Spatial-Temporal Epidemiology of the Syphilis Epidemic in Relation to Neighborhood-Level Structural Factors in British Columbia, 2005–2016

• Published in: Sexually transmitted diseases Vol. 46; no. 9; pp. 571 - 578
• Main Authors: Salway, Travis, Gesink, Dionne, Lukac, Christine, Roth, David, Ryan, Venessa, Mak, Sunny, Wang, Susan, Newhouse, Emily, Hayden, Althea, Bharmal, Aamir, Hoyano, Dee, Morshed, Muhammad, Grennan, Troy, Gilbert, Mark, Wong, Jason
• Format: Journal Article
• Language: English
• Published: United States Copyright American Sexually Transmitted Diseases Association 01-09-2019; Lippincott Williams & Wilkins Ovid Technologies
• Subjects: Bisexuality; Central business districts; Clustering; Clusters; Density; Epidemics; Epidemiology; Health risk assessment; Homosexuality; Infectious diseases; Men; Neighborhoods; Population statistics; Sexually transmitted diseases; Spatial analysis; Spatial distribution; STD; Syphilis; British Columbia Canada; Canada
• ISSN: 0148-5717; 1537-4521
• DOI: 10.1097/OLQ.0000000000001034

BACKGROUNDSpatial clusters of syphilis have been observed within several jurisdictions globally; however, the degree to which they are predicted by the spatial distributions of gay, bisexual, and other men who have sex with men (GBM) and testing remains unknown. We sought to describe the spatial-temporal epidemiology of infectious syphilis and identify associations between neighborhood-level factors and rates of syphilis, in British Columbia, Canada.
METHODSWe used ArcGIS to map infectious syphilis cases among men (2005 to 2016), SaTScan to detect areas with significantly elevated rates of syphilis, and spatial regression to identify associations between neighborhood-level factors and rates of syphilis.
RESULTSFive clusters were identifieda core in downtown Vancouver (incidence rate ratio [IRR], 18.0; 2007–2016), 2 clusters adjacent to the core (IRR, 3.3; 2012–2016; and IRR, 2.2; 2013–2016), 1 cluster east of Vancouver (IRR, 2.1; 2013–2016), and 1 cluster in Victoria (IRR, 4.3; 2015–2016). Epidemic curves were synchronized across cluster and noncluster regions. Neighborhood-level GBM population estimates and testing rates were both associated with syphilis rates; however, the spatial distribution of syphilis was not fully explained by either of these factors.
CONCLUSIONSWe identified two novel ecologic correlates of the spatial distribution of infectious syphilis—density of GBM and rates of syphilis testing—and found that these factors partially, though not entirely, explained the spatial distribution of clusters. Residual spatial autocorrelation suggests that greater syphilis testing coverage may be needed and low-barrier GBM-affirming testing should be expanded to regions outside the core.