Patient groups in intensive care units (ICUs) exhibit considerable diversity in terms of their health conditions and demographic profiles. This variance has been inadequately addressed, leading to a notable disparity in applying scientific findings to practical care. Elements contributing to critical illness include pre-disposing factors, such as multimorbidity, polypharmacy, and frailty, and precipitating factors such as diagnosis and illness severity. The complex interplay of these factors in determining patient trajectories and outcomes is poorly understood.
The first aim of the project will be to bring together healthcare datasets to understand the scale of the multimorbidity burden in the critical care population, alongside other pre-disposing and precipitating factors. I will aim to derive distinct patient clusters and explore the relationships between these clusters and patterns of multiple organ dysfunction and outcomes including early and late mortality, care-transitions and hospital readmission. This will help elucidate mechanistic pathways and identify targets for tailored interventions.
The second aim will be to use prediction modelling and causal inference techniques, such as target trial emulation, to identify and explore customised treatments within the different patient clusters identified that can lead to improved outcomes.
