Group B Streptococcus Infections Caused by Improper Sourcing and Handling of Fish for Raw Consumption, Singapore, 2015-2016

Abstract

We assessed microbial safety and quality of raw fish sold in Singapore during 2015-2016 to complement epidemiologic findings for an outbreak of infection with group B Streptococcus serotype III sequence type (ST) 283 associated with raw fish consumption. Fish-associated group B Streptococcus ST283 strains included strains nearly identical (0-2 single-nucleotide polymorphisms) with the human outbreak strain, as well as strains in another distinct ST283 clade (57-71 single-nucleotide polymorphisms). Our investigations highlight the risk for contamination of freshwater fish (which are handled and distributed separately from saltwater fish sold as sashimi) and the need for improved hygienic handling of all fish for raw consumption. These results have led to updated policy and guidelines regarding the sale of ready-to-eat raw fish dishes in Singapore.

Keywords: GBS; Singapore; bacteria; clonal complex; consumption; food safety; group B Streptococcus; handling; humans infections; outbreak; raw fish; sequence type; streptococci.

Figure 1

SPCs for sliced fish samples collected from various eating establishments during investigation of group B Streptococcus infections, Singapore, 2015–2016. Dashed horizontal line indicates regulatory limit of Singapore for SPCs for ready-to-eat foods (<5 log₁₀ CFU/g) (14). Top and bottom of boxes in plots indicate 25th and 75th percentiles, horizontal lines indicate medians, and whiskers indicate minimum and maximum values. *p<0.05. †Food stalls housed within larger eating establishments that include hawker centers, coffee shops, and eating houses. Open circle indicates an outlier. NS, not significant (p>0.05); RTE, ready to eat; SG, Singapore government; SPCs, standard plate counts.

Figure 2

Escherichia coli (A and C) and Staphylococcus aureus (B and D) counts in fish samples (muscle and surface swabs) collected from fresh produce markets during investigation of group B Streptococcus infections, Singapore, 2015–2016. Solid horizontal lines indicate ICMSF limit for E. coli count in fresh fish intended for cooking (<2.7 log₁₀ CFU/g or <4.7 log₁₀ CFU/100 cm²) (23). Dashed horizontal line indicates Singapore regulatory limit for E. coli count ready-to-eat foods (<1.3 log₁₀ CFU/g) (14). Top and bottom of boxes in plots indicate 25th and 75th percentiles, horizontal lines indicate medians, and whiskers indicate minimum and maximum values. Open circles indicate outliers. *p<0.05. †Companies that supplied sashimi grade fish to restaurants and snack bars. ‡Fish stalls at ports and wet markets, as well as fresh produce sections of supermarkets, excluding sashimi and sushi counters of supermarkets. ICMSF, International Commission on Microbiological Specifications of Foods; NS, not significant (p>0.05); SG, Singapore government; SPCs, standard plate counts.

Figure 3

SPCs for fish samples (muscle) collected from fresh produce markets during investigation of group B Streptococcus infections, Singapore, 2015–2016. Solid horizontal line indicates ICMSF limit for SPCs in fresh fish intended for cooking (<7 log₁₀ CFU/g) (23). Dashed horizontal line indicates Singapore regulatory limit for SPCs for ready-to-eat foods (<5 log₁₀ CFU/g) (14). Top and bottom of boxes in plots indicate 25th and 75th percentiles, horizontal lines indicate medians, and whiskers indicate minimum and maximum values. Open circles indicate outliers. *p<0.05. †Companies that supplied sashimi grade fish to restaurants and snack bars. ‡Fish stalls at ports and wet markets, as well as fresh produce sections of supermarkets, excluding sashimi and sushi counters of supermarkets. ICMSF, International Commission on Microbiological Specifications of Foods; SG, Singapore government; SPCs, standard plate counts.

Figure 4

Phylogenetic analysis of group B Streptococcus (GBS) infections caused by improper sourcing and handling of raw fish for raw consumption, Singapore, 2015–2016. A) Maximum-likelihood single-nucleotide polymorphism (SNP)–based tree for GBS ST283 strains relative to the SG-M1 reference human outbreak strain. Year, location, and source (human or fish) for isolates are indicated. Twelve strains from 6 fish (SGEHI2015-NWC941, SGEHI2015–95, LG01, LG02, LG04, and LG06) and 4 fish tank water samples (LG07, LG08, LG09, and LG11) were nearly identical to the local reference outbreak strain SG-M1 (no SNP, 0 and 12 indels, respectively). Scale bar indicates distance over 4,093 total SNPs. B) Enlargement of blue subtree from bottom of tree in panel A showing fish GBS ST283 isolates that were different (57–71 SNPs and 11–33 indels) from the human outbreak strain. ST, sequence type.