Abstract:
|
We propose a case cohort design for longitudinal data as a cost-effective way to study the effects of covariates on repeated observations of rare binary outcomes. Our design entails choosing a random sub-cohort at the beginning of the study and following them repeatedly over time. Although some members in this sub-cohort may experience events over the study period, we refer to it as the "control cohort." The "case cohort," on the other hand, is a random sample of subjects not in the control cohort, who have experienced at least one event during the study period. This design is an extension of the case-cohort design (Prentice, 1986) and the bidirectional case-crossover design (Navidi, 1998). We illustrate it using data from a study of childhood asthma. The purpose of the study is to assess the role of immune function in impacting asthma risk directly, and in modifying the risk associated with various environmental exposures. Immune function is generally assessed through biomarkers. As the evaluation of biomarkers is expensive and labor-intensive, a cost effective sub-sampling strategy is very useful. The efficiency of our design is compared to a full cohort analysis.
|