Weekly Forecasting of Yellow Fever Occurrence and Incidence via Eco-Meteorological Dynamics

Joseph L. Servadio, Matteo Convertino, Mark Fiecas, Claudia Muñoz-Zanzi

Research output: Contribution to journalArticlepeer-review

Abstract

Yellow Fever (YF), a mosquito-borne disease, requires ongoing surveillance and prevention due to its persistence and ability to cause major epidemics, including one that began in Brazil in 2016. Forecasting based on factors influencing YF risk can improve efficiency in prevention. This study aimed to produce weekly forecasts of YF occurrence and incidence in Brazil using weekly meteorological and ecohydrological conditions. Occurrence was forecast as the probability of observing any cases, and incidence was forecast to represent morbidity if YF occurs. We fit gamma hurdle models, selecting predictors from several meteorological and ecohydrological factors, based on forecast accuracy defined by receiver operator characteristic curves and mean absolute error. We fit separate models for data before and after the start of the 2016 outbreak, forecasting occurrence and incidence for all municipalities of Brazil weekly. Different predictor sets were found to produce most accurate forecasts in each time period, and forecast accuracy was high for both time periods. Temperature, precipitation, and previous YF burden were most influential predictors among models. Minimum, maximum, mean, and range of weekly temperature, precipitation, and humidity contributed to forecasts, with optimal lag times of 2, 6, and 7 weeks depending on time period. Results from this study show the use of environmental predictors in providing regular forecasts of YF burden and producing nationwide forecasts. Weekly forecasts, which can be produced using the forecast model developed in this study, are beneficial for informing immediate preparedness measures.

Original languageEnglish (US)
Article numbere2023GH000870
JournalGeoHealth
Volume7
Issue number10
DOIs
StatePublished - Oct 2023

Bibliographical note

Publisher Copyright:
© 2023 The Authors. GeoHealth published by Wiley Periodicals LLC on behalf of American Geophysical Union.

Keywords

  • Brazil
  • Yellow Fever
  • ecometeorology
  • forecasting
  • gamma hurdle model
  • infectious disease

PubMed: MeSH publication types

  • Journal Article

Fingerprint

Dive into the research topics of 'Weekly Forecasting of Yellow Fever Occurrence and Incidence via Eco-Meteorological Dynamics'. Together they form a unique fingerprint.

Cite this