Fig 1.
Map of Samoa with census regions, roads, and health facilities.
Samoa is comprised of two primary islands, Upolu to the east and Savaii to the west. About three quarters of the population and the capital town of Apia are located on Upolu. The black lines indicate roads associated with populated areas, which are predominantly located around the periphery of the islands except for the urban areas where inland settlements are increasing. The islands are subdivided into four statistical census regions, 43 political districts (outlined in light grey), and >300 villages (not depicted). The four colored geographic areas differentiate the four census regions: Apia Urban Area (AUA), North West Upolu (NWU), Rest of Upolu (ROU), and Savaii (SAV). AUA is most dense with ~25% of the total population. Apia is the capital seat of government and includes the Ministry of Health headquarters, Samoa Typhoid Fever Public Health Laboratory, and Tupua Tamasese Meaole (TTM) national hospital, which processes the blood cultures for Upolu (insert). NWU is mixed peri-urban and rural in population density and separates Apia from the inter-island ferry (navy dashed line) and adjacent international airport (not shown). The Malietoa Tanumafili II (MT2) hospital processes blood cultures for Savaii. The health facility icons indicate the two central hospitals with clinical microbiology laboratories, TTM and MT2, and the 6 district hospitals and 4 peripheral health centers. Base layer map available from https://planet.openstreetmap.org.
Table 1.
Comparison of age and sex variables of acute cases associated with an S. Typhi whole genome sequence (WGS) versus acute cases missing WGS data.
Fig 2.
Spatial-temporal and genomic analysis of typhoid fever cases (burden) in Samoa from April 2018 through June 2020.
(A) Using ArcGIS Pro v2.9, household GPS coordinates of acute cases of S. Typhi (n = 260) from April 27, 2018 to June 9, 2020 are plotted on a map of Samoa and colored according to S. Typhi genotype/sub-lineage. A point density heatmap is inferred from case coordinates. The four prominent hotspots are identified with arrows labeled 1–4, which also indicate aerial perspectives shown in the reconstructions in panel B. (B) Coordinates were binned into space-time cubes of 1km2 x 2-months (see panel C and S1 Text for additional details) and tested for significant clustering by the emerging hotspot analysis implemented in ArcGIS Pro. The resulting 3D space-time hotspots were visualized from an aerial view of 20,000 meters and 10,000 meters focusing on the four typhoid hotspots identified in panel A. In the 3D renderings, each vertical column on the map represents a stack of cubes over 1 km2 area and with a height representing 2 months of time, totaling 13 time steps, or all 26 months of this study. Aggregation of points in both space and time generate a space-time hotspot, visible as a red cube or band of several adjacent cubes. The intensity of red shading indicates statistical confidence. (C) To generate space-time cubes, points were graphed in a 3-dimensional plane based on coordinates and time, and then binned into cubes of space and time dimensions. Counts within each cube were compared to nearby cubes to extrapolate hot and cold spots with statistical support (99%, 95%, 90%, not significant). Hotspots in the x-y plane occur within a time slice. Hotspots inferred from the z-direction fall within a time series. Adapted with permission from Esri (Redlands, CA: Environmental Systems Research Institute, Inc.). Base layer map available from https://planet.openstreetmap.org.
Table 2.
Results of the Average Nearest Neighbor (ANN) statistic performed on geographic and phylogenetic groupings of S. Typhi acute infections.
Fig 3.
Maximum-likelihood phylogeny, associated epidemiologic variables, and epidemiologic linkages among S. Typhi from Samoa from April 2018 through June 2020.
(A) Maximum-likelihood phylogeny from de novo assembled whole genome sequences of Samoan S. Typhi genotype 3.5.4 isolates from April 27, 2018 through June 9, 2020 (N = 186, S1 Table). Core genome single nucleotide polymorphisms (SNPs) were identified relative to the 2012 Samoa Reference genome H12ESR00394-001A (GCA_001118185.2). Genotype/sub-lineage is indicated by isolate label shading (ring 1). Two outlier isolates are not shaded. Asymptomatic shedders of S. Typhi are denoted with a pink star. The patient age group (ring 2), patient sex (ring 3), census region of residence (ring 4), and year of isolation (ring 5) are annotated in the circumscribing colored rings. (B) Epidemiologic linkages (EL) identified through case household investigations are compared against genomic relatedness of the isolates based on core genome phylogeny. The visible isolates were extracted from the tree in Panel A. In total 2/12 epidemiologic linkages are not supported by the genomic variation observed in the isolates from those samples (EL5/dark green and EL11/pink groups are not similar). Three repeat positive (RP) blood cultures isolated ~1 month apart are also compared, showing close genetic similarity.
Table 3.
Comparison of epidemiologic variables relating to person, place, and time for each Samoa dominant S. Typhi genotype/sub-lineage.
Fig 4.
Bar graphs depicting the number and proportion of blood culture-confirmed cases of typhoid fever in Samoa over time from April 27, 2018 through June 9, 2020 by genotype/sub-lineage.
(A-B) The frequency and proportion of acute cases (N = 260) occurring each month are colored by genotype/sub-lineage (n = 172) or greyed if no genome was available (n = 88). Genotype 3.5.4 sub-lineages 1–3, comprising a total of 166/172 sequenced isolates from acute cases, are identified as orange, yellow, and blue bars, respectively. Minor genotypes (purple bar) include genotype 4.1 (4/186), genotype 3.5.3 (1/186), and one genotype 3.5.4 outlier (1/186) that did not fall into one of the three sub-lineages. (C-F) The frequency and proportion of acute cases occurring over 2-month intervals in each census region. Three major events impacting population mobility–increases during the 2019 Pacific Games in Nov-Dec 2019, and restrictions during the measles lockdown from Nov-Dec 2019 [35] and COVID-19 lockdown from Mar-Apr 2020 [36]–are labeled as light red, light purple, and light green background colors, respectively, over the two-month time intervals affected.
Table 4.
Suspected epidemiologic linkages detected through active surveillance and associated genotype and pairwise SNP distance between isolates.