Acute air pollution effects: consequences of exposure distribution and measurements.

Acute effect air pollution studies estimate the effect of short-term change in exposure on a health outcome. The two designs most commonly used in air pollution epidemiology are panel studies and time-series studies. Typically, both designs rely on ambient concentration measurements and not the personal exposures of individuals. This article discusses how panel studies and time-series studies are related and reviews the use of ambient concentrations versus personal exposure measurements in the analyses. This work suggests that for estimating acute effects, ambient concentration measurements are quite adequate in time-series studies. In addition, time-series studies have ample power relative to panel studies, in spite of the ecologic nature of their design. Panel studies have the benefit of being able to use all the information from personal exposures in the analysis, but they are much more costly and difficult to conduct. Furthermore, nontraditional panel studies, where multiple repeat panels are followed over time, require additional considerations in the analysis.