Dynamic clinical prediction models for cardiac surgery

Hickey, Graeme ORCID: 0000-0002-4989-0054, Grant, Stuart W, Caiado, Camila C, Kendall, Simon, Dunning, Joel, Poullis, Mike, Buchan, Iain and Bridgewater, Ben
(2013) Dynamic clinical prediction models for cardiac surgery. 2013, Brighton, UK.

This is the latest version of this item.

[img] Text
Presentation - 469.pdf - Unspecified

Download (1MB)


OBJECTIVES Over its lifespan, EuroSCORE became systematically miscalibrated due to a continuous fall in observed mortality despite patients becoming relatively more high-risk. We aimed to explore some potential frameworks for fitting prediction models for in-hospital mortality following cardiac surgery that dynamically adjust for case-mix in a heterogeneous patient population, and compare these to the standard application of static prediction models. METHODS Data from the Society for Cardiothoracic Surgery in Great Britain and Ireland database were analyzed for procedures performed at all NHS and some private hospitals in England and Wales between April-2001 to March-2011. The study outcome was all-cause in-hospital mortality. Four cross-sectional multiple logistic regression models were fit ranging from static to dynamic generalized linear modelling. Covariate adjustment was made using risk factors included in the logistic EuroSCORE prediction model. RESULTS The association between in-hospital mortality and the risk factors varied with time. Notably, the intercept coefficient has been steadily decreasing over the study period consistent with decreasing observed mortality. Some risk factors such as extracardiac arteriopathy and chronic pulmonary disease has been relatively stable overtime, whilst females have been associated with higher risk relative to the static model. CONCLUSIONS It is known that prediction models can lose calibration. Periodic model updating is necessary but may be better implemented using a less arbitrary modelling approach such as dynamical modelling.

Item Type: Presentation Material
Depositing User: Symplectic Admin
Date Deposited: 27 May 2015 10:07
Last Modified: 17 Dec 2022 01:45
URI: https://livrepository.liverpool.ac.uk/id/eprint/2012100

Available Versions of this Item