Cost-Effectiveness of Remote Patient Monitoring for Hypertension Management Among Older Adults with Hypertension: A Semi-Markov Modeling Study

OBJECTIVES: To estimate incremental costs, incremental quality-adjusted life years (QALYs) and incremental cost-effectiveness ratio (ICER) of remote patient monitoring (RPM) for hypertension management among older adults.

METHODS: A semi-Markov model with monthly cycle length was developed using published models and our team's clinical judgment. The model had ten disease states: hypertension, myocardial infarction (MI), heart failure (HF), stroke, transient ischemic attack (TIA), other cardiovascular disease (Other), chronic kidney disease (CKD), chronic cardiovascular disease (Chronic CVD), advanced CKD, and death. All patients started in the hypertension state. Transition probabilities and hazard ratios were derived from analyses of Medicare fee-for-service (2018–2020) data. Healthcare costs (2022 dollars) and health utilities were obtained from published literature. Lifetime time horizon, 3% discounting, and the healthcare perspective were used. The model was estimated under three treatment scenarios: no RPM use, RPM use wherein RPM only directly affected event probabilities (RPM 1^st scenario), and RPM use wherein RPM affected event and mortality probabilities (RPM 2^nd scenario). Cost-effectiveness was determined at willingness-to-pay thresholds of $50,000, $100,000, and $150,000 per QALY. Probabilistic sensitivity analyses (PSA) were performed. R/RStudio was used.

RESULTS: The total costs (total QALYs) associated with no RPM, RPM 1^st scenario, and RPM 2^nd scenario were, respectively, $179,900.40 (6.09), $195,853.70 (6.12), and $217,770.50 (6.75). Relative to no RPM, the incremental cost, incremental QALY and ICER were, respectively: RPM 1^st scenario: $15,953.30, 0.035 and $450,483.43/QALY; RPM 2^nd scenario: $37,870.09, 0.661 and 57,253.77/QALY. The PSA found RPM use (1^st scenario) to be cost-effective 0% of the time at each of the three thresholds, while RPM use (2^nd scenario) was cost-effective in 32.5%, 99.5% and 100% of simulations at $50,000/QALY, $100,000/QALY and $150,000/QALY, respectively.

CONCLUSIONS: RPM use was cost-effective when RPM’s direct effects on mortality were included; however, it was not cost-effective under a conservative scenario wherein RPM only affected cardiovascular event probabilities.