Uncovering infant group B streptococcal (GBS) disease clusters in the UK and Ireland through genomic analysis: a population-based epidemiological study.

BACKGROUND The true frequency of hospital outbreaks of invasive group B streptococcal (iGBS; Streptococcus agalactiae) disease in infants is unknown. We used whole genome sequencing (WGS) of iGBS isolates collected during a period of enhanced surveillance of infant iGBS disease in the UK and Ireland to determine the number of clustered cases. METHODS Potentially linked iGBS cases from infants with early (<7 days of life) or late-onset (7-89 days) disease were identified from WGS data (HiSeq 2500 platform, Illumina) from clinical sterile site isolates collected between 04/2014 and 04/2015. We assessed time and place of cases to determine a single-nucleotide polymorphism (SNP) difference threshold for clustered cases. Case details were augmented through linkage to national hospital admission data and hospital record review by local microbiologists. RESULTS Analysis of sequences indicated a cut-off of ≤5 SNP differences to define iGBS clusters. Among 410 infant iGBS isolates, we identified 7 clusters (4 genetically-identical pairs with 0 SNP differences, one pair with 3 SNP differences, one cluster of 4 cases with ≤1 SNP differences) of which 4 clusters were uncovered for the first time. The clusters comprised 16 cases, of which 15 were late-onset (of 192 late-onset cases with sequenced isolates) and one an early-onset index case. Serial intervals between cases ranged from 0-59 (median 12) days. CONCLUSIONS Approximately one in 12 late-onset infant iGBS cases were part of a hospital cluster. Over half of clusters were previously undetected, emphasising the importance of routine submission of iGBS isolates to reference laboratories for cluster identification and genomic confirmation.