Version 7

The genotype of each isolate was identified by compiling the genotype generated by each VNTR locus. The phylogenetic tree was generated using an unweighted pair group method with arithmetic mean (UPGMA) method^{35 (link)} that is embedded in BioNumerics version 7.6 (Applied Maths, Austin, TX). The cluster analysis was performed using the UPGMA with a minimum spanning tree (MST) and distance matrices for categorical data by BioNumerics version 7.6. The Geographic distribution map template was downloaded from www.vectortemplates.com (Graphics Factory CC, Western Cape, Sourth Africa) with permission, and the added pie charts were constructed using BioNumerics version 7.6. Hunter and Gaston discrimination index (HGDI) was calculated using Ridom EpiCompare software version 1.0 (www.ridom.de/epicompare/) to elucidate the discriminatory power of the genotyping methods, which explained the probability of two unrelated and different isolates sampled from the test population grouping as different subtypes by a specific typing method^{28 (link)}. The 95% confidence intervals (CI) were calculated according to the method previously described^{36 (link)}.

Clinical isolates of L. monocytogenes identified in hospital and private laboratories are submitted to the PHO Laboratory for molecular subtyping and surveillance purposes. Food samples and environmental swabs collected by public health unit officials in relation to a L. monocytogenes clinical case are also submitted to the PHO Laboratory, and testing is performed according to the Health Canada method MFHPB-30 [38 ]. During the outbreak investigations molecular subtyping by both pulsed field gel electrophoresis (PFGE) and WGS was performed in accordance with PulseNet Canada standard operating procedures; both methods were used as the PHO Laboratory was transitioning methods during this time. During the first outbreak, PFGE was performed for the clinical, food and environmental L. monocytogenes isolates at the PHO Laboratory and WGS was performed at the National Microbiology Laboratory (NML) in Winnipeg, Manitoba, Canada. PFGE and WGS for the second outbreak were both performed at the PHO Laboratory. Comparisons were generated using BioNumerics version 7.6 and nodes were calculated using UPGMA (Unweighted Pair Group Method with Arithmetic Mean), which is the standard analytical process for PulseNet Canada laboratories. Comparative analysis of WGS results from both outbreaks one and two were performed by the NML.

Multilocus sequence typing (MLST) was performed by sequencing seven house-keeping genes of C. difficile (adk, atpA, dxr, glyA, recA, sodA and tpi) as previously reported by Griffiths D et al. [11 (link)]. Sequence types (STs) and clades of C. difficile strains were confirmed by querying on http://pubmlst.org/ website. A minimum spanning tree generated from BioNumerics version 7.6 was used to show the genetic diversity of the MLST data derived from this study.
Capillary gel electrophoresis-based PCR ribotyping was implemented according to a previous report by Fawley WN et al. [17 (link)]. Ribotypes were identified by querying on the WEBRIBO web-based database (http://webribo.ages.at). The novel ribotype was named as “Chongqing Ribotype” (CQR) plus two Arabic numbers (e.g., CQR01).