Spatiotemporal Heterogeneity in the Distribution of Chikungunya and Zika Virus Case Incidences and Risk Factors During Their Epidemics in Barranquilla, Colombia, between 2014 and 2016: An Ecological Study

Abstract

Chikungunya virus (CHIKV) and Zika virus (ZIKV) have recently emerged as global infections with consequential disability adjusted life years (DALYs) and economic burden. This study aimed to explore the spatiotemporal heterogeneity in the occurrence of CHIKV and ZIKV outbreaks throughout Barranquilla, Colombia during 2014 and 2016 and explored the potential for clustering. Incidence data were fitted using multiple Bayesian Poisson models based on a suite of explanatory variables as potential risk factors and multiple options for random effects. A best fit model was used to analyse the case incidence risk for both epidemics to identify any risk factors during their epidemics. Neighbourhoods in the northern region of Barranquilla were hotspots for the outbreaks of both CHIKV and ZIKV. Additional hotspots occurred in the south-western and central regions of the CHIKV and ZIKV outbreaks, respectively. Multivariate conditional autoregressive models strongly identified higher socioeconomic strata (SES) and residing in detached houses as risk factors for ZIKV case incidences. These novel findings challenge the belief that these infections are driven by social vulnerability and merits further study both in Barranquilla and throughout the tropical and subtropical regions of the world. 

Epidemics of chikungunya, Zika, and COVID-19 reveal bias in case-based mapping

“…2 A large proportion of infections caused by ZIKV, the Asian lineage of CHIKV, and SARS-CoV-2 are subclinical, yet they contribute substantially to transmission. [4][5][6] Standard spatial studies of these viruses [7][8][9][10][11] and other pathogens nevertheless focus exclusively on cases, neglecting subclinical infections and thereby precluding a full understanding of the spatial contours of immunity. Further, such studies are unable to distinguish between uninfected persons and those with subclinical infections, as only cases are evaluated.…”

“…For a naïve population that experiences an epidemic, infection risk is equivalent to seroprevalence. The incidence rate is commonly misinterpreted in spatial health research as the risk of infection [7][8][9] or disease. 10,11 Below, we provide three forms of the same relationship between these three measures:…”

“…Controlling epidemic spread requires accurate data on the movement of pathogens through populations. Standard spatial studies of infectious diseases use passively collected, individual-level data for cases (symptomatic infections) from health facilities after cases present for medical treatment (1)(2)(3)(4)(5). Then, by using census data to obtain the total population in an area, these studies estimate incidence rates (attack rates, incidence proportions) as the ratio of cases to the total population.…”

“…However, because this approach does not capture subclinical (clinically inapparent) infections, this incidence approach may not recapitulate the spatial contour of infections, which may have a distinct pattern and magnitude. These issues may be compounded when the incidence rate, estimated from passively collected and hence incomplete case data, is used to infer infection risk (1)(2)(3) or disease risk (4,5) in policy decision-making on epidemic control.…”