When Google got flu wrong

Butler, Declan

doi:10.1038/494155a

Cited by 468 publications

(289 citation statements)

References 0 publications

Supporting

Mentioning

281

Contrasting

Unclassified

Order By: Relevance

“…In 2009, Google Flu Trends (GFT), a digital disease detection system that uses the volume of selected Google search terms to estimate current influenza-like illnesses (ILI) activity, was identified by many as a good example of how big data would transform traditional statistical predictive analysis (12). However, significant discrepancies between GFT's flu estimates and those measured by the Centers for Disease Control (CDC) in subsequent years led to considerable doubt about the value of digital disease detection systems (13). Although multiple articles have identified methodological flaws in GFT's original algorithm (14)(15)(16) and have led to incremental improvements (14,16) (see also googleresearch.…”

mentioning

confidence: 99%

Accurate estimation of influenza epidemics using Google search data via ARGO

Yang

Santillana

Kou

2015

Proc. Natl. Acad. Sci. U.S.A.

363

443

View full text Add to dashboard Cite

show abstract

mentioning

confidence: 99%

Accurate estimation of influenza epidemics using Google search data via ARGO

Yang

Santillana

Kou

2015

Proc. Natl. Acad. Sci. U.S.A.

363

443

View full text Add to dashboard Cite

show abstract

“…GFT was extended to other countries and its estimates confirmed to be accurate. However, GFT yielded inaccurate data during several periods [19,20]. In 2009, it produced lower estimates at the start of the H1N1 pandemic; in 2013 its estimates were almost twice those from the CDC.…”

Section: Accepted Manuscriptmentioning

confidence: 99%

“…As a result, GFT is currently closed to the public. GFT appeared to be sensitive to uncommon flu epidemics, to media coverage, to changes in the internet users' habits and to modifications of the algorithm in the Google search engine [11,20]. Consequently, other studies proposed to combine traditional surveillance systems and web data, to benefit from the advantages of both systems.…”

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

“…Second, readers may be interested due to indirect reasons such as news coverage. Prior work disagrees on the impact of such influences; for example, Dukic et al found that adding news coverage to their methicillinresistant Staphylococcus aureus (MRSA) model had a limited effect [58], but recent Google Flu Trends failures appear to be caused in part by media activity [94]. Finally, both diseases have a relatively short incubation period (influenza at 1-4 days and dengue at 3-14); soon-to-be-ill readers may be observing the illness of their infectors or those who are a small number of degrees removed.…”

Section: Successful Nowcastingmentioning

confidence: 99%

Global Disease Monitoring and Forecasting with Wikipedia

Generous

Fairchild²,

Deshpande³

et al. 2016

OJPHI

View full text Add to dashboard Cite

Infectious disease is a leading threat to public health, economic stability, and other key social structures. Efforts to mitigate these impacts depend on accurate and timely monitoring to measure the risk and progress of disease. Traditional, biologically-focused monitoring techniques are accurate but costly and slow; in response, new techniques based on social internet data, such as social media and search queries, are emerging. These efforts are promising, but important challenges in the areas of scientific peer review, breadth of diseases and countries, and forecasting hamper their operational usefulness. We examine a freely available, open data source for this use: access logs from the online encyclopedia Wikipedia. Using linear models, language as a proxy for location, and a systematic yet simple article selection procedure, we tested 14 location-disease combinations and demonstrate that these data feasibly support an approach that overcomes these challenges. Specifically, our proof-of-concept yields models with r 2 up to 0.92, forecasting value up to the 28 days tested, and several pairs of models similar enough to suggest that transferring models from one location to another without retraining is feasible. Based on these preliminary results, we close with a research agenda designed to overcome these challenges and produce a disease monitoring and forecasting system that is significantly more effective, robust, and globally comprehensive than the current state of the art.

show abstract

When Google got flu wrong

Cited by 468 publications

References 0 publications

Accurate estimation of influenza epidemics using Google search data via ARGO

Accurate estimation of influenza epidemics using Google search data via ARGO

Leveraging hospital big data to monitor flu epidemics

Global Disease Monitoring and Forecasting with Wikipedia

Contact Info

Product

Resources

About