This is a proposal to address statistical design and analysis issues related to occupational and environmental morbidity and mortality studies. In general, modest elevations in morbidity and mortality risk associated with exposure to particular toxic and hazardous substances found in the environment will be hard to detect in population or community-based studies. In some cases, epidemiological studies of occupationally exposed cohorts are extremely useful in quantifying environmental risks because workers are generally exposed to higher levels of toxic materials than the general population. These higher levels of exposure make it easier to detect and characterize modest elevations in morbidity and mortality risk associated to toxic and hazardous substances found in the workplace. However, occupational studies are also plagued with a form of selection bias known as the healthy worker effect (HWE) which can distort the shape of the exposure-response relationship and lead to bias towards the null hypothesis of no causal association between exposure and disease onset. Mr. Tchetgen will investigate statistical methods that can be used to consistently estimate the causal morbidity or mortality effect of occupational exposure while accounting for the HWE. In addition to methodologic work, Mr Tchetgen will have the opportunity to use these novel methods to re-analyze data collected in a morbidity and mortality study of occupational exposure to metalworking fluids in the automobile industry. Although some emphasis will be placed on the metalworking fluid case study, this research is most relevant to environmental health studies in general where risk assessment is of prime importance.
| Tchetgen, Eric J Tchetgen; Robins, James (2010) The semiparametric case-only estimator. Biometrics 66:1138-44 |
