P20 Health Technology Assessment Based on Real-World Data: Estimating Lifetime Survival and Costs With Empirical Examples from Taiwan

Analyzing RWD, aim to estimate (1) how many QALY or LY (life year) and lifetime medical costs (LMC) could be saved from effective prevention of catastrophic illnesses, including cancer, and (2) how much would be spent per QALY/LY in diagnosis and treatment of them. Inter-linkages Taiwan’s different databases, mortality registry, cancer registry, reimbursement data of National Health Insurance, we established longitudinal cohorts of different catastrophic illnesses and estimated the LE by extrapolating the survival after the end of follow-up using simulated age-, sex-, and calendar year-matched referents from national lifetables and a rolling-over algorithm of constructed restricted cubic spline model based on logit transformed of survival ratio between the index and referents cohorts. We multiplied the second functions (QoL, LMC) with the lifetime survival function and summed up for QALE and LMC after adjustment for consumer price index of each year and annual discount rate for extrapolating years. The results showed that effective prevention would save 12.1-13.5 QALY and 300000 USD for patients under maintenance dialysis, followed by spinal cord injury (13-14 life-years & 100000 USD) and cancer of different organ-systems (2.7-18 QALYs & 26.5-97.3 USD). The cost-effectiveness ratios of healthcare services for comatose patients under prolonged mechanical ventilation, dialysis, spinal cord injury, major trauma, and cancers were $100,000, 16.2-20.0, 5.5-5.9, 3.4-3.6, and 2.9-11.9 USD per QALY or LY. Instead of direct comparison of QALE or LE, we propose to quantify difference-in-differences obtained from comparison of loss-of QALE or loss-of-LE to control confounding from different distributions of age, sex, and calendar year of medical technology advancement while evaluating the effectiveness of two technologies. HTA using RWD is feasible after quantification of lifetime survival plus another second function. Extending the evaluation to include societal impact would further improve justice and efficiency of resources distribution and the sustainability of universal coverage system.