Fig. 1

Schematic diagram to detect pre-outbreak signals of HFMD based on the L-DNM method. a The historical information of clinics hospitalization due to HFMD infection from January 1, 2009 to December 31, 2018, was collected from public records in Tokyo, Hokkaido and Osaka, Japan. b According to the DNM theory, the process of HFMD outbreaks is divided into three stages, including the normal stage, the pre-outbreak stage and the outbreak stage. The sudden increase in the DNM score indicates a transition from the normal stage to the pre-outbreak stage, i.e., the critical point before the upcoming outbreak of HFMD that results in an increase in clinical visits. c Based on the historical and current clinical records and geographic characteristics of a region, the DNM score provides an early warning signal of an upcoming outbreak of HFMD as a real-time indicator