Evaluation of the Japanese school health surveillance system for influenza.

In order to evaluate the Japanese nationwide school absenteeism surveillance system for pediatric influenza in comparison with the national sentinel surveillance for influenza, we used surveillance guidelines (Centers for Disease Control and Prevention, 1998) to determine the efficacy of the school health surveillance system (SHSS). Data regarding school absenteeism (age 4-15 years old) was compared with data regarding influenza-like illness (ILI) per sentinel sites during the second to the 11th weeks of 1998 and 1999. Despite the system's high simplicity and acceptability, telecommunication costs were estimated at US$ 490,000 (1998). Representativeness of schoolchildren was very accurate, but ILI for pre-school children (4-6 years) remained uncountable. Sensitivity, specificity, and positive predictive value of the SHSS compared to sentinel surveillance were calculated as 80%, 100%, and 100%, respectively (P=0.004). Although the SHSS was found to provide accurate surveillance data during periods of high influenza activity, non-influenza virus infections (e.g., adenovirus, rotavirus, and Norwalk virus, etc.) may become mixed in the SHSS data. Evaluation using this system should be continued employing a new case definition excluding gastrointestinal symptoms.