Dust storms frequently occur in El Paso, Texas, but little is known about their respiratory health effects. The association between dust storm and hospitalization for respiratory illness will be studied through a case-crossover study design. The goal is to ascertain the effect of dust storms, adjusted for weather and other pollutants, and to explore the potential harmfulness for different subgroups of residents. A case-crossover study allows patients to serve as their own controls, in order to reduce confounding. The history, relationship, and application of the case-control and case-crossover study designs are reviewed in this thesis. The estimating equations for conditional logistic regression are derived in order to obtain the relative risk for evaluating the impact of dust storms in this case-crossover study design. Primary findings are reported. This thesis contributes to filling the gap in scientific research about the dust storms's effect on respiratory health.
KEYWORDS: Dust storms, case-control, case-crossover, respiratory health.
