Laboratory tests for influenza and the respiratory syncytial virus may be used as a proxy for associated hospitalizations in Canada

Abstract In Canada, the national number of hospital admissions associated with influenza and respiratory syncytial virus (RSV) is usually available about three months after admission. This delay hampers real-time analysis involving hospitalization data, like, for example, epidemic forecasting. Here, using a mixed-effects model on 15 years of data covering about 70% of the Canadian population, we show that these hospitalizations can be approximated using the number of laboratory tests positive for influenza A and RSV, a data stream publicly reported much more rapidly, about two weeks after symptoms onset.