Standard epidemiological methods to understand and improve Apis mellifera health

In this paper, we describe the use of epidemiological methods to understand and reduce honey bee morbidity and mortality. Essential terms are presented and defined and we also give examples for their use. Defining such terms as disease, population, sensitivity, and specificity, provides a framework for epidemiological comparisons. The term population, in particular, is quite complex for an organism like the honey bee because one can view "epidemiological unit" as individual bees, colonies, apiaries, or operations. The population of interest must, therefore, be clearly defined. Equations and explanations of how to calculate measures of disease rates in a population are provided. There are two types of study design; observational and experimental. The advantages and limitations of both are discussed. Approaches to calculate and interpret results are detailed. Methods for calculating epidemiological measures such as detection of rare events, associating exposure and disease (Odds Ratio and Relative Risk), and comparing prevalence and incidence are discussed. Naturally, for beekeepers, the adoption of any management system must have economic advantage. We present a means to determine the cost and benefit of the treatment in order determine its net benefit. Lastly, this paper presents a discussion of the use of Hill's criteria for inferring causal relationships. This framework for judging cause-effect relationships supports a repeatable and quantitative evaluation process at the population or landscape level. Hill's criteria disaggregate the different kinds of evidence, allowing the scientist to consider each type of evidence individually and objectively, using a quantitative scoring method for drawing conclusions. It is hoped that the epidemiological approach will be more broadly used to study and negate honey bee disease.