Are hospital outcomes worse for people living with dementia? : An analysis of Hospital Episode Statistics

Reeves, David and Hann, Mark and Holland, Fiona and Ahmed, Faraz and Morbey, Hazel and Keady, John and Reilly, Siobhan (2019) Are hospital outcomes worse for people living with dementia? : An analysis of Hospital Episode Statistics. In: 29th Alzheimer Europe Conference - Making Valuable Connections, 2019-10-23 - 2019-10-25.

Abstract

Concern over the quality of care provided to people living with dementia during hospital stays has led to a wide range of initiatives to improve care. A key indicator frequently referenced as evidence of poorer care is that people living with dementia have longer mean stays in hospital compared to other patients. However, previous analyses of hospital spell data have only weakly controlled for demographic, health and other differences between the groups. Using routine data (Hospital Episode Statistics) for hospital admissions in 2016/17 for ~400,000 people living with dementia and ~2million people without across 187 large acute hospitals, we constructed a range of patient and hospital spell-level variables including demographics (age, sex, ethnicity, residential and income deprivation); spell characteristics (type of admission, time of year and day of week, discharge destination and delay); pre-existing health (comorbidities, previous admissions); and reason for admission and treatment burden. People living with dementia accounted for 15% of all hospital admissions. They were on average older (75% aged 80 or older versus 40%), more likely to be female (58% versus 51%), to be admitted as an emergency (96% versus 78%), be discharged to a care home (8% versus 2%), and have different conditions for admission but similar overall numbers of comorbidities. Overall, people with dementia spent on average 3.6 days longer in hospital (mean length of stay 9.5 days compared to 5.9), but after controlling for confounding factors this reduced to 0.5 days, with the remaining difference possibly accounted for by residual confounding. Longer hospital stays reported for people living with dementia can be mostly or fully accounted for by pre-admission differences in patient characteristics and health, rather than differences in the hospital care received. The analysis is currently being extended to examine other patient outcomes including hospital mortality, emergency re-admissions and hospital-acquired conditions.