Detecting Outbreaks of Ross River Virus Disease Using Automated Monitoring
Author Information
Author(s): Rochelle E Watkins, Serryn Eagleson, Bert Veenendaal, Graeme Wright, Aileen J Plant
Primary Institution: Curtin University of Technology
Hypothesis
This study investigates the correspondence between retrospective epidemiological evaluation of notifications of Ross River virus disease and the signals produced by cumulative sum-based automated monitoring methods.
Conclusion
The negative binomial cusum provides greater sensitivity for detecting outbreaks of Ross River virus disease compared to EARS algorithms, although it may be less timely.
Supporting Evidence
- The negative binomial cusum had a greater area under the ROC curve compared to EARS algorithms.
- Major outbreaks of RRv disease occurred approximately every four years between 1991 and 2004.
- Case notifications often remain elevated for 6 months following seasonal outbreaks.
Takeaway
The study looked at how well different computer methods can find outbreaks of a disease called Ross River virus. One method worked better at spotting outbreaks than the others.
Methodology
The study compared the performance of EARS algorithms and a negative binomial cusum using historical disease notification data.
Potential Biases
Potential bias due to the subjective nature of expert evaluations and the retrospective design.
Limitations
The study relied on retrospective data and expert opinion, which may not capture all outbreaks accurately.
Participant Demographics
The study focused on Ross River virus disease notifications in Western Australia, primarily affecting middle-aged adults.
Statistical Information
P-Value
0.0002
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website