Applying normalized p-control charts for early detection of local COVID-19 outbreaks
- 1School of Industrial & Intelligent Systems Engineering, The Iby and Aladar Fleischman Faculty of Engineering, Tel Aviv University, Tel Aviv, Israel.
- 2Department of Industrial & Information Systems Engineering, Faculty of Engineering, Bar-Ilan University, Ramat Gan, Israel.
Abstract
The COVID-19 pandemic underscored the urgent need for early detection of local outbreaks to enable better resource allocation and effective mitigation. Outbreak identification largely depended on local infection rates, which were highly variable and influenced by daily testing volumes. Effective pandemic management requires accurate, adaptive methods that reflect the dynamic and localized nature of these rates, ensuring timely and appropriate interventions. This study seeks to improve the early detection and localization of COVID-19 outbreaks through Statistical Process Control (SPC) and health data analytics. We propose an adaptive outbreak-detection framework that combines a normalized p-control chart with an Amplitude-based labeler (ABL). For each area and day, the monitored statistic is a standardized positivity-rate score, computed from the observed daily positivity rate, the number of tests performed that day, and an area-specific baseline positivity estimate derived from recent historical observations. This calibration is followed by Phase II sequential monitoring using fixed three-sigma limits on the normalized scale and Western Electric decision rules. The framework accommodates varying daily sample sizes, captures regional interconnectedness, and supports parameter tuning using a temporally ordered train/test design. Accounting for daily testing fluctuations and spatial variability, it enhances the accuracy of local outbreak detection. Compared with the benchmark outbreak-detection methods considered in this study, the proposed framework was evaluated on its ability to predict outbreak onset within the following three days, achieving overall accuracy of 0.81 and averaged precision and recall of 0.83 and 0.79, respectively. Performance was further evaluated using the Fβ-measure, enabling a flexible trade-off between precision and recall depending on the relative cost of errors. SPC-based methods like ours can significantly improve early outbreak detection and localization. In the post-vaccination era, when targeted responses are key to controlling localized spread while avoiding widespread restrictions, health data analytics offer a valuable tool for informed decision-making.
Related
- Effect of Measurement Errors on One‐Sided Synthetic‐RZ Control Charts for Monitoring the Ratio of Two Normal Variables
- Frayed Demazure weaves for Poisson-compatible cluster structures on Bott--Samelson charts
- Critique of "Use of roster charts in the investigation and prosecution of nurses ..." by John O' Quigley
- Weyl defects in skein theory and quantum cluster charts
- Approximation of Probability Geometries via Convex Families of Coordinate Charts
BibTeX
@article{foox2026applying,
title = {{Applying normalized p-control charts for early detection of local COVID-19 outbreaks}},
author = {Yuval Foox and Dan Yamin and Hila Chalutz-BenGal and Irad Ben-Gal},
journal = {PLOS One},
year = {2026},
month = oct,
doi = {10.1371/journal.pone.0357183},
url = {https://doi.org/10.1371/journal.pone.0357183},
note = {CC BY},
}
RIS
TY - JOUR
TI - Applying normalized p-control charts for early detection of local COVID-19 outbreaks
AU - Yuval Foox
AU - Dan Yamin
AU - Hila Chalutz-BenGal
AU - Irad Ben-Gal
PY - 2026
DA - 2026/10/20
JO - PLOS One
DO - 10.1371/journal.pone.0357183
UR - https://doi.org/10.1371/journal.pone.0357183
AB - The COVID-19 pandemic underscored the urgent need for early detection of local outbreaks to enable better resource allocation and effective mitigation. Outbreak identification largely depended on local infection rates, which were highly variable and influenced by daily testing volumes. Effective pandemic management requires accurate, adaptive methods that reflect the dynamic and localized nature of these rates, ensuring timely and appropriate interventions. This study seeks to improve the early detection and localization of COVID-19 outbreaks through Statistical Process Control (SPC) and health data analytics. We propose an adaptive outbreak-detection framework that combines a normalized p-control chart with an Amplitude-based labeler (ABL). For each area and day, the monitored statistic is a standardized positivity-rate score, computed from the observed daily positivity rate, the number of tests performed that day, and an area-specific baseline positivity estimate derived from recent historical observations. This calibration is followed by Phase II sequential monitoring using fixed three-sigma limits on the normalized scale and Western Electric decision rules. The framework accommodates varying daily sample sizes, captures regional interconnectedness, and supports parameter tuning using a temporally ordered train/test design. Accounting for daily testing fluctuations and spatial variability, it enhances the accuracy of local outbreak detection. Compared with the benchmark outbreak-detection methods considered in this study, the proposed framework was evaluated on its ability to predict outbreak onset within the following three days, achieving overall accuracy of 0.81 and averaged precision and recall of 0.83 and 0.79, respectively. Performance was further evaluated using the Fβ-measure, enabling a flexible trade-off between precision and recall depending on the relative cost of errors. SPC-based methods like ours can significantly improve early outbreak detection and localization. In the post-vaccination era, when targeted responses are key to controlling localized spread while avoiding widespread restrictions, health data analytics offer a valuable tool for informed decision-making.
SN - 1932-6203
ER -