Early warning of influenza epidemic based on influenza sentinel surveillance

Objective To explorer the method of early warning of influenza epidemic based on influenza sentinel surveillance. Methods The Influenza-like illness (ILI) proportion of total outpatients and the positive rate of Influenza viruses for every week of a year in sentinel hospital were calculated. The correlation between them were analyzed. The threshold of ILI proportion for Influenza epidemic early warning were calculated by using the positive rate in peak season when it passed 40% of the maximum seasonal level. Results The average ILI proportion of total outpatients in 2010 was 1.62% (95%CI: 1.43%-1.82%), and the average positive rate of influenza viruses was 26.77% (95%CI: 20.06%-33.49%). The positive rate threshold of influenza viruses during peak season was 32.31%, ILI proportion threshold of 1.73% was calculated for Influenza epidemic early warning. Conclusion Influenza epidemic early warning by using ILI proportion is simple, feasible, sensitive and specific. It is suitable for influenza epidemic early warning in local medical agencies conducting ILI sentinel surveillance.