Reconstructing Mayotte 2018-19 Rift Valley Fever outbreak in humans by combining serological and surveillance data

BackgroundRift Valley Fever (RVF) is a zoonosis that affects large parts of Africa and the Arabian Peninsula. RVF virus (RVFV) is transmitted to humans through contacts with infected animals, animal products, mosquito bites or aerosols. Its pathogenesis in humans ranges from asymptomatic forms to potentially deadly haemorrhagic fevers, and the true burden of human infections during outbreaks is generally unknown.MethodsWe build a model fitted to both passive surveillance data and serological data collected throughout a RVF epidemic that occurred in Mayotte Island in 2018-2019.ResultsWe estimate that RVFV infected 10,797 (95% CrI 4,728-16,127) people aged >= 15years old in Mayotte during the entire outbreak, among which only 1.2% (0.67%-2.2%) were reported to the syndromic surveillance system. RVFV IgG seroprevalence in people >= 15years old was estimated to increase from 5.5% (3.6%-7.7%) before the outbreak to 12.9% (10.4%-16.3%) thereafter.ConclusionsOur results suggest that a large part of RVFV infected people present subclinical forms of the disease and/or do not reach medical care that could lead to their detection by the surveillance system. This may threaten the implementation of exhaustive RVF surveillance and adequate control programs in affected countries. Plain language summaryRift Valley Fever (RVF) is a disease caused by a virus transmitted from livestock animals to humans by mosquito bites, aerosols or direct contact with infected animals or animal products. In some parts of Africa and the Arabian Peninsula, the virus can lead to large outbreaks in both humans and animals. Despite some infected people developing severe forms of the disease, some experience no or mild symptoms. Therefore, infection is often not detected by surveillance systems based on the reporting of symptoms by patients. Here, we use data collected during a RVF outbreak that occurred in 2018-2019 in Mayotte Island, in the Indian Ocean, to model the course of the outbreak in humans. We estimate that, throughout the epidemic, only 1.2% of infected people were detected by the surveillance system. Our results highlight that most human cases may go unreported during RVF outbreaks, making it difficult to monitor the burden of infections. Bastard et al. estimate the extent of zoonotic human infections during the 2018-2019 Rift Valley Fever outbreak in Mayotte, France, combining passive surveillance and serological data. Estimating the case reporting fraction shows that syndromic surveillance fails to capture most infections during the epidemic.