Background
Dementia represents a significant burden on healthcare systems and prevalence is expected to increase rapidly due to population ageing. Australia is lacking up-to-date information on dementia prevalence and incidence, challenging health service planners.
Aim
This study uses multiple-linked administrative datasets to measure dementia in a large cohort of older people.
Methods
The 45 and Up Study collected baseline survey data (2006-2009) for 266,028 participants aged 45 years and over in New South Wales1. These data were linked with: deaths, hospitalisations2, Pharmaceutical Benefits Scheme (PBS) claims3 and Aged Care data4 for the period 2006-2014. Dementia was estimated from diagnosis codes within hospitalisations/ aged care assessments and dementia-specific medication claims. Age-specific incidence and prevalence were calculated.
Results
9110 cases of dementia were identified in the cohort. Age-specific incidence rates ranged from 0.3 per 1000 person-years in those 55-59 years old to 72.6 per 1000 person-years in those 90+ years old. Estimated prevalence of dementia was 0.1% and 18.9% in persons aged 55-59 and 90+ years, respectively. Incidence and prevalence were lower than published estimates, for example in people aged 85-89 years, the calculated crude incidence was about 80% of the global estimate.5 Pharmaceutical data were important for detecting younger-onset dementia (under 65 years).
Conclusions
This study demonstrates the feasibility of using linked administrative data to measure dementia across a range of ages. The relative importance of different linked datasets varied by age. Further linkages, for example with GP medical records, could improve estimates particularly in younger age groups.