Statistics Seminar: Zeny Feng - ATQ: Alarm time quality, an evaluation metric for assessing timely epidemic detection models within a school absenteeism-based surveillance system
Title: ATQ: Alarm time quality, an evaluation metric for assessing timely epidemic detection models within a school absenteeism-based surveillance system
Speaker: Zeny Feng, University of Guelph
Abstract: Wellington-Dufferin-Guelph Public Health (WDGPH) has conducted an absenteeism-based influenza surveillance program in the WDG region of Ontario, Canada since 2008, using a 10% absenteeism threshold to raise an alarm for the implementation of mitigating measures. A recent study indicated that model-based alternatives, such as distributed lag seasonal logistic regression models,provided improved alarms for detecting an upcoming epidemic. However model evaluation and selection was primarily based on alarm accuracy, measured by the false alarm rate (FAR), and failed to optimize timeliness. Here, a new metric that simultaneously evaluates epidemic alarm accuracy and timeliness is proposed. The ATQ assessed alarms on a gradient, where alarms raised incrementally before or after an optimal day were considered informative, but were penalized for lack of timeliness. Summary statistics of ATQ,average alarm time quality (AATQ) and first alarm time quality (FATQ), were used for model evaluation and selection. Alarms raised by ATQ and FAR selected models were compared. A simulation study that mimics the WDG population and influenza demographics was conducted for further evaluation of the proposed metric. Our proposed method was applied to the daily elementary school absenteeism and laboratory-confirmed influenza case data collected by WDGPH.
Date/Time: Tuesday November 8, 2022, 3:30 - 5:00
Location: UH 112