Comparison of statistical algorithms for the detection of outbreaks
PHE ePoster Library. Morbey R. 09/12/17; 186479; 147
Dr. Roger Morbey
Dr. Roger Morbey
Login now to access Regular content available to all registered users.

You may also access this content "anytime, anywhere" with the Free MULTILEARNING App for iOS and Android
Abstract
Rate & Comment (0)
Abstract Big DataThe Real-time syndromic surveillance team (ReSST) monitors anonymised data from GP services, NHS 111 calls and emergency departments to detect threats to public health. This involves over 12,000 daily time series, aggregated into syndromes locally and nationally. Therefore ReSST uses automated statistical detection methods.Detection methods need a high sensitivity and specificity across a wide range of syndromes. The time series we monitor range from most days having a zero count to over 10,000 a day. Also, seasonality and day of the week effects vary by syndrome and data source. Furthermore, outbreaks can vary from one day spikes to gradual rises lasting over a year. Therefore, we have compared methods across a range of scenarios.Best PracticeWe have compared three methods; the 'RAMMIE' method used by ReSST, the 'flexible Farrington' method used by PHE for laboratory surveillance and the cumulative sum method popular in other countries. We modelled synthetic syndromic datasets based on our historical data. Thus, we were able to create scenarios with different; seasonality, day of week effects, data volume and outbreaks. We then used simulations to estimate sensitivity and specificity.We will describe which method worked best for each scenario. Subsequently we can ensure that ReSST is using the best detection method for each syndrome. Also our results will have wider application for laboratory surveillance and for other countries. Furthermore, researchers can use our synthetic datasets to test any new methods developed in the future.
Abstract Big DataThe Real-time syndromic surveillance team (ReSST) monitors anonymised data from GP services, NHS 111 calls and emergency departments to detect threats to public health. This involves over 12,000 daily time series, aggregated into syndromes locally and nationally. Therefore ReSST uses automated statistical detection methods.Detection methods need a high sensitivity and specificity across a wide range of syndromes. The time series we monitor range from most days having a zero count to over 10,000 a day. Also, seasonality and day of the week effects vary by syndrome and data source. Furthermore, outbreaks can vary from one day spikes to gradual rises lasting over a year. Therefore, we have compared methods across a range of scenarios.Best PracticeWe have compared three methods; the 'RAMMIE' method used by ReSST, the 'flexible Farrington' method used by PHE for laboratory surveillance and the cumulative sum method popular in other countries. We modelled synthetic syndromic datasets based on our historical data. Thus, we were able to create scenarios with different; seasonality, day of week effects, data volume and outbreaks. We then used simulations to estimate sensitivity and specificity.We will describe which method worked best for each scenario. Subsequently we can ensure that ReSST is using the best detection method for each syndrome. Also our results will have wider application for laboratory surveillance and for other countries. Furthermore, researchers can use our synthetic datasets to test any new methods developed in the future.
    This eLearning portal is powered by:
    This eLearning portal is powered by MULTIEPORTAL
Anonymous User Privacy Preferences

Strictly Necessary Cookies (Always Active)

MULTILEARNING platforms and tools hereinafter referred as “MLG SOFTWARE” are provided to you as pure educational platforms/services requiring cookies to operate. In the case of the MLG SOFTWARE, cookies are essential for the Platform to function properly for the provision of education. If these cookies are disabled, a large subset of the functionality provided by the Platform will either be unavailable or cease to work as expected. The MLG SOFTWARE do not capture non-essential activities such as menu items and listings you click on or pages viewed.


Performance Cookies

Performance cookies are used to analyse how visitors use a website in order to provide a better user experience.


Save Settings