A comparison of case-crossover and case-control designs in a study of risk factors for hemorrhagic fever with renal syndrome.

The case-crossover design is a new method for studying acute effects of transient exposures, in which cases serve as their own controls. To assess the validity, strengths, and weaknesses of the design, we used both the case-crossover method and the traditional case-control method of multivariate analysis with data collected from 196 cases and 295 hospital controls to study risk factors for hemorrhagic fever with renal syndrome in Korean soldiers. Both case-control and case-crossover methods showed that living in primitive field conditions, exposure to dust, and exposure to rodents were risk factors; use of insecticides or insect repellents was protective. Odds ratios from the case-crossover method were generally somewhat higher for risk factors and lower for protective factors. The case-crossover technique has great potential as a powerful, cost-effective way to assess risk factors for conditions such as infectious diseases and injuries, while avoiding some of the bias and logistical problems of traditional study designs.