Projecting incidence of post-stroke cognitive impairment and dementia in the Irish population aged 40–89 years in 2015 to 2025

Abstract

BackgroundPost-stroke cognitive impairment (PSCI) is a common consequence of stroke, leading to reduced quality of life and increased care needs. However, cognitive impairment receives less attention in stroke rehabilitation relative to physical disability. We aimed to apply estimates of PSCI incidence to the Irish population and project the number of patients who could potentially benefit from cognitive rehabilitation. MethodsWe developed the StrokeCog deterministic model to estimate incidence of PSCI in the population aged 40–89 years living in Ireland in 2015, and project cumulative incidence in this cohort over 10 years. Population data, estimates, and projections to 2025 were obtained from the Central Statistics Office. Age-specific and sex-specific stroke incidence was estimated with 2015 stroke hospital discharge data (n=6155). Transition probabilities across five health states defined by cognitive impairment, physical disability, and dementia were estimated with data from participants reporting stroke in the English Longitudinal Study on Ageing (n=523). An annual stroke recurrence risk of 5% was assumed. FindingsProjections show that the Irish population aged 40–89 years in 2015 (2·05 million) will have a cumulative incidence of stroke of 3·7% by 2025 (74 948), of whom 19 082 (25·5%) will die from stroke and 20 580 (27·5%) from another cause. Of 35 287 survivors, 17 614 (49%) are predicted to have cognitive impairment without dementia, and 8677 (24·6%) to have dementia. InterpretationIn 2025, three quarters of Irish people who have survived a stroke in the preceding 10 years will have cognitive impairment and could potentially have benefited from cognitive rehabilitation. The model will be developed further to project yearly incidence and prevalence of PSCI in the Irish population to 2046, and to include a probabilistic sensitivity analysis to allow for uncertainty in any estimates; it will also be used to evaluate cost-effectiveness of cognitive rehabilitation. The model could also be adapted for use with UK populations. FundingHealth Research Board in Ireland (grant number ICE-2015-1048 and award RL-15-1579).