환경 오염원으로 인한 만성건강영향의 리버스 엔지니어링(역공학)

Often information on the exposed are unavailable when suspicions are raised on the development of chronic diseases from exposures to small area sources in the community or in the workplace. Without the appropriate information on the denominators of the risk rate, we tried to come up with testable case-only approaches to examine no effect assumptions of past exposures. With the assumptions of no-bias in the report of case and stability of population characteristics, we have examined testable hypothesis of no exposure effects from small area sources recursively based on previously known exposure scenarios. Two groups of casesubgroup approaches, one with standardized ratios of cases of Ever/Now Exposed and Ever/Now Unexposed, the other with odds ratio of two Independent Exposure Characteristics among diseased, are proposed in this study for the investigation of the differences in exposure subgroups on the disease outcome. As in the reverse engineering, where the cause of success or failures in the performance of engineered products is understood by going back from the finished products to the assembled parts, the case-subgroup analysis is to understand the potential causal role of exposure scenarios with only the data on cases available, by going back from the diseased to their presumed exposure sources based on known mechanisms.