Clinical Report: Leveraging Community Syndromic Surveillance to Predict School Closures Due to Enterovirus
Overview
This retrospective study developed a district-level long short-term memory (LSTM) framework intended to provide early warning of abnormal enterovirus-like (EV-like) syndromic activity and associated class suspension events in Taipei. The model used routine pediatric outpatient data and district-specific adaptive thresholds to generate alarms generally earlier than abnormal class suspension days.
Background
Enterovirus infections cause recurrent pediatric epidemics and commonly present as hand-foot-and-mouth disease or herpangina. Taiwan typically experiences seasonal peaks during May to June and September to October, although increasing variability and year-round activity have been observed. Evidence cited by the authors indicated that infections declined during the COVID-19 pandemic and subsequently rebounded sharply. Traditional approaches using fixed thresholds may not adequately capture evolving or nonlinear disease patterns.
Data Highlights
Participating clinics: 110
Taipei City districts: 12
Surveillance period: January 2022 to January 2025
Preschool age group: 0–6 years
Primary school age group: 7–12 years
Model lookback window: 30 days
Forecast horizon: 14 days
Preschool-data accuracy: 0.94
Primary-school-data accuracy: 0.96
Postpandemic increase cited as background: 619.5%
Key Findings
Evidence cited by the study indicated that enterovirus infections in Taiwan decreased by 67.1% during the COVID-19 pandemic and subsequently increased by 619.5%.
Adaptive residual thresholds were used to identify unexpected increases relative to each district’s modeled baseline.
Model-generated alarms frequently preceded periods with abnormal class suspensions.
Forecasting performance was stronger in districts with relatively stable activity and more difficult where epidemic waves were larger or more variable.
The framework also achieved average accuracies of 0.87 for herpangina and 0.97 for hand-foot-and-mouth disease among preschool children.
Clinical Implications
Combining pediatric outpatient syndromic data with school suspension records may support earlier identification of districts entering periods of elevated enterovirus activity. Model-generated alarms could inform local preparedness and targeted public health responses, but they do not confirm outbreaks or predict suspensions at individual schools or classes. The framework requires validation in other settings and with additional predictors before broader implementation.
Conclusion
District-level LSTM forecasting with adaptive residual thresholds captured local patterns in EV-like syndromic activity and generally provided warning before abnormal class suspension days. The findings support further evaluation of this approach as a scalable surveillance tool for timely responses, targeted class suspension decisions, and preparedness among schools and families.
Related Resources & Content
Using Community Syndromic Surveillance to Anticipate Enterovirus-Related School Class Suspensions: A Real-Time LSTM Approach — Rayguru C, Jian HL, Peng YF, et al. International Journal of Infectious Diseases. 2026;170:108918. doi:10.1016/j.ijid.2026.108918.
Aggregated EV-Like Illness Data Analyzed in the Study — Available through Figshare. doi:10.6084/m9.figshare.32017062.
Software Used for the Study’s LSTM Surveillance Framework — Available through Figshare. doi:10.6084/m9.figshare.32006418.
Federal prosecutors allege that a Florida physician and research staff fabricated clinical trial records that were submitted into database systems used to evaluate investigational drugs.
