Influenza Forecasting with Google Flu Trends

Dugas, Andrea; Jalalpour, Mehdi; Gel, Yulia R.; Levin, Scott R.; Torcaso, Fred; Igusa, Takeru; Rothman, Richard E.

doi:10.1371/journal.pone.0056176

Cited by 304 publications

(196 citation statements)

References 31 publications

Supporting

Mentioning

190

Contrasting

Unclassified

Order By: Relevance

“…Moreover, it is possible to perform long-term forecasting of visitor flow for individual health centers (or emergency departments) in the same way. This prediction can be used to estimate the centers' workload and revenue and to prevent overcrowding [17,28].…”

Section: Discussionmentioning

confidence: 99%

“…After non-adaptive regression, autoregressive models are the second most popular option [17,4,2,20,18,1,56,44,55,68,69,50,74,28]. This type of model is commonly used in the Box-Jenkins methodology [7] and allows the use of the data correlation structure.…”

Section: Methodsmentioning

confidence: 99%

“…However, ARIMA requires a large number of observations and certain assumptions about the time series. Hence, alternative autoregressive approaches are also used [50,74,28]. For instance, integer autoregressive models yield good results for a short morbidity series [50].…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Long-Term Forecasting of Influenza-Like Illnesses in Russia

Kondratyev

Цыбалова²

2013

Int. J. of Pure and Appl. Math.

View full text Add to dashboard Cite

This paper compares the feasible methods for the long-term forecasting of the incidence rates of influenza-like illnesses (ILI) and acute respiratory infections (ARI), which is important for strategic management. A literature survey shows that the most appropriate techniques for long-term ILI & ARI morbidity projections are the following well-known statistical methods: simple averaging of observations, point-to-point linear estimates, Serfling-type regression models, autoregressive models such as autoregressive integrated moving average (ARIMA) models, and generalized exponential smoothing using the Holt-Winters approach. Using these methods and official data on the total number of ILI & ARI cases per week in 2000-2012 in Moscow, St. Petersburg, Novosibirsk, Yekaterinburg, Nizhny Novgorod and Yakutsk, we developed oneyear projections and evaluated their accuracy. Different methods yielded the best results, depending on the time series. Generally, it is preferable to use the Serfling model. The Serfling model forecasts almost matched the pointto-point linear estimates. In certain cases, ARIMA outperformed the Serfling model. Simple averaging can ensure a fairly good prediction when the ILI & ARI time series do not exhibit a trend. The results of exponential smoothing were poorer than those of other techniques.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Long-Term Forecasting of Influenza-Like Illnesses in Russia

Kondratyev

Цыбалова²

2013

Int. J. of Pure and Appl. Math.

View full text Add to dashboard Cite

show abstract

“…Since its launch in the United States in 2008, GFT predictions have proven to be very accurate when compared to CDC reports. Moreover, GFT data are available 7-10 days before those of the CDC [12]. GFT was extended to other countries and its estimates confirmed to be accurate.…”

Section: Accepted Manuscriptmentioning

confidence: 94%

“…Many studies have assessed the use of internet-user activity data because they can produce real-time indicators [10][11][12][13][14][15][16][17][18]. Several data sources have been explored, including Wikipedia, Twitter or Google search-engine data.…”

Section: Accepted Manuscriptmentioning

confidence: 99%

Leveraging hospital big data to monitor flu epidemics

Bouzillé

Poirier

Campillo‐Gimenez

et al. 2018

Computer Methods and Programs in Biomedicine

View full text Add to dashboard Cite

Background and ObjectiveInfluenza epidemics are a major public health concern and require a costly and time-consuming surveillance system at different geographical scales. The main challenge is being able to predict epidemics. Besides traditional surveillance systems, such as the French Sentinel network, several studies proposed prediction models based on internet-user activity. Here, we assessed the potential of hospital big data to monitor influenza epidemics. MethodsWe used the clinical data warehouse of the Academic Hospital of Rennes (France) and then built different queries to retrieve relevant information from electronic health records to gather weekly influenza-like illness activity. ACCEPTED MANUSCRIPT A C C E P T E D M A N U S C R I P T ResultsWe found that the query most highly correlated with Sentinel network estimates was based on emergency reports concerning discharged patients with a final diagnosis of influenza (Pearson's correlation coefficient (PCC) of 0.931). The other tested queries were based on structured data (ICD-10 codes of influenza in Diagnosis-related Groups, and influenza PCR tests) and performed best (PCC of 0.981 and 0.953, respectively) during the flu season 2014-15. This suggests that both ICD-10 codes and PCR results are associated with severe epidemics. Finally, our approach allowed us to obtain additional patients' characteristics, such as the sex ratio or age groups, comparable with those from the Sentinel network. ConclusionsHospital big data seem to have a great potential for monitoring influenza epidemics in near realtime. Such a method could constitute a complementary tool to standard surveillance systems by providing additional characteristics on the concerned population or by providing information earlier. This system could also be easily extended to other diseases with possible activity changes.Additional work is needed to assess the real efficacy of predictive models based on hospital big data to predict flu epidemics.

show abstract

Development of a Machine Learning Model Using Multiple, Heterogeneous Data Sources to Estimate Weekly US Suicide Fatalities

et al. 2020

View full text Add to dashboard Cite

IMPORTANCE Suicide is a leading cause of death in the US. However, official national statistics on suicide rates are delayed by 1 to 2 years, hampering evidence-based public health planning and decision-making. OBJECTIVE To estimate weekly suicide fatalities in the US in near real time. DESIGN, SETTING, AND PARTICIPANTSThis cross-sectional national study used a machine learning pipeline to combine signals from several streams of real-time information to estimate weekly suicide fatalities in the US in near real time. This 2-phase approach first fits optimal machine learning models to each individual data stream and subsequently combines predictions made from each data stream via an artificial neural network. National-level US administrative data on suicide deaths, health

show abstract

Influenza Forecasting with Google Flu Trends

Cited by 304 publications

References 31 publications

Long-Term Forecasting of Influenza-Like Illnesses in Russia

Long-Term Forecasting of Influenza-Like Illnesses in Russia

Leveraging hospital big data to monitor flu epidemics

Development of a Machine Learning Model Using Multiple, Heterogeneous Data Sources to Estimate Weekly US Suicide Fatalities

Contact Info

Product

Resources

About