Counterpoint: moving from potential-outcomes thinking to doing--changing research planning to enable successful...

The potential-outcomes framework is an appealing new approach that imposes a degree of formal conceptual modeling beyond traditional epidemiologic methods for assessing associations between air pollution and health. However, it introduces a number of additional factors to consider when selecting intervention and especially control conditions that call for forward-thinking research designs. We propose that researchers seeking to implement the potential-outcomes framework consider the use of prospective designs that provide more opportunities to establish well-defined intervention and control populations and determine causal relationships between air quality and health. In implementing these prospective research designs, collaboration between researchers and those who implement the interventions can improve the understanding of how a planned intervention actually occurs, thereby improving the characterization of emissions and air quality responses to the intervention. By looking ahead, epidemiologists can take advantage of upcoming regulatory interventions to design successful health outcomes research programs.