Defining a genetic relatedness cut-off to exclude recent MRSA transmission: a genomic epidemiology analysis
Published version
Peer-reviewed
Repository URI
Repository DOI
Change log
Authors
Abstract
Background Whole-genome sequencing (WGS) of bacterial pathogens such as methicillin-resistant Staphylococcus aureus (MRSA) has emerged as a transformative tool for the detection of outbreaks. WGS can be applied in genomic epidemiology investigations to confirm or refute bacterial outbreaks, supporting targeted and efficient infection control interventions. Recent patient-to-patient transmission can be ruled out, for infection control purposes, if the genetic relatedness between their bacterial isolates (as measured by the number of single nucleotide polymorphisms (SNPs)), is greater than a defined threshold. Here, we aim to define a SNP cut-off above which recent transmission can be ruled out. Methods Our findings were informed by three approaches applied to two patient cohorts, whose MRSA isolates were sequenced. First, we applied a linear mixed model to estimate the S. aureus substitution rate and within-host diversity in a cohort of 459 people with multiple isolates sequenced per individual. Second, we applied a simulated transmission model to this genomic dataset. Third, we used a second cohort of 778 patients to determine the genetic distance to capture 95% of epidemiologically linked cases. Findings The linear mixed model estimated a substitution rate of 5 whole-genome (wgSNPs), or 3 core-genome SNPs (cgSNPs), per genome per year and quantified the within-host diversity to be 19 wgSNPs (95 percentile), or 10 cgSNPs, giving a combined SNP cut-off of 24 wgSNPs (13 cgSNPs) to detect transmission within six months. The transmission model suggested a 17 wgSNPs (12 cgSNPs) cut-off to detect 95% of transmission events within the same timeframe. In the second cohort, we determined that a 22 wgSNPs (11 cgSNPs) captured 95% of epidemiologically linked cases within six months. Interpretation Based on multiple lines of evidence, we propose a conservative a 25 whole-genome or 15 core-genome SNP cut-off above which recent transmission can be ruled out.
Description
Keywords
Journal Title
Conference Name
Journal ISSN
2666-5247
Volume Title
Publisher
Publisher DOI
Sponsorship
National Institute for Health and Care Research (HICF-T5-342)