Poster Presentation Australian Society for Microbiology Annual Scientific Meeting 2023

Dynamic genomic thresholds for outbreak detection in Salmonella. (#180)

Michael Payne 1 , Qinning Wang 2 , Amy Jennison 3 , Vitali Sintchenko 2 4 , Ruiting Lan 1
  1. The School of Biotechnology and Biomolecular Sciences, The University of New South Wales, Sydney, New South Wales, Australia
  2. Centre for Infectious Diseases and Microbiology - Public Health, Institute of Clinical Pathology and Medical Research – NSW Health Pathology, Westmead Hospital, Sydney, New South Wales, Australia
  3. Public Health Microbiology, Forensic and Scientific Services, Queensland Department of Health, Coopers Plains, Queensland, Australia
  4. Sydney Institute for Infectious Diseases, Sydney Medical School, University of Sydney, Sydney, NSW, Australia

Salmonella is one of the leading causes of foodborne disease in Australia and globally. Infections often occur in point source outbreaks where a single source causes many infections. The detection, tracing and control of these outbreaks is therefore key to limiting the public health impact of Salmonella.  

Surveillance of Salmonella in public health is rapidly moving to genomic typing methods, such as core genome multilocus sequence typing (cgMLST). One complication is that outbreaks are often composed of groups of closely related, but not identical, isolates which are assigned to more than one cgMLST sequence types (ST). A solution is to group STs together using a genetic difference threshold. The selection of an appropriate threshold remains a challenge. A fixed threshold lacks sensitivity and specificity in outbreak detection. A dynamic threshold is desirable as it can take account of differences in the population diversity of the outbreak strain.

In this study we developed an algorithm that incorporates temporal thresholds and background population diversity to define a dynamic threshold that can vary between potential outbreak clusters examined. The algorithm can also utilise our previously described genomic nomenclatures using multilevel genome typing to provide standardised naming of potential outbreak clusters, so that they can be tracked over time and between jurisdictions. We analysed two months of genome sequencing data from Salmonella Typhimurium isolates (N=517) from NSW and Qld and identified 15 potential outbreak clusters (225 isolates). We also analysed 6873 isolates from the UK over 4 years and identified 66 potential outbreak clusters (1063 isolates). Interestingly some of these clusters reoccurred across 2 or 3 years indicating potential reservoirs that may cause repeated outbreaks.

This novel algorithm should allow outbreaks to be detected more rapidly and precisely. In turn this will allow more targeted public health investigation of potential outbreaks, saving time and precious resources and ultimately helping to reduce the public health impact of Salmonella outbreaks.