University of Edinburgh

Dr Luna De Ferrari

National Institute for Health and Care Research

DL_2022_029

Increasing numbers of people have multiple long-term conditions (MLTC). A real point of risk for people with MLTC is when admitted to hospital, where the risks of harm are higher for this group. In this project, we want to use advanced computing techniques to understand the journeys of patients through hospital, using the electronic records made when people are moved between wards or specialist areas. This will highlight areas where delays are experienced, and also allow us to compare very clear pathways (e.g. a broken hip needing surgery) with more complex ones (e.g. a patient admitted with confusion). This mapping work will allow us to better understand key pressure points in the hospital system for people with MLTC. Later in our research programme, we will use this work to develop new tools to predict the risk of harm when admitted to the hospital. These tools could be useful for clinicians to develop better services and care pathways for patients in the future.

Almost half of adults and a large majority of older people have multimorbidity (multiple long-term conditions). People with multimorbidity are the both the highest users of healthcare and the most likely to be harmed by medication-related and other healthcare-related adverse events such as falls and delirium. Our research programme has been developed in partnership with members of the public with experience of multimorbidity and/or caring, and with healthcare professionals. The ways people with multimorbidity travel through hospital care are complex. In this project, we want to use AI techniques grounded in process mining, to better understand these pathways. This is widely used outside of healthcare to understand complex systems. Understanding these pathways will highlight key problem pathways, such as those with delays in reaching appropriate places of care. While this project will be useful for understanding the delivery of care in our hospitals, this is the first stage in our programme to develop prediction tools to identify people at the highest risk of serious adverse events in hospital, which we would ultimately want to translate into clinical care.

Public Health Research

11/08/2023

De-Personalised

TRE