The lay abstract featured today (for Control charts for threshold correlated count data in disease infection number monitoring by Nannan Li, Cong Li and Jing Wan) is from Quality and Reliability Engineering International with the full article now available to read here.
, , . Control charts for threshold correlated count data in disease infection number monitoring. Qual Reliab Eng Int. 2024; 1–14. https://doi.org/10.1002/qre.3526
Abstract
Timely monitoring is essential for effective disease control and public health management. Early
outbreak detection and swift responses inform optimal resource allocation for effective public health protection. Disease incidence datasets often consist of time series counts, which refer to data that record the occurrences of certain events over time. These types of datasets are unique in that they often display threshold characteristics, meaning that there is a specific point at which the frequency of disease cases changes significantly, indicating a transition to a different state or phase. Currently, there is limited research on the monitoring of such data in the field of statistical process control. Control chart plays a crucial role in the realm of statistical process control. A effective control chart should accurately differentiate outbreak from non-outbreak periods, alerting promptly to outbreaks and avoiding false alarms. In this paper, we propose a new threshold-based control chart for threshold autoregressive models, which has been proven to be effective. The term ”autoregressive” simply means that the model uses historical data points as a basis for predicting future ones. Several existing efficient control charts are also employed for comparison. We conduct an extensive simulation study to assess the performance of our proposed chart. Finally, the method is applied to the meningitis data that motivated this investigation.
outbreak detection and swift responses inform optimal resource allocation for effective public health protection. Disease incidence datasets often consist of time series counts, which refer to data that record the occurrences of certain events over time. These types of datasets are unique in that they often display threshold characteristics, meaning that there is a specific point at which the frequency of disease cases changes significantly, indicating a transition to a different state or phase. Currently, there is limited research on the monitoring of such data in the field of statistical process control. Control chart plays a crucial role in the realm of statistical process control. A effective control chart should accurately differentiate outbreak from non-outbreak periods, alerting promptly to outbreaks and avoiding false alarms. In this paper, we propose a new threshold-based control chart for threshold autoregressive models, which has been proven to be effective. The term ”autoregressive” simply means that the model uses historical data points as a basis for predicting future ones. Several existing efficient control charts are also employed for comparison. We conduct an extensive simulation study to assess the performance of our proposed chart. Finally, the method is applied to the meningitis data that motivated this investigation.