1223 IDENTIFYING SCOTLAND'S CARE HOME POPULATION LIVING WITH DEMENTIA - IS DATA LINKAGE USEFUL?

Abstract

Abstract Background People living with dementia are more likely to move into care homes. The true prevalence of dementia among care home residents in Scotland is not known. People living with dementia often interact with multiple social and healthcare services, thus routine data may offer a way to enhance understanding. Aim To compare national health and social care data sources recording dementia status for Scottish care home residents. Methods A retrospective cohort study of adult (≥ 18 years) care home residents in Scotland during financial years 2012/13 and 2013/14. An indexing process linked data from the Scottish Care Home Census (SCHC) to Community Health Index numbers to allow linkage to healthcare datasets. Anonymised individual data was accessed in a secure environment, within the National Safe Haven. A linked dataset with acute/general and psychiatric hospitalisations (SMR01, SMR04), prescriptions (Prescribing Information System), Scottish Patients at Risk of Admission and Readmission (SPARRA) data, and National Records of Scotland (NRS) mortality records was analysed. Dementia recording was studied across these datasets. Results In 2012/13 and 2013/14, 31,589 and 31,504 care home residents were included for analysis. In 2012/13, 17,548 (55.5%) had dementia according to SCHC. PIS and SMR01 confirm 4,701 (26.8%) and 4,254 (24.3%) SCHC dementia records, respectively. SMR04 and SPARRA confirm 1,830 (10.4%) and 964 (5.5%). Among 2012/13 residents, 19,593 (62.0%) have at least one dementia record across datasets. Of these, 10,445 (53.3%) have one record – 83.9% SCHC records, 7.3% SMR01 records, and 5.0% PIS records. Of 15,781 residents who die within 5 years from 2012/13, 6,984 (44.3%) have death records confirming dementia. Results for 2013/14 are similar. Conclusion Routine data enhances dementia ascertainment amongst care home residents, with most confirmation from general hospitalisations and prescriptions. Primary care data and analysis of more financial years would enable further exploration of dementia recording patterns.