Massachusetts researchers recently conducted a study to design, implement and test a flexible methodology to generate surveillance data that could provide a realistic temporal clustering of disease cases to evaluate protocols for outbreak detection.
The team of researchers constructed a detailed representation of the Boston area based on data about locations, activity patterns and individuals. The goal of the study, which was published on Wednesday in the journal BMC Medical Informatics and Decision Making, was to evaluate gaps in the systematic evaluation of outbreak detection protocols, BioMedCentral reports.
The study used the simulation of influenza-like illness transmission to produce 100 years of in silico ILI data. The team tested six different surveillance systems for detecting cases from the simulated disease data. Performance was measured by inserting test outbreaks into the streams of the surveillance to analyze the timeliness and likelihood of detection.
The detection of outbreaks ranged from 21 to 95 percent, with increased coverage failing to linearly improve detection probability for all the systems. While relaxing the decision threshold for signaling outbreaks increased false positives significantly, it also led to earlier outbreak detection and improved overall outbreak detection, according to BioMedCentral reports.
The authors concluded that geographical distribution may be more important than coverage level and that detailed simulations of disease transmission can be configured to represent almost all conceivable scenarios.
According to the study, the simulations are a powerful tool to evaluate surveillance system performance and other methods used to detect outbreaks.