Project 171914

In recent years, many public health agencies have started to operate automated surveillance systems. Due to the large volume of data, surveillance analysts are not able to review all data manually. Instead, they use statistical algorithms that screen the data automatically and issue alerts to draw an epidemiologist's attention to potential disease outbreaks. It is well recognized that not all of these algorithms perform equally well when it comes to detecting outbreaks quickly and accurately. Moreover, the same algorithm may perform better or worse under different conditions. To ensure the effectiveness of automated surveillance systems, it is important to understand the factors that determine the performance of detection algorithms and to select the best algorithm for a given context. The goal of this project is to evaluate the performance of different outbreak detection algorithms, by experimenting with different algorithms and different types of surveillance data and then analyzing the results of these experiments. This knowledge will serve as an information resource for public health practitioners as they select algorithms for different surveillance systems or develop new algorithms.