Objective: To illustrate the efficiency of cumulative sum (CUSUM) in pre-warning of the influenza peak in Beijing.
Methods: CUSUM was used to analyze the data of influenza like illness (ILI), and the results of the influenza laboratory surveillance was regarded as the gold standard to judge the approaching of the influenza peak.
Results: The surveillance was launched in 421 hospitals in Beijing during the 2009 to 2010 influenza season, while the influenza laboratory surveillance was launched by 7 collaborative laboratories. From Jun. 2009 to Apr. 2010, the average ILI percentage in the 421 hospitals was 2.56%. In the study, 19 262 pharyngeal swab samples were collected from the ILI cases in 11 hospitals and 5 045 of them were tested positive for the influenza virus, with the novel swine-origin influenza A H1N1 virus dominating. After analyzing of the ILI surveillance data with CUSUM, it was found that the ILI surveillance in Beijing could make a satisfactory early warning for the approaching of the influenza peak referring to the gold standard based on the influenza laboratory results.
Conclusion: It could give the prediction and early warning for the influenza peak efficiently and precisely, by using CUSUM to analyze the influenza surveillance data of Beijing.