You are here:
Technical Manual for MCestimate
Citation:
Etterson, M. Technical Manual for MCestimate. U.S. Environmental Protection Agency, Washington, DC, EPA/600/B-13/164, 2013.
Impact/Purpose:
MCnestimate is a tool for estimating probabilities of failure in the presence of competing risks. It was originally developed to facilitate the study of causes of avian nest-failure. However, competing risks arise in virtually all stochastic processes and MCnestimate can be applied in many contexts outside of avian nest failure.
Description:
The analysis of competing risks has a long history in human health research, but has received limited formal attention in wildlife ecology and ecotoxicology (Heisey and Patterson 2006). Informally, however, the concept of competing risks has arisen many times in these disciplines (Abbott 1925, Ricklefs 1969). A common and accessible motivating example for defining competing risks is the concept of cause-of-death. Thus, when considering the possibility that an animal will suffer mortality in either acute or chronic response to chemical exposure, one must also consider the possibility that the animal will die due to disease, predation, or other causes, the occurrence of which will alter the rate at which animals succumb to exposure to a toxicant. Although such cause-specific failure analyses remain an important application, competing risks naturally arise in virtually all stochastic processes. The model described under this QAPP provides a unified common framework for analysis of data arising from serial observations on subjects in the presence of competing risks. The three examples below illustrate some the diverse contexts in which such data arise.