Whilst the sensitivity and specificity of traditional laboratory-based surveillance can be readily estimated, the situation is less clear cut for syndromic surveillance. Syndromic surveillance indicators based upon presenting symptoms, chief complaints or preliminary diagnoses are designed to provide public health systems with support to detect multiple potential threats to public health. There is however, no gold standard list of all the possible ‘events’ that should have been detected. This is especially true in emergency response where systems are designed to detect possible events for which there is no directly comparable historical precedent.
Objective
To devise a methodology for validating the effectiveness of syndromic surveillance systems across a range of public health scenarios, even in the absence of historical example datasets.