Related Experiment Video
Updated: Jul 12, 2026
05:38
Rapid Molecular Detection and Differentiation of Influenza Viruses A and B
Published on: January 30, 2017
15.9K
Quickest Detection of COVID-19 Pandemic Onset
P Braca1, D Gaglione1, S Marano2
1NATO STO CMRE, Research Department, La Spezia, 19126, Italy.
IEEE Signal Processing Letters
|June 24, 2021
Summary
This study presents a simple quickest-detection test for detecting changes in data, useful for real-time analysis. Applied to COVID-19 data, it reliably signaled infection flare-ups early, aiding public health decisions.
Area of Science:
- Statistics
- Epidemiology
- Public Health
Background:
- Traditional quickest-detection tests may not suit complex scenarios with changing data.
- Real-time monitoring is crucial for managing public health crises like pandemics.
Purpose of the Study:
- To develop an accessible version of Page's CUSUM (Cumulative Sum) quickest-detection test.
- To adapt the test for composite hypotheses and time-varying data statistics.
- To evaluate its efficacy in early detection of infectious disease outbreaks.
Main Methods:
- Developed a recursive formulation of the decision statistic for efficient on-line analysis.
- Implemented an easily-implementable version of Page's CUSUM quickest-detection test.
- Validated the approach using publicly available COVID-19 epidemiological data.
Main Results:
- The developed test provides reliable early warnings of infection flare-ups.
- The early detection was sufficiently timely to inform decision-making on public health interventions.
- The recursive nature of the statistic facilitates continuous, real-time monitoring.
Conclusions:
- The enhanced Page's CUSUM test is a practical tool for on-line analysis of time-varying data.
- This method offers a valuable early warning system for infectious disease outbreaks.
- The findings support the use of this tool for timely public health policy decisions.
