Current challenges and implications for dengue, chikungunya and Zika seroprevalence studies worldwide: A scoping review

Abstract

Background: Arboviral infections are a public health concern and an escalating problem worldwide. Estimating the burden of these diseases represents a major challenge that is complicated by the large number of unapparent infections, especially those of dengue fever. Serological surveys are thus required to identify the distribution of these diseases and measure their impact. Therefore, we undertook a scoping review of the literature to describe and summarize epidemiological practices, findings and insights related to seroprevalence studies of dengue, chikungunya and Zika virus, which have rapidly expanded across the globe in recent years.

Methodology/principal findings: Relevant studies were retrieved through a literature search of MEDLINE, WHOLIS, Lilacs, SciELO and Scopus (2000 to 2018). In total, 1389 publications were identified. Studies addressing the seroprevalence of dengue, chikungunya and/or Zika written in English or French and meeting the inclusion and exclusion criteria were included. In total, 147 studies were included, from which 185 data points were retrieved, as some studies used several different samples. Most of the studies were exclusively conducted on dengue (66.5%), but 16% were exclusively conducted on chikungunya, and 7 were exclusively conducted on Zika; the remainder were conducted on multiple arboviruses. A wide range of designs were applied, but most studies were conducted in the general population (39%) and in households (41%). Although several assays were used, enzyme-linked immunosorbent assays (ELISAs) were the predominant test used (77%). The temporal distribution of chikungunya studies followed the virus during its rapid expansion since 2004. The results revealed heterogeneity of arboviruses seroprevalence between continents and within a given country for dengue, chikungunya and Zika viruses, ranging from 0 to 100%, 76% and 73% respectively.

Conclusions/significance: Serological surveys provide the most direct measurement for defining the immunity landscape for infectious diseases, but the methodology remains difficult to implement. Overall, dengue, chikungunya and Zika serosurveys followed the expansion of these arboviruses, but there remain gaps in their geographic distribution. This review addresses the challenges for researchers regarding study design biases. Moreover, the development of reliable, rapid and affordable diagnosis tools represents a significant issue concerning the ability of seroprevalence surveys to differentiate infections when multiple viruses co-circulate.

Fig 1. Flowchart of dengue, chikungunya and Zika seroprevalence studies used in the review.

Fig 2. Distribution of laboratory test combinations across studies.

HI: Hemagglutination inhibition; IFA: Immunofluorescence assay; IIFT: Indirect immunofluorescence test; PCR: Polymerase chain reaction; RT-PCR: Reverse transcription PCR; MIA: Microsphere immunoassays; NS1: Non-structural protein 1 antigen test.

Fig 3

A) Distribution of arboviruses seroprevalence studies worldwide. B) Distribution of dengue seroprevalence studies number worldwide, 1989–2017, N = 149. C) Distribution of chikungunya seroprevalence studies number worldwide, 1989–2017, N = 54. D) Distribution of Zika seroprevalence studies number worldwide, 2007–2017, N = 10.

Fig 4. Distribution of dengue, chikungunya and Zika studies according to the year of survey (1989–2017; N = 185).

If a survey was conducted over several years, we plotted this study for each given year.

Fig 5. Dengue seroprevalence rates by assay and continent, according to subpopulations.

Each square and associated 95% confidence interval were derived from an individual study (see S1 Appendix). If a study was conducted in different places within a given country, only the overall mean seroprevalence was included in this figure.

Fig 6. Chikungunya seroprevalence rates by assay and continent, according to subpopulations.

Each square and associated 95% confidence interval were derived from an individual study (see S1 Appendix). If a study was conducted in different places within a given country, only the overall mean seroprevalence was included in this figure.

Fig 7. Zika seroprevalence rates by assay and continent, according to subpopulations.

Each square and associated 95% confidence interval were derived from an individual study (see S1 Appendix). If a study was conducted in different places within a given country, only the overall mean seroprevalence was included in this figure.