, respectively. This is an important finding, as patients voluntarily continued to take measurements even when not required, especially for BG, driving up operational costs. To address this issue, we suggest shortening the telemonitoring interval to six months, when clinical parameters stabilise (3, 12), and conducting nurse follow-up meetings every three months after the telemonitoring period to maintain the self-management behaviour learned through telemonitoring (4, 12).

Additionally, the delegation of measurement interpretation to nurses (12, 26) or the introduction of artificial intelligence for automated responses (27) could significantly lower operational costs. As nurse practitioners gain expertise, they could gradually handle teleconsultations, allowing GPs to focus primarily on making any changes to therapy that are needed (4, 12, 26).

Due to our focus on an elderly demographic, we specifically calculated patient participation costs, omitting the impractical comparison of productivity losses. In our study the patients devoted 18.8 hours (95% CI 16.6-21.0) annually to participate in telemonitoring. This is less than in previous studies where patients reported spending 10-12 hours per year on self-monitoring BP and 13-46 hours on self-monitoring BG (28, 29), and more than the expected 12 hours based on our measurement protocol (3). Implementing a less intensive BP measurement protocol (i.e., once a week) (12) and shortening the telemonitoring interval could further optimise patient participation costs.

One critical aspect from a societal perspective is the impact of telemonitoring on out-of-pocket costs. Previous studies found that these payments made up almost one quarter of all AH and T2D treatment costs in Slovenia (30, 31). In our study, both groups had an average annual out-of-pocket cost of €345, with the telemonitoring group reporting significantly lower expenses after a 12-month follow-up. The reduction in medical device costs was in line with expectations, as telemonitored patients received BP and BG monitors with BG test strips, while decreases in expenses for food, dietary supplements, and specialist check-ups could be associated with improved patient education received during training and teleconsultations (32).

Notably, there was no significant reduction in self-reported transportation costs. Older individuals in the intervention group continued regular preventative activities, including visits to their GPs for other health consultations. This was deemed essential for ethical reasons, given the study population’s various associated health conditions. However, a more restrictive approach might prove feasible and efficient in younger populations with isolated AH or T2D (17–19).

The strength of this study lies in its integration into a randomised controlled trial with elderly participants from diverse backgrounds, an underexplored demographic in telemedicine research. Nevertheless, limitations include a small sample size and the inclusion of motivated participants, potentially limiting generalisability. Moreover, we only examined the change in out-of-pocket costs at the end of a 12-month period, while a baseline assessment should be performed to compare groups and verify the results. Additionally, there were costs related to educating GPs and registered nurses on the proper use of the telemedicine platform and devices. Given that this was a one-time expense that fell significantly with increased patient volume, we have excluded it from our calculations for clarity. In future research, it is advisable to estimate costs associated with unpaid caregiver time and delve into further medical aspects of telemonitoring’s cost-saving potential, encompassing the prevention of secondary complications and hospital admissions.

In conclusion, this study provides valuable insights into the cost structure of integrating telemonitoring into established clinical pathways for older people with AH and T2D in primary care. To optimise the cost structure of telemonitoring, strategies like shortening the telemonitoring period, developing a national telemonitoring platform, using patient devices, integrating artificial intelligence into platforms, and involving nurse practitioners as telemedicine centre coordinators should be explored. Future research should build on these findings, testing new models and estimating the savings resulting from telemonitoring to provide evidence-based insights into the economic impact of telemonitoring in primary care.