Real-time analysis of health data: from revolutionising personalised medicine to empowering public health (Online)

Date: Monday 27 November 2023, 12.30PM
Location: Online
Section Group Meeting

Share this event

Join our forthcoming seminar to explore the transformative impact of real-time health data in personalised medicine and syndromic surveillance. The capacity to harness, store, and analyse real-time health data from sources such as electronic health records, electronic prescription services, and health-related mobile apps is reshaping personalised healthcare, while real-time data from primary care providers, emergency departments, and ambulance services, is being widely utilized for the early identification of potential health threats that may necessitate public health interventions. Our expert speakers will introduce and explore cutting-edge statistical methodologies for analysing real-time data on both individual and population levels.
Ken Cheung: Monotonicity is often a reasonable assumption in regression and statistical learning in health systems. While monotone regression or isotonic regression has a long history since the pool-adjacent-violators algorithm (PAVA), current monotone regression methods are limited to relatively low dimensional problem. In this talk, I will introduce a Bayesian monotone regression method, called iPIPE, which is applicable to situations where the dimension is higher than what the isotonic regression literature typically considers.  I will describe applications of iPIPE in notification optimization in mHealth, cervical cancer screening using PCR assays, and risk prediction using EHR. 
Angela Noufaily: Public health authorities can provide more effective and timely interventions to protect populations during health events if they have effective multi-purpose surveillance systems. These systems rely on aberration detection algorithms to identify potential threats within large datasets. We discuss the evolution of the Quasi-Poisson regression-based Farrington Flexible algorithm used by the UK Health Security Agency for weekly monitoring of laboratory-confirmed infectious diseases, and compare it with the ‘rising activity, multilevel mixed effects, indicator emphasis’ (RAMMIE), UKHSA’s daily syndromic surveillance algorithm, and the ‘Early Aberration Reporting System’ (EARS) method used at the US Centre for Disease Control and Prevention.
Prof Ken Cheung, Columbia University: iPIPE: Bayesian monotone regression
Dr Angela Noufaily, University of Warwick: Regression-based statistical aberration detection algorithms for case count outcomes
Contact Rute Vieira
Non-Fellows                     £20
Fellows                             £15
CStats                              £12
Concessionary Fellows   £10