Cost-effectiveness of digital therapeutics for essential hypertension

Abstract

Hypertension increases the risk of cardiovascular and other diseases. Lifestyle modification is a significant component of nonpharmacological treatments for hypertension. We previously reported the clinical efficacy of digital therapeutics (DTx) in the HERB-DH1 trial. However, there is still a lack of cost-effectiveness assessments evaluating the impact of prescription DTx. This study aimed to analyze the cost-effectiveness of using prescription DTx in treating hypertension. We developed a monthly cycle Markov model and conducted Monte Carlo simulations using the HERB-DH1 trial data to investigate quality-adjusted life-years (QALYs) and the cost of DTx for hypertension plus guideline-based lifestyle modification consultation treatment as usual (TAU), comparing DTx + TAU and TAU-only groups with a lifetime horizon. The model inputs were obtained from the HERB-DH1 trial, published or publicly available data, and expert assumptions. The incremental cost-effectiveness ratio (ICER) per QALY was used as the benchmark for cost-effectiveness. We performed probabilistic sensitivity analyses (PSAs) using the Monte Carlo simulation with two million sets. The DTx + TAU strategy produced 18.778 QALYs and was associated with ¥3,924,075 ($34,122) expected costs, compared with 18.686 QALYs and ¥3,813,358 ($33,160) generated by the TAU-only strategy over a lifetime horizon, resulting in an ICER of ¥1,199,880 ($10,434)/QALY gained for DTx + TAU. The monthly cost and attrition rate of DTx for hypertension have a significant impact on ICERs. In the PSA, the probability of the DTx arm being a cost-effective option was 87.8% at a threshold value of ¥5 million ($43,478)/QALY gained. In conclusion, the DTx + TAU strategy was more cost-effective than the TAU-only strategy.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 54 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.