Abstract
Objective
To estimate the additional impact of dementia on in-patient length of stay (LOS) and related costs in Irish acute hospitals. Both principal and secondary diagnosis effects are estimated and valued.
Methods
This is a cross-sectional study based on administrative data collected on all public hospital in-patient discharges in Ireland for people aged 65 years and older in 2019. Coarsened exact matching (CEM) was undertaken to account for observed confounders between dementia and non-dementia groups, while generalised linear modelling (GLM) was used to compare differences in LOS.
Results
Patients with a principal diagnosis of dementia spent on average 17.5 (CI: 15.42, 19.56; p < .01) d longer in hospital than similar patients with no principal diagnosis of dementia. LOS was 6.7 (CI: 6.31, 7.14; p < .01) d longer for patients with a secondary diagnosis of dementia compared to similar patients with no secondary diagnosis of dementia. The additional annual cost of care for patients in hospitals with a secondary (principal) diagnosis of dementia was €62.0 million (€13.2 million).
Conclusions
This study highlights the economic impact of extended LOS for patients with dementia in Irish acute hospitals. Addressing specific dementia-related needs of people in hospital is likely to optimise resource use and decrease health care costs in acute care settings.
Keywords:
Disclosure statement
No conflict of interest.
Funding
None.
Notes
1 Table S1, in the Supplemental Material, presents key descriptive statistics for each group in Comparisons 1 and 2 before CEM.
2 The base category for admission source is transferred from other source. The base category for age is 65–74 years. The base category for Elixhauser comorbidities is those patients with no comorbidities. ***Denotes significant at 1%.
3 The base category for admission source is transferred from other source. The base category for age is 65-74 years. The base category for Elixhauser comorbidities is those patients with no comorbidities. ***Denotes significant at 1%; **Denotes significant at 5%.
4 ***Denotes significant at 1%.