Dataset Information

Improved Real-Time Influenza Surveillance: Using Internet Search Data in Eight Latin American Countries.

ABSTRACT:

Background

Novel influenza surveillance systems that leverage Internet-based real-time data sources including Internet search frequencies, social-network information, and crowd-sourced flu surveillance tools have shown improved accuracy over the past few years in data-rich countries like the United States. These systems not only track flu activity accurately, but they also report flu estimates a week or more ahead of the publication of reports produced by healthcare-based systems, such as those implemented and managed by the Centers for Disease Control and Prevention. Previous work has shown that the predictive capabilities of novel flu surveillance systems, like Google Flu Trends (GFT), in developing countries in Latin America have not yet delivered acceptable flu estimates.

Objective

The aim of this study was to show that recent methodological improvements on the use of Internet search engine information to track diseases can lead to improved retrospective flu estimates in multiple countries in Latin America.

Methods

A machine learning-based methodology that uses flu-related Internet search activity and historical information to monitor flu activity, named ARGO (AutoRegression with Google search), was extended to generate flu predictions for 8 Latin American countries (Argentina, Bolivia, Brazil, Chile, Mexico, Paraguay, Peru, and Uruguay) for the time period: January 2012 to December of 2016. These retrospective (out-of-sample) Influenza activity predictions were compared with historically observed flu suspected cases in each country, as reported by Flunet, an influenza surveillance database maintained by the World Health Organization. For a baseline comparison, retrospective (out-of-sample) flu estimates were produced for the same time period using autoregressive models that only leverage historical flu activity information.

Results

Our results show that ARGO-like models' predictive power outperform autoregressive models in 6 out of 8 countries in the 2012-2016 time period. Moreover, ARGO significantly improves on historical flu estimates produced by the now discontinued GFT for the time period of 2012-2015, where GFT information is publicly available.

Conclusions

We demonstrate here that a self-correcting machine learning method, leveraging Internet-based disease-related search activity and historical flu trends, has the potential to produce reliable and timely flu estimates in multiple Latin American countries. This methodology may prove helpful to local public health officials who design and implement interventions aimed at mitigating the effects of influenza outbreaks. Our methodology generally outperforms both the now-discontinued tool GFT, and autoregressive methodologies that exploit only historical flu activity to produce future disease estimates.

SUBMITTER: Clemente L

PROVIDER: S-EPMC6470460 | biostudies-literature | 2019 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Improved Real-Time Influenza Surveillance: Using Internet Search Data in Eight Latin American Countries.

Clemente Leonardo L Lu Fred F Santillana Mauricio M

JMIR public health and surveillance 20190404 2

<h4>Background</h4>Novel influenza surveillance systems that leverage Internet-based real-time data sources including Internet search frequencies, social-network information, and crowd-sourced flu surveillance tools have shown improved accuracy over the past few years in data-rich countries like the United States. These systems not only track flu activity accurately, but they also report flu estimates a week or more ahead of the publication of reports produced by healthcare-based systems, such a ...[more]

PMID: 30946017

Similar Datasets

Project description:BackgroundPhysical activity is a cornerstone in the prevention and treatment of obesity. There are relatively few studies that explore the effect of accelerometer-determined moderate-to-vigorous physical activity (MVPA) on neck circumference (NC), most of them confined to single high-income countries. The present study investigated the association of accelerometer-determined MVPA with NC in adolescents and adults from eight Latin American countries, which are mostly upper-middle income countries.MethodsThe sample consisted of 2370 participants (47.8% male) from the Latin American Study of Nutrition and Health, a multicenter cross-sectional nutrition and health surveillance study of a nationally representative sample from eight Latin American countries (Argentina, Brazil, Chile, Colombia, Costa Rica, Ecuador, Peru, and Venezuela). Times (min/day) in MVPA (defined as time accumulated at ≥1952 activity counts/min) was assessed by ActiGraph GT3X+ accelerometer over 7 days. NC for adolescent was categorized as abnormal if circumference was > 34.5 cm for boys and > 31.25 for girls, whereas for adults the cut-off points for abnormal were > 39 cm for men and > 35 cm women. Multilevel logistic models, including country and region as random effects and adjusted for sex, age, socioeconomic level, and educational level, were used to study the association between MVPA and NC.ResultsThe average time of MVPA was 34.88 min/day, ranging from 31.16 in Venezuela to 40.27 in Chile. Concerning NC, 37.0% of the sample was classified as having elevated NC. Chile was the country with the highest percentage of people with elevated NC (56.9%), and Colombia had the lowest percentage (24.8%). Overall, the MVPA (min/day) was associated with elevated NC (OR = 0.994, CI95% = 0.990-0.998). In Costa Rica and Peru, there were significant associations between MVPA and NC when analyzed by country.ConclusionsThe present study provided evidence of significant associations between MVPA and NC in adolescents and adults from Latin America, independent of sex, age, socioeconomic level, and educational level. This analysis of accelerometry data and NC represents the first examination of these associations in eight Latin America countries. Further research is required to understand the differences between countries in the observed associations.Trial registrationClinicalTrials.Gov NCT02226627 . Retrospectively registered on August 27, 2014.

Project description:IntroductionThe increased availability of influenza surveillance data in recent years justifies an actual and more complete overview of influenza epidemiology in Latin America. We compared the influenza surveillance systems and assessed the epidemiology of influenza A and B, including the spatio-temporal patterns of influenza epidemics, in ten countries and sub-national regions in Latin America.MethodsWe aggregated the data by year and country and characteristics of eighty-two years were analysed. We calculated the median proportion of laboratory-confirmed influenza cases caused by each virus strain, and compared the timing and amplitude of the primary and secondary peaks between countries.Results37,087 influenza cases were reported during 2004-2012. Influenza A and B accounted for a median of 79% and, respectively, 21% of cases in a year. The percentage of influenza A cases that were subtyped was 82.5%; for influenza B, 15.6% of cases were characterized. Influenza A and B were dominant in seventy-five (91%) and seven (9%) years, respectively. In half (51%) of the influenza A years, influenza A(H3N2) was dominant, followed by influenza A(H1N1)pdm2009 (41%) and pre-pandemic A(H1N1) (8%). The primary peak of influenza activity was in June-September in temperate climate countries, with little or no secondary peak. Tropical climate countries had smaller primary peaks taking place in different months and frequently detectable secondary peaks.ConclusionsWe found that good influenza surveillance data exists in Latin America, although improvements can still be made (e.g. a better characterization of influenza B specimens); that influenza B plays a considerable role in the seasonal influenza burden; and that there is substantial heterogeneity of spatio-temporal patterns of influenza epidemics. To improve the effectiveness of influenza control measures in Latin America, tropical climate countries may need to develop innovative prevention strategies specifically tailored to the spatio-temporal patterns of influenza in this region.

Dataset Information

Improved Real-Time Influenza Surveillance: Using Internet Search Data in Eight Latin American Countries.

Background

Objective

Methods

Results

Conclusions

Publications

Improved Real-Time Influenza Surveillance: Using Internet Search Data in Eight Latin American Countries.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets