Social network and genomic analysis of an OXA-48 carbapenemase-producing Enterobacterales hospital ward outbreak in Ireland, 2018-2019.

Background: Nosocomial transmission and outbreaks of carbapenemase-producing Enterobacterales (CPE) represent a challenge to healthcare systems. In July 2018, a CPE hospital ward outbreak was declared. Our aim was to investigate transmission patterns, using social network analysis and genomics in a nosocomial CPE outbreak.

Methods: A retrospective descriptive analysis of all patients (cases and contacts) admitted to a ward experiencing a CPE outbreak (2018-2019) was undertaken. A case had a negative CPE admission screen, and subsequent positive test. A contact shared a multi-bed area and/or facility with a case (>4 hours). Social networks, including genomics data and ward locations, were constructed. Network metrics were analysed.

Findings: Forty-five cases and 844 contacts were analysed. The median age of cases was 78 years (IQR 67-83), 58% (n=26) were male and 100% had co-morbidities. The median outbreak ward length-of-stay (LOS) was 17 days (IQR 10-34). OXA-48 CPE was confirmed in all cases and from 26 environmental samples. Social networks identified clusters by time, gender and species/sequence type/plasmid. Network metrics indicated potential superspreading involving a subset of patients with behavioural issues.

Conclusion: Social networks elucidated high resolution transmission patterns involving two related OXA-48 plasmids, multiple species/genotypes and potential super-spreading. Interventions prevented intra-hospital spread. An older patient cohort, extended hospital LOS and frequent intra-ward bed transfers, coupled with suboptimal ward infrastructure, likely prolonged this outbreak. We recommend social network analysis contemporaneously with genomics (on case and environmental samples) for complex nosocomial outbreaks and bespoke care plans for patients with behavioural issues on outbreak wards.