COST-EFFECTIVENESS ANALYSIS OF AI-BASED 12-LEAD ELECTROCARDIOGRAM SCREENING FOR LEFT VENTRICULAR SYSTOLIC DYSFUNCTION IN SOUTH KOREA
Author(s)
Tae Jin Lee, PhD1, Namhee Kim, MPH2, yunjeong kwon, MPH3, YOUNGSIL LEE, Ph.D.1.
1Seoul National University, Seoul, Korea, Republic of, 2Seoul National Univeristy, Seoul, Korea, Republic of, 3Seoul National University, seoul, Korea, Republic of.
1Seoul National University, Seoul, Korea, Republic of, 2Seoul National Univeristy, Seoul, Korea, Republic of, 3Seoul National University, seoul, Korea, Republic of.
OBJECTIVES: Artificial intelligence-based electrocardiography (AI-ECG) uses standard 12-lead electrocardiogram data to identify individuals at risk of asymptomatic left ventricular dysfunction (ALVD) with greater diagnostic accuracy than conventional ECG. This study evaluated the cost-effectiveness of an AI-ECG-based screening strategy for ALVD in the Korean healthcare setting.
METHODS: A cost-utility analysis was conducted from the healthcare system perspective for asymptomatic adults aged ≥40 years undergoing routine health screening. The comparator reflected current clinical practice, in which all individuals receive a standard 12-lead ECG and 25% undergo transthoracic echocardiography (TTE) based on clinical judgment. A lifetime decision tree-Markov model with annual cycles was developed, including five health states: No ALVD, Untreated ALVD, Treated ALVD, Symptomatic Heart Failure, and Death. AI-ECG diagnostic performance was based on a Korean study (sensitivity 90.6%, specificity 99.4%). Treatment effectiveness and health utilities were obtained from systematic reviews and meta-analyses. Costs were estimated using National Health Insurance reimbursement schedules, healthcare utilization data, and expert consultation. One-way and probabilistic sensitivity analyses were performed to assess parameter uncertainty.
RESULTS: The AI-ECG strategy yielded lower lifetime costs than conventional screening (KRW 301,430 vs. KRW 394,185) and greater health benefits (18.9874 vs. 18.9848 QALYs), indicating dominance. Cost savings and QALY gains were consistent across age and sex subgroups. The model was most sensitive to the probability of heart failure progression in untreated ALVD, mortality risk in untreated ALVD, and progression in treated ALVD. In probabilistic sensitivity analysis (10,000 simulations), AI-ECG was dominant in 80.8% of iterations.
CONCLUSIONS: AI-ECG-based screening for ALVD among asymptomatic adults aged ≥40 years participating in health screening is a cost-saving strategy that improves health outcomes while reducing healthcare costs compared with conventional screening. Despite reliance on some international data due to limited Korean evidence, this study provides the first economic evaluation of AI-ECG-based ALVD screening reflecting real-world Korean health screening practice.
METHODS: A cost-utility analysis was conducted from the healthcare system perspective for asymptomatic adults aged ≥40 years undergoing routine health screening. The comparator reflected current clinical practice, in which all individuals receive a standard 12-lead ECG and 25% undergo transthoracic echocardiography (TTE) based on clinical judgment. A lifetime decision tree-Markov model with annual cycles was developed, including five health states: No ALVD, Untreated ALVD, Treated ALVD, Symptomatic Heart Failure, and Death. AI-ECG diagnostic performance was based on a Korean study (sensitivity 90.6%, specificity 99.4%). Treatment effectiveness and health utilities were obtained from systematic reviews and meta-analyses. Costs were estimated using National Health Insurance reimbursement schedules, healthcare utilization data, and expert consultation. One-way and probabilistic sensitivity analyses were performed to assess parameter uncertainty.
RESULTS: The AI-ECG strategy yielded lower lifetime costs than conventional screening (KRW 301,430 vs. KRW 394,185) and greater health benefits (18.9874 vs. 18.9848 QALYs), indicating dominance. Cost savings and QALY gains were consistent across age and sex subgroups. The model was most sensitive to the probability of heart failure progression in untreated ALVD, mortality risk in untreated ALVD, and progression in treated ALVD. In probabilistic sensitivity analysis (10,000 simulations), AI-ECG was dominant in 80.8% of iterations.
CONCLUSIONS: AI-ECG-based screening for ALVD among asymptomatic adults aged ≥40 years participating in health screening is a cost-saving strategy that improves health outcomes while reducing healthcare costs compared with conventional screening. Despite reliance on some international data due to limited Korean evidence, this study provides the first economic evaluation of AI-ECG-based ALVD screening reflecting real-world Korean health screening practice.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
EE166
Topic
Economic Evaluation, Medical Technologies
Disease
Cardiovascular Disorders (including MI, Stroke, Circulatory)