A Bayesian Outbreak Detection Method for Influenza-Like Illness

García, Yury E.; Christen, J. Andrés; Capistrán, Marcos A.

doi:10.1155/2015/751738

Cited by 6 publications

(8 citation statements)

References 25 publications

(27 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Outbreaks of new (or old) diseases typically display an initial exponential rate of growth [14]. DUDE compares the set of cases in the monitor window with a baseline set of cases and generates an alarm if the presence of an unmodeled disease better explains the difference than simple variation.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

A Bayesian approach for detecting a disease that is not being modeled

et al. 2020

View full text Add to dashboard Cite

Over the past decade, outbreaks of new or reemergent viruses such as severe acute respiratory syndrome (SARS) virus, Middle East respiratory syndrome (MERS) virus, and Zika have claimed thousands of lives and cost governments and healthcare systems billions of dollars. Because the appearance of new or transformed diseases is likely to continue, the detection and characterization of emergent diseases is an important problem. We describe a Bayesian statistical model that can detect and characterize previously unknown and unmodeled diseases from patient-care reports and evaluate its performance on historical data.

show abstract

Section: Discussionmentioning

confidence: 99%

“…A distinctive feature of an infectious outbreak is an initial period of exponential growth. [14] describes a Bayesian system, based on moving windows of ILI records, to detect a period of exponential growth that can signal the start of an outbreak. That system uses Bayes' factors to determine when it is likely that an outbreak has started.…”

Section: Introductionmentioning

confidence: 99%

A Bayesian approach for detecting a disease that is not being modeled

et al. 2020

View full text Add to dashboard Cite

show abstract

“…To our knowledge, no other study has been conducted on the performance of BODA. However, other algorithms using the Bayesian framework were used to detect outbreaks in various studies, most of which showed good performance (Garcia, Christen, & Capistran, 2015; Liao, Xu, Wang, & Liu, 2017; Texier et al., 2016, 2019). The BODA model has two advantages over traditional models, one using the Bayesian framework, which also uses the prior distribution to calculate the alarm threshold.…”

Section: Discussionmentioning

confidence: 99%

Performance of Bayesian outbreak detection algorithm in the syndromic surveillance of influenza‐like illness in small region

Aghaali

Kavousi

Shahsavani

et al. 2020

Transbound Emerg Dis

View full text Add to dashboard Cite

Early warning for Infectious disease outbreak is an important public health policy concern, and finding a reliable method for early warning remains one of the active fields for researchers. The purpose of this study was to evaluate the performance of the Bayesian outbreak detection algorithm in the surveillance of influenza‐like illness in small regions. The Bayesian outbreak detection algorithm (BODA) and modified cumulative sum control chart algorithm (CUSUM) were applied to daily counts of influenza‐like illness in Tehran, Iran. We used data from September 2016 through August 2017 to provide background counts for the algorithms, and data from September 2017 through August 2018 used for testing the algorithms. The performances of the BODA and modified CUSUM algorithms were compared with the results coming from experts' signal inspections. The data of syndromic surveillance of influenza‐like illness in Tehran had a median daily counts of 7 (IQR = 3–14). The data showed significant seasonal trends and holiday and day‐of‐the‐week effects. The utility of the BODA algorithm in real‐time detection of the influenza outbreak was better than the modified CUSUM algorithm. Moreover, the best performance was when a trend included in the analysis. The BODA algorithm was able to detect the influenza outbreaks with 4–5 days delay, with the least false‐positive alarm. Applying the BODA algorithm as an outbreak detection method in influenza‐like syndromic surveillance might be useful in early detection of the outbreaks in small regions.

show abstract

“…In this paper q = 0.4 is used to get 90% Confidence Interval by Equation 7• Model 4: Growth Value The fourth model is using RI growth value as a measurement to identify abrupt growing RIs. The anomaly label using growth value is given in Equation (8). It means the labeled weeks have RI growth value exceeding the median of the positive growth values.…”

Section: • Model 1: Simple Thresholdmentioning

confidence: 99%

“…In 2015, García YE utilized Bayesian model selection and Bayesian regression to detect outbreaks for ILI using surveillance data 8 . The method was applied to both the Spanish influenza outbreaks in USA San Francisco in 1918 and the acute respiratory illnesses (ARI) from Mexico San Luis Potosí for validation.…”

Section: Introductionmentioning

confidence: 99%

A Novel Data-Driven Methodology for Influenza Outbreak Detection and Prediction

Yan²

2020

SSRN Journal

View full text Add to dashboard Cite

Influenza is an infectious disease that leads to an estimated 5 million severe illness cases and 650,000 respiratory deaths worldwide each year. Early detection and prediction of influenza outbreaks are crucial to efficient resource planning to save patients' lives and healthcare costs. This paper proposes a novel data-driven methodology for influenza outbreaks detection and prediction. The doctor's diagnosis-based prescription dataset of Influenza-Like Illness (ILI) from more than 3,000 clinics in Malaysia is used in this study because the prescription data are reliable and can be captured timely. A new Region Index (RI) of the influenza outbreak is proposed based on the prescription dataset. With the newly proposed RI metric, statistical and machine learning models are developed to detect and predict influenza outbreaks. Cross-validation is conducted to evaluate the prediction model performance. The proposed methods are also validated by real-world evidence. It is proved to be sensitive and accurate in influenza outbreak prediction with 80 − 90% accuracy, 70 − 80% recall, and 70 − 80% precision scores.

show abstract

A Bayesian Outbreak Detection Method for Influenza-Like Illness

Cited by 6 publications

References 25 publications

A Bayesian approach for detecting a disease that is not being modeled

A Bayesian approach for detecting a disease that is not being modeled

Performance of Bayesian outbreak detection algorithm in the syndromic surveillance of influenza‐like illness in small region

A Novel Data-Driven Methodology for Influenza Outbreak Detection and Prediction

Contact Info

Product

Resources

About