Introduction: Dementia and fall-related hip fractures both contribute significantly to the burden of illness within elderly populations in Australia and elsewhere. The research presented here uses a large probabilistically linked dataset from NSW, Australia to estimate the prevalence of dementia within hip fracture patients and investigate the impact of dementia on hospitalisation length of stay (LOS) and survival.
Method: The cases considered were NSW residents aged 65 years and above who experienced a fall related hip fracture between 1 July 2000 and 30 June 2009. The prevalence of dementia was calculated for the incident hip fracture using two methods to infer dementia status. Cox proportional hazards regression modelling was used to estimate the relative rate of discharge from a hospitalisation episode, and the relative mortality rate of hip fracture patients suffering dementia versus those who were cognitively intact. Additional covariates used in the models included sex, age group at admission, the Charlson Comorbidity Index and separation mode.
Results: Of the 44,143 fall-related incident hip fracture cases considered, between 24% (observed diagnosis) to 29% (inferred diagnosis) of these people had dementia. The median LOS for patients with dementia was shorter than those without dementia, but there was a strong interaction with age. The rate of discharge from the fracture-related hospitalisation episode of the cases with dementia was 40% greater (95% CI 1.4-1.5) than the non-demented group. Similarly, the relative mortality rate of those with dementia was greater (2.4, 95% CI 2.3-2.6) than the non-demented group. Both Cox analyses indicated evidence for main effects of age at admission and comorbidity, as well as interaction effects between age group and dementia status.
Conclusion: The use of linked datasets with tens of thousands of cases enables the calculation of precise estimates of various parameters. People with dementia constitute a significant proportion of the total population of elderly hip fracture patients in hospitals (up to 29%). Their mortality rate is greater than those without a diagnosis of dementia and their hospital length of stay is shorter, particularly if they are discharged to a residential aged care facility.
Copyright © 2012 Elsevier Ltd. All rights reserved.