A systematic review of cluster detection mechanisms in syndromic surveillance: Towards developing a framework of cluster detection mechanisms for EDMON system

Abstract

Time lag in detecting disease outbreaks remains a threat to global health security. Currently, our research team is working towards a system called EDMON, which uses blood glucose level and other supporting parameters from people with type 1 diabetes, as indicator variables for outbreak detection. Therefore, this paper aims to pinpoint the state of the art cluster detection mechanism towards developing an efficient framework to be used in EDMON and other similar syndromic surveillance systems. Various challenges such as user mobility, privacy and confidentiality, geographical location estimation and other factors have been considered. To this end, we conducted a systematic review exploring different online scholarly databases. Considering peer reviewed journals and articles, literatures search was conducted between January and March 2018. Relevant literatures were identified using the title, keywords, and abstracts as a preliminary filter with the inclusion criteria and a full text review were done for literatures that were found to be relevant. A total of 28 articles were included in the study. The result indicates that various clustering and aberration detection algorithms have been developed and tested up to the task. In this regard, privacy preserving policies and high computational power requirement were found challenging since it restrict usage of specific locations for syndromic surveillance.