Existing and potential infection risk zones of yellow fever worldwide: a modelling analysis

The high resolution maps of risk and incidence of yellow fever produced in this study highlight the success of vaccination in reducing incidence in high-risk regions, including Brazil, Cameroon, and Togo. The maps also identify areas with high predicted average annual case numbers, where vaccination coverage in 2016 was less than the recommended threshold to prevent outbreaks, such as large parts of Nigeria, DR Congo, and South Sudan. These maps provide an evidence base to prioritise areas for vaccination and vector control programmes in current risk zones. Receptivity to yellow fever virus transmission in areas outside risk zones was also mapped, including areas where yellow fever has been controlled, such as Central America, or has never persisted, such as southern Asia. Our findings highlight areas where public health authorities should be most vigilant for potential spread or importation events. Furthermore, the areas of high receptivity near the edge of the current risk zone on the southeast coast of Brazil (states of Bahia, Minas Gerais, São Paulo, Espírito Santo, and Rio de Janeiro) reflect a need to geographically expand the existing risk limits because locally acquired cases have recently been reported in this area.28x28WHO. Scientific and Technical Advisory Group on Yellow Fever Risk Mapping. Fourth teleconference report.; Jan 27, 2017. ()
Google ScholarSee all References

Rogers and colleagues3x3Rogers, DJ, Wilson, AJ, Hay, SI, and Graham, AJ. The global distribution of yellow fever and dengue. Adv Parasitol. 2006;
62: 181–220
Crossref | PubMed | Scopus (90)
| Google ScholarSee all References

We included distributions of A aegypti and non-human primate hosts in our model, because countries where both exist are thought to be most vulnerable to the introduction and establishment of yellow fever virus.12x12Reisen, WK. Landscape epidemiology of vector-borne diseases. Annu Rev Entomol. 2010;
55: 461–483
Crossref | PubMed | Scopus (115)
| Google ScholarSee all References

Yellow fever has both urban and sylvatic transmission cycles, and the spatial and temporal distribution of cases fluctuates substantially. The periodicity of yellow fever epidemic activity is driven in part by the rapid depletion and slow replenishment of susceptible hosts, as illustrated by upsurges at irregular intervals in parts of west and east Africa. Our model did not attempt to capture temporal and spatial spikes in cases of yellow fever; the focus was on predicting spatial variation in the underlying risk of infection. Our outputs represent the period from 1970 to 2016, averaged over the large fluctuations that occur. Thus, the results might smooth over important secular trends across time, and our estimated incidence for a particular location will be different from the 2016 incidence, or even a 5-year average. Indeed, the five most recently reported outbreaks of yellow fever began in areas of variable predicted risk. Additionally, since all reports of yellow fever virus infection in human beings were included in the model, irrespective of whether infection was the result of a sylvatic, intermediate, or urban transmission cycle, the model predicts apparent infection risk from any transmission cycle. The BRT model is capable of encompassing different relations in the data arising from these distinct transmission cycles, with the given covariates.

Our model predicts receptivity to yellow fever virus transmission based on the relation between locations where yellow fever has been reported from 1970 to 2016, and the values of environmental and biological covariates at those locations. For this reason, predictions of high receptivity in southeast Asia should be interpreted with caution. Potential variables that distinguish current risk zones in Africa and the Americas from A aegypti-inhabited areas of southeast Asia might be missing from our analysis. Indeed, most theories that have been used to explain the absence of yellow fever in Asia involve biological factors rather than climatic or environmental ones. These factors include cross-protection from other flaviviruses, lower competence of local populations of A aegypti, and competition between other flaviviruses within mosquito cells,29x29Amaku, M, Coutinho, FAB, and Massad, E. Why dengue and yellow fever coexist in some areas of the world and not in others?. Biosystems. 2011;
106: 111–120
Crossref | PubMed | Scopus (18)
| Google ScholarSee all References

Our study involved multiple stages of analysis, each containing potential sources of bias and uncertainty, much of which was difficult to account for or quantify. Each stage of the analytical process involved fitting a model to an independent, fixed dataset. Since each of these models was fitted to fixed data, not the output of previous models, it was not meaningful to propagate uncertainty through these steps. However, there were some steps where it was important to account for or report uncertainties. In fitting the model of risk for apparent yellow fever virus infection, uncertainty in the spatial locations of reported cases was propagated through the model via Monte Carlo simulation. Similarly, uncertainty in the final step of the analytical process—calibrating spatial predictions of yellow fever risk against GBD 2015 estimates of annual cases—was estimated and reflected in reported results.

High-quality spatial data on yellow fever are lacking, largely because of diagnostic complexity and limitations of health-care systems in many affected countries. Our occurrence dataset included human yellow fever virus infections confirmed via both genetic and serological diagnostic methods. As a result of cross-reactivity among flaviviruses, the precision of serological diagnostic techniques is limited, particularly when considering historical records. Additionally, we assumed that estimates of vaccination coverage, combined with data on population size, were proportional to the number of people susceptible to yellow fever virus infection. Translating maps of unvaccinated individuals into maps of susceptible people is complicated by the acquisition of immunity via natural infection, which is difficult to quantify. Modelling efforts will be improved as the volume and quality of geographical data on yellow fever increases, ideally using diagnostics that distinguish past infection from vaccination, and from other flavivirus infections. For now, the work presented here provides the best available evidence base for identifying populations most vulnerable to yellow fever.

To develop the most cost-effective vaccination strategies that prevent outbreaks and minimise adverse events, vaccination policy makers require a clear understanding of geographical disease risk. Within risk zones, our model improves understanding of geographical risk and comes at a time when current policies may require re-evaluation in view of the demonstrated inadequacy of emergency stocks and surge capacity of vaccine manufacture to meet the needs of the recent outbreak in Angola.7x7Vasconcelos, PFC and Monath, TP. Yellow fever remains a potential threat to public health. Vector Borne Zoonotic Dis. 2016;
16: 566–567
Crossref | PubMed | Scopus (14)
| Google ScholarSee all References

Source link

قالب وردپرس

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *


Adblock Detected

Please consider supporting us by disabling your ad blocker