ارزشیابی عملکرد الگوریتم میانگین متحرک وزن داده شده نمایی در شناسایی طغیان های آنفلونزا در ایران

Background: Timely response to influenza outbreaks using Influenza like illness (ILI) data is one of the most important priorities for public health authorities. The aim of this study was to evaluate the performance of the Exponentially Weighted Moving Average (EWMA) for timely detection of influenza outbreaks in Iran using simulated approaches from January 2010 to December 2015. Methods: Simulated influenza outbreaks were generated using ILI data as syndromic data of influenza according to potentially occurred outbreaks including uniform, linear and exponential distribution of corresponding epidemic curves. The performance of variety λ (smoothing parameter) values of EWMA was measured using sensitivity, specificity, false alarm rate, likelihood ratios and area under the receiver operating characteristics (ROC) curve. Results: The overall sensitivity of EWMA in detection of uniformly distributed influenza outbreaks was 70% (95%CI: 30, 80). The corresponding values for linear and exponential distribution of epidemic curves were 84% (95%CI: 75, 92) and 51% (95%CI: 40, 62). EWMA with λ=0.9 had the best performance for timely detection of influenza outbreaks in comparison to other smoothing parameters. Conclusion: Findings revealed that EWMA works well in detection of influenza outbreaks. However, national influenza surveillance systems need to use different outbreak detection methods for detecting aberrations in influenza-like illnesses activity.