Delirium involves a sudden onset of confusion, memory loss, inability to focus and difficulty speaking. It is relatively common in older patients in the community, with approximately 200,000 cases every year in the UK. Delirium often precipitates hospital admission and results in loss of quality of life, and increased death rates and healthcare costs. Most research on delirium has focused on hospitalized patients. However there is little data on delirium in older people living in the community and no trials of prevention have been reported in primary care.
We will use anonymised patient medical records to improve our understanding of the causes of delirium in the community and to develop tools to help prevent delirium. Our main aim is to develop tools to help general practices identify patients at higher risk of delirium, so that risk factors (e.g. certain medications) can be avoided. These prediction tools will also help to target information to carers about how to cope if an episode of delirium does develop. Such tools would be of great value for aiding GPs to provide personalised care to their patients at risk of delirium. This project is funded by the National Institute for Health Research.
Delirium often develops following potentially avoidable triggers, especially in patients with cognitive impairments. Risk factors in hospital settings are well studied and evidence suggests that interventions can reduce delirium incidence by up to 35% in hospital settings. In contrast, avoidable factors in community acquired delirium have been little studied. Electronic clinical records, such as the NIHR part-funded Clinical Practice Research Database (CPRD) provide cost-efficient opportunities for studying delirium in the community. We will develop a tool to help general practices identify patients at higher risk of delirium. A case-control analysis, using conditional logistic regression, will identify factors associated with delirium. Prediction models will be trained on the dataset ending in 2014, and validated on more recent (2015 onward) data, using established prediction performance measures. These prediction tools will also help to target information for carers about how to cope if an episode of delirium does develop. The tools will aim to identify those at higher risk of first episodes of delirium, and also those who have had previous episodes. This work is supported by NIHR Research for Patient Benefit project funding (PB-PG-1215-20022).
Incident delirium
- All-cause mortality
- Health-care use (hospitalisations and length of stay, GP visits, medication, tests)
David Melzer - Chief Investigator - University of Exeter
Joao Delgado - Collaborator - University of Exeter
Joe Butchart - Collaborator - University of Exeter
Jose M Valderas - Collaborator - University of Exeter
Kirsty Bowman - Collaborator - University of Exeter
Lindsay Jones - Collaborator - University of Exeter
Rodney Stephen Taylor - Collaborator - University of Exeter
Ruben Mujica Mota - Collaborator - University of Exeter
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation