Stochastic approach for sufficient component cause interaction applies to longitudinal studies with right censoring

• Main Authors: Lin, Kuan-I, 林冠逸
• Other Authors: Lin, Sheng-Hsuan
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/8ukdcu

碩士 === 國立交通大學 === 統計學研究所 === 107 === In epidemiologic area, mechanistic interactions between exposures and diseases is one of the most critical issues. Previous literatures have proposed a stochastic sufficient component cause (SCC) framework. In this theory, it treats each sufficient cause as a stochastic process instead of a time-invariant random variable. However, the mechanistic interactions such as synergism and agonism can not be identified and estimated. Therefore, in this study, we combine stochastic SCC and marginal SCC as stochastic mSCC and additional assumptions. By this idea, we can identify synergism and agonism. We further provided six approaches to identify and estimate synergism and agonism based on an additive hazard model and complementary log model. In addition, the confounder is easily adjusted by appropriate covariates into a regression model. In our proposed three models, the simulations have proven their valid test. That is, their type I error rate α=0.05. The power of additive hazard model increases as the follow-up time increases and it is higher than other two models. Next, we applies this method to a Taiwanese cohort dataset to investigate the mechanistic interaction among hepatitis B and C viruses on the incidence of hepatocellular carcinoma. The hazard of people with agonism is 1.28×10^(-5) (95% CI∶6.97×10^(-6)-1.87×10^(-5)), and the cumulative hazard of people is 7.41×10^(-2) (95% CI∶4.09×10^(-2)-1.07×10^(-1) ), which is approximately 3.5 times stronger than that of synergism. We make it possible to quantify synergism and agonism in right-censored data.