DODGE: automated point source bacterial outbreak detection using cumulative long term genomic surveillance

Area of Science:

• Genomic epidemiology
• Public health surveillance
• Foodborne disease outbreak detection

Background:

• Accurate and timely recognition of foodborne disease outbreaks is crucial for public health surveillance.
• Whole genome sequencing (WGS) provides high-resolution typing for bacterial pathogens, aiding outbreak detection.
• Current methods lack dynamic tools for selecting and adjusting genetic thresholds for WGS data analysis.

Purpose of the Study:

• To introduce DODGE (Dynamic Outbreak Detection for Genomic Epidemiology), an algorithm for dynamically selecting and comparing genetic thresholds in WGS data.
• To enable integrated analysis across jurisdictions by naming predicted outbreak clusters using established genomic nomenclature.
• To assess DODGE's performance in analyzing expanding datasets over time.

Main Methods:

• Development of the DODGE algorithm for dynamic threshold selection and comparison.
• Application of DODGE to two real-world Salmonella WGS surveillance datasets (Australia: 2 months; UK: 9 years).
• Evaluation of DODGE's ability to detect known outbreaks and assess investigation clusters.

Main Results:

• DODGE successfully analyzed expanding WGS datasets and identified investigation clusters.
• Two known UK outbreaks were detected by DODGE earlier than their official reporting time.
• A minority of isolates were identified as investigation clusters in both datasets.

Conclusions:

• DODGE demonstrates potential to enhance the effectiveness and timeliness of genomic surveillance for foodborne diseases.
• The algorithm facilitates integrated analysis by enabling standardized naming of genomic clusters.
• DODGE offers a dynamic approach to managing genetic thresholds for outbreak detection.