Ecological methods investigate group level aggregate data. These methods are motivated mainly due to the relative low cost and convenience of the data, limited availability of individual-level data, design limitations of individual studies, interest in ecologic effect, and simplicity of analysis and presentation (Morgenstern 1995). For very rare events (e.g., death by suicide) that occur at rates on the order of 1 in 10,000, there may be few options for routine drug surveillance. One approach is to use ecological data that attempt to relate changes in prescription rates of particular drugs or classes of drugs to the AE rate of interest. These more global associations clearly do not support causal inferences, but can be hypothesis generating and help support inferences drawn from other studies, in some cases based on surrogate endpoints (e.g., suicide attempts and/or ideation). In some cases, natural experiments such as a black box warning provide an opportunity to evaluate the positive or negative consequences of decreased access to the drug on the event of interest. Here we may compare national rates of the AE before and after the public health warning, to determine if the warning has had the anticipated effect. If the warning is specific to a strata of the population, comparison of changes in that strata versus those for which the warning did not apply provides stronger inferences.