Abstract
Aim: Older adults living with dementia occupy approximately one quarter of acute hospital beds in England. The risk of hospital readmission within six months of discharge increases with comorbidities, reduced mobility, and limited interdisciplinary collaboration across care sectors. Subsequently, hospital readmission can increase the risk of mortality in this population. This study aimed to quantify the clinical determinants of readmission and mortality in older adults with dementia in England.
Method: A retrospective cohort study was conducted using anonymised data from adults in England with a diagnosis of dementia, aged 65 and over. Individuals were identified through primary care electronic health records in the Clinical Practice Research Datalink (CPRD) GOLD, between April 1997 and November 2018. Readmissions within 180 days were identified using linked Hospital Episode Statistics. Adjusted logistic regression assessed factors associated with readmission, and Cox proportional hazards regression identified predictors of one-year mortality following readmission.
Results: The cohort included 24,956 individuals with dementia from 253 primary care centres (mean age 81.93 years; 61.6% female). Chronic obstructive pulmonary disease (odds ratio [OR]=1.26, 95% confidence interval [CI]: 1.15-1.39), type 1 diabetes (OR=1.21, CI: 1.12-1.30), and chronic kidney disease (OR=1.14, CI: 1.07-1.22) were strongly associated with readmission. Medication review in primary care within one year before index hospitalisation (OR=1.08, CI: 1.02-1.14), and primary care consultation within two weeks of discharge (OR=1.21, CI: 1.15-1.28) were associated with readmission. One-year mortality following readmission was associated with age (hazard ratio [HR]=3.20, CI: 2.49-4.11 for ages 90+ versus 65-69), multiple long-term conditions (HR=1.21, CI: 1.06-1.39 for 4-5 conditions versus none), prescriptions for antipsychotic medication (HR=1.37, CI: 1.22-1.53), and care home residence (HR=1.33, CI: 1.10-1.62).
Conclusion: Knowledge of clinical factors associated with readmission and mortality can inform advanced care planning between health and social care professionals, people living with dementia and their families.
Comments
Thank you for this…
Thank you for this interesting and well presented poster. It's great to see such a large dataset used. As the data spans over 20 years, I was wondering if the factors associated with readmission and mortality changed over time or if they remained consistent over the 20 year period?
Thank you
Thank you for your comment.
The aim was to identify the overall factors associated with readmission and subsequent mortality across the entire study period. Although the study spans about 20 years, individuals had an average of around 1.5 person-years from their index hospital admission to the end of follow-up. So the models capture the overall associations rather than changes in these associations over time. Additional analyses would be needed for this, which is feasible to do in future work.