Key points are not available for this paper at this time.
BACKGROUND: Many hospital laboratories have the technical capacity to perform microbial whole-genome sequencing but lack bioinformatic expertise to analyse sequence data. Sending isolates to reference laboratories creates delays that can be highly detrimental to outbreak responses. The Wellington Regional Hospital laboratory, which lacks on-site bioinformaticians, implemented real-time nanopore-based genome sequencing that has detected several hospital outbreaks at an early stage. This has required off-site analysis, often taking weeks. Solu Genomics, a cloud-based automated bioinformatic platform, requires no bioinformatic or command-line expertise and accepts basecalled sequence files or genome assemblies. AIM: The aim of this study was to use Solu to replicate the analysis of two prior neonatal unit outbreaks detected by on-site genome sequencing, as if they had occurred now, and compare the output to 'manual' bioinformatic analysis. METHODS: Surveillance isolates that had been sequenced up until the beginning of each outbreak were loaded into Solu to replicate the background genomic data available when each outbreak occurred. The 13 methicillin-resistant Staphylococcus aureus (MRSA) and seven Klebsiella variicola outbreak isolates were then uploaded. FINDINGS: Including upload time, each Solu analysis was completed in under 40 min. The phylogenetic trees generated showed distinct clustering of outbreak isolates, with overall tree topologies similar to the manual analyses. Median pairwise single-nucleotide variant distances were 12 (range: 4-27) and 6 (range: 1-11) for the MRSA and K. variicola outbreaks, respectively, versus 6 (range: 0-14) and 18 (range: 0-54) for the manual analyses. CONCLUSION: Solu Genomics delivers high-resolution, actionable outbreak analysis for common hospital bacterial pathogens within minutes, transforming routine nanopore sequencing into infection prevention and control (IPC)-ready insights without on-site bioinformatics.
Bloomfield et al. (Sat,) studied this question.