Evaluation of a Heart Failure Telemonitoring Program Through a Microsimulation Model: Cost-Utility Analysis

Abstract

Background Heart failure (HF) is a major public health issue in Canada that is associated with high prevalence, morbidity, and mortality rates and high financial and social burdens. Telemonitoring (TM) has been shown to improve all-cause mortality and hospitalization rates in patients with HF. The Medly program is a TM intervention integrated as standard of care at a large Canadian academic hospital for ambulatory patients with HF that has been found to improve patient outcomes. However, the cost-effectiveness of the Medly program is yet to be determined. Objective This study aims to conduct a cost-utility analysis of the Medly program compared with the standard of care for HF in Ontario, Canada, from the perspective of the public health care payer. Methods Using a microsimulation model, individual patient data were simulated over a 25-year time horizon to compare the costs and quality-adjusted life years (QALYs) between the Medly program and standard care for patients with HF treated in the ambulatory care setting. Data were sourced from a Medly Program Evaluation study and literature to inform model parameters, such as Medly’s effectiveness in reducing mortality and hospitalizations, health care and intervention costs, and model transition probabilities. Scenario analyses were conducted in relation to HF severity and TM deployment models. One-way deterministic effectiveness analysis and probabilistic sensitivity analysis were performed to explore the impact on the results of uncertainty in model parameters. Results The Medly program was associated with an average total cost of Can $102,508 (US $77,626) per patient and total QALYs of 5.51 per patient compared with the average cost of Can $97,497 (US $73,831) and QALYs of 4.95 per patient in the Standard Care Group. This led to an incremental cost of Can $5011 (US $3794) and incremental QALY of 0.566, resulting in an incremental cost-effectiveness ratio of Can $8850 (US $6701)/QALY. Cost-effectiveness improved in relation to patients with advanced HF and with deployment models in which patients used their own equipment. Baseline and alternative scenarios consistently showed probabilities of cost-effectiveness greater than 85% at a willingness-to-pay threshold of Can $50,000 (US $37,718). Although the results showed some sensitivity to assumptions about effectiveness parameters, the intervention was found to remain cost-effective. Conclusions The Medly program for patients with HF is cost-effective compared with standard care using commonly reported willingness-to-pay thresholds. This study provides evidence for decision makers on the use of TM for HF, supports the use of a nurse-led model of TM that embeds clinically validated algorithms, and informs the use of economic modeling for future evaluations of early-stage health informatics technology.

Cost-effectiveness of a Telemonitoring Program for Patients With Heart Failure During the COVID-19 Pandemic in Hong Kong: Model Development and Data Analysis

“…We further examined the impact of the monthly cost of the telemonitoring system in an extended one-way sensitivity analysis, and we found that telemonitoring-mediated care remained highly cost-effective if the monthly cost of telemonitoring was below US $467. Our findings were consistent with a cost-utility study of a telemonitoring-mediated HF care system in Canada in that the telemonitoring strategy was highly acceptable to be cost-effective, with an ICER of US $6701/QALY (WTP threshold=US $37,718/QALY) [ 16 ]. Our study further evaluated the interacting impact of two key parameters (risk ratios of events with telemonitoring vs without telemonitoring), which represented the relative effectiveness of telemonitoring in lowering HF-related hospitalization and all-cause mortality, on the cost-effective acceptance of the telemonitoring strategy.…”

“…The implementation cost is a modifiable factor when introducing a new technology in a health care system. In this study, telemonitoring was assumed to have a monthly cost of US $50 based on the estimated cost of a currently available smartphone-based telemonitoring system in Canada [ 16 , 18 ]. We further examined the impact of the monthly cost of the telemonitoring system in an extended one-way sensitivity analysis, and we found that telemonitoring-mediated care remained highly cost-effective if the monthly cost of telemonitoring was below US $467.…”

“…The monthly outpatient cost was estimated from the findings of a retrospective observational study on the total management cost (including hospitalization cost and ambulatory care cost) of patients with HF (n=73) recruited from a public hospital in Hong Kong [ 27 ]. The implementation cost of telemonitoring per capita (US $80) and monthly cost of telemonitoring (US $50) were approximated from the reported costs of a smartphone-based telemonitoring system [ 16 ], including a smartphone, blood pressure monitor, weight scale, and licensing fee. The implementation cost was a one-time charge, while the monthly cost of telemonitoring was a recurrent cost for maintenance of the app.…”

“…Markov modeling is recommended for evaluating the outcomes of diseases that might progress, improve, or relapse through transition over a series of health states [ 14 ]. The cost-effective application of telemonitoring for the management of HF was demonstrated by Markov model–based analyses prior to the era of COVID-19 [ 15 , 16 ], and the patients’ medical avoidance was therefore not evaluated as an influential factor. In this study, COVID-related medical avoidance was considered in the model-based analysis.…”

A Divided Old Age through Research on Digital Technologies

“…Similarly, other studies found that participants reported higher need of support compared to decliners (Flink, Brandberg and Ekstedt, 2019;Foster et al, 2015). In other studies, study participants reported to enjoy engaging in activities that could contribute to science and help other people and society (Bongartz et al, 2017;Coley et al, 2018).…”