Centre for Primary Care Seminar - Charles Reynard - Advanced cardiovascular risk prediction in the acute care setting: Protecting and improving clinical prediction models
|Starts:||14:00 24 Sep 2019|
|Ends:||15:00 24 Sep 2019|
|What is it:||Seminar|
|Organiser:||Faculty of Biology, Medicine and Health|
|Who is it for:||University staff, Current University students|
|Speaker:||Dr Charles Reynard|
Clinical prediction models (CPM) are a corner stone of clinical care in the emergency department. We utilise algorithms to predict mortality in chest infections, risk of blood clots, likelihood of an ankle fracture, and risk of a heart attack. However they all share a flaw, they become less accurate over time through an effect frequently described as 'calibration drift'.
This is thought to be due to changing population demographics, changing behaviour of clinicians and also changing disease incidence. This weakness stems in part from the static population in which the majority of these algorithms were originally derived.
This project focuses around a chest pain CPM called the Troponin-only Manchester Acute Coronary Syndrome diagnostic algorithm (T-MACS). This algorithm uses a bio-marker, symptoms and an electrocardiograph to immediately rule in or rule out a heart attack (acute myocardial infarction). It is currently being implemented across Greater Manchester and once fully deployed we expect 30,000 patients per year to be entered into the database (GM-TMACS).
In this project we will aim to use the GM-TMACS database to:
(i) identify the optimal methodology for updating CPMs; meta-modelling, one at a time co-efficient updating, or dynamic hierarchical Bayesian updating
(ii) validate the methodology with T-MACS
(iii) seek to simultaneously enhance T-MACS - incorporate geographical location into the CPM, variables from other models, and include a long-term heart health outcome
Dr Charles Reynard
Organisation: School of Medical Sciences
Travel and Contact Information
Seminar Room, Fifth Floor