Objective: Low-dose aspirin is standard care in patients with a history of cardiovascular disease (CVD). But, the use of low-dose aspirin in primary prevention has not yet been fully established in Japan although meta-analyses and US/European guidelines support its use in persons at increased CVD risk. This study assessed the health economic consequences of the use of low-dose aspirin in the primary prevention of CVD in Japan.
Patients and Methods: Based on results reported in two recent meta-analyses of Hayden (2002) and Eidelman (2003), a Markov model was constructed to predict the cost-effectiveness of low-dose aspirin in the primary prevention of CVD. The model consists of 5 health states: 1) no history of CVD, 2) history of stroke, 3) history of myocardial infarction, 4) history of CVD, and 5) death, with a 10-year time horizon and 1-year cycles. Direct costs from the insurers' perspective were used, while health outcome was expressed in Life-Years Gained (LYG). 'Discounting Rate' with 3% was applied on effectiveness and costs.
Results: For patients with a 1-year risk of coronary heart disease (CHD) of 1.5% (10-year risk of ±15%), the model demonstrated 'dominance' of the 'aspirin' arm versus 'no aspirin' arm; the 10-year costs were Japanese Yen (JPY) 634,000 (Euro 4,857) and JPY 518,000 (Euro 3,968) in the 'no aspirin' arm and 'aspirin' arm, respectively, while LYG was 8.33 and 8.36, respectively. Low-dose aspirin treatment saved on average JPY 116,000 (Euro 889) [95% confidence interval (CI) JPY 57,077-175,151] per patient. Dominance was demonstrated (non-significant) in the first year of treatment and, low-dose aspirin was dominant to 'no aspirin' arm from an annual risk of 0.20%. Other results of sensitivity analysis on gastrointestinal (GI) bleeding rate, stroke rate, cost of each event and discounting showed the robustness of the results.
Conclusions: Administering low-dose aspirin to patients with a 1-year risk of CHD of 1.5% and more is significantly costsaving from the insurers' perspective in Japan.
View full abstract