Predicting dengue incidence leveraging internet-based data sources. A case study in 20 cities in Brazil

Abstract

The dengue virus affects millions of people every year worldwide, causing large epidemic outbreaks that disrupt people’s lives and severely strain healthcare systems. In the absence of a reliable vaccine against dengue or an effective treatment to manage the illness in humans, most efforts to combat dengue infections have focused on preventing its vectors, mainly the Aedes aegypti mosquito, from flourishing across the world. These mosquito-control strategies need reliable disease activity surveillance systems to be deployed. Despite significant efforts to estimate dengue incidence using a variety of data sources and methods, little work has been done to understand the relative contribution of the different data sources to improved prediction. Additionally, scholarship on the topic had initially focused on prediction systems at the national- and state-levels, and much remains to be done at the finer spatial resolutions at which health policy interventions often occur. We develop a methodological framework to assess and compare dengue incidence estimates at the city level, and evaluate the performance of a collection of models on 20 different cities in Brazil. The data sources we use towards this end are weekly incidence counts from prior years (seasonal autoregressive terms), weekly-aggregated weather variables, and real-time internet search data. We find that both random forest-based models and LASSO regression-based models effectively leverage these multiple data sources to produce accurate predictions, and that while the performance between them is comparable on average, the former method produces fewer extreme outliers, and can thus be considered more robust. For real-time predictions that assume long delays (6–8 weeks) in the availability of epidemiological data, we find that real-time internet search data are the strongest predictors of dengue incidence, whereas for predictions that assume short delays (1–3 weeks), in which the error rate is halved (as measured by relative RMSE), short-term and seasonal autocorrelation are the dominant predictors. Despite the difficulties inherent to city-level prediction, our framework achieves meaningful and actionable estimates across cities with different demographic, geographic and epidemic characteristics.

Plasma metabolomics analyses highlight the multifaceted effects of noise exposure and the diagnostic power of dysregulated metabolites for noise-induced hearing loss in steel workers

“…Furthermore, the diagnostic efficiency of the dysregulated metabolites was evaluated. As a good method in machine learning, LASSO regression has been widely used in cancer marker studies (Hua et al, 2022;Koplewitz et al, 2022). In this study, with LASSO regression analysis, we constructed two risk models with the seven key metabolites for NIHL status predication.…”

Saúde Pública: Princípios e práticas

“…Possui 4 sorotipos conhecidos: DENV 1, DENV 2, DENV 3 e DENV 4, apresentando como transmissor o mosquito Aedes aegypti, que tem preferência por áreas urbanas e de elevada densidade populacional para sua proliferação (CAVALLI et al, 2019). Uma variedade de condições externas demonstrou afetar a transmissão da dengue como a precipitação, temperatura e padrões climáticos sazonais que convergem para a propagação da doença, desenvolvimento e vida útil dos mosquitos transmissores (KOPLEWITZ et al, 2022).…”

Early Warning and Prediction of Scarlet Fever in China Using the Baidu Search Index and Autoregressive Integrated Moving Average With Explanatory Variable (ARIMAX) Model: Time Series Analysis

“…An increasing number of individuals are inclined to search for health-related information on the internet before seeking medical services, opening up the potential for early disease surveillance by monitoring fluctuations in the frequency of specific search keywords [11]. Internet-derived data revealed substantial potential in the application of infectious disease surveillance worldwide [12,13]. Google's influenza monitoring data became available 2 weeks before official announcements in the United States [14].…”