Measuring the prevalence of 60 health conditions in older Australians in residential aged care with electronic health records: a retrospective dynamic cohort study

Abstract

Background : The number of older Australians using aged care services is increasing, yet there is an absence of reliable data on their health. Multimorbidity in this population has not been well described. A clear picture of the health status of people using aged care is essential for informing health practice and policy to support evidence-based, equitable, high-quality care. Our objective was to describe the health status of older Australians living in residential aged care facilities (RACFs) and develop a model for monitoring health conditions using data from electronic health record systems. Methods : Using a dynamic retrospective cohort of 9436 RACF residents living in 68 RACFs in New South Wales and the Australian Capital Territory from 2014-2017, we developed an algorithm to identify residents’ conditions using: aged care funding assessments; medications administered; and clinical notes from their facility electronic health record (EHR). We generated age and sex-specific prevalence estimates for 60 health conditions. Agreement between conditions recorded in aged care funding assessments and those documented in residents’ EHRs was evaluated using Cohen’s Kappa. Cluster analysis was used to describe combinations of health conditions (multimorbidity) occurring among residents. Results : Using all data sources, 93% of residents had some form of circulatory disease, with hypertension the most common (62%). Most residents (93%) had a mental or behavioural disorder, including dementia (58%) or depression (54%). For most conditions, EHR data identified approximately twice the number of people with the condition compared to aged care funding assessments. Agreement between data sources was highest for multiple sclerosis, Huntington’s disease and dementia. The cluster analysis identified seven groups with distinct combinations of health conditions and demographic characteristics and found that the most complex cluster represented a group of residents that had on average the longest lengths of stay in residential care. Conclusions : The prevalence of many health conditions among RACF residents in Australia is underestimated in previous reports. Aged care EHR data have the potential to be used to better understand the complex health needs of this vulnerable population and can help fill the information gaps needed for population health surveillance and quality monitoring.

Transforming routinely collected residential aged care provider data into timely information: Current and future directions

“…Recent research has demonstrated that routinely collected data from aged care providers can be harnessed to provide valuable, up-to-date and comprehensive monitoring of medication and health conditions. [4][5][6][7][8][9]12,15 For example, one study utilised aged care data to investigate adherence to clinical practice guidelines for the use of antipsychotics in residents with dementia, and demonstrated considerable areas for improvement. 8 Currently, real-time predictive analytics dashboards are being developed that aim to utilise aged care routinely collected data to better identify fall risk, monitor quality of life, support end-oflife care and early detection of health deterioration.…”

“…Recent research has demonstrated the benefit of linking these aged care data sets to assess longitudinal patterns of medication use and compliance with clinical guidelines related to resident's conditions. [4][5][6][7]9,11,15 These data sets also have the potential to be linked across providers and with external community-based service and State/National data sets, enabling a comprehensive overview of residents' use of additional services such as general practitioner visits and hospitalisations.…”

“…For example, information about having cognitive impairment can be recorded at least 15 different ways after standardising text case. 7 Coding algorithms can be written to standardise these free-text fragments for research use. One publicly available algorithm specific to aged care in Australia supports the analysis of free-text data, ACFI data and medication administration data on residents' health conditions, and was developed based on a large and representative sample of care homes.…”

“…New research has demonstrated how to utilise these routinely collected data sets to examine trends and variation in health status of residents, quality of care and adherence to clinical guidelines. [2][3][4][5][6][7][8][9][10][11][12] The aim of this paper was to provide clinicians, researchers, aged care policymakers and aged care providers with an understanding of the strengths of these data sets, as well as identifying areas that require future development to maximise their potential to drive improvements in resident care and outcomes.…”

Increased number of deaths within 24 h of admission during a period of social restriction related to the COVID-19 pandemic: A retrospective service evaluation in a metropolitan palliative care unit

“…This supports the trend we observed with respect more deaths from dementia, given the high prevalence of this disease within aged care facilities. 34 Patients with dementia display negative psychological impacts of social isolation [35][36][37] and may be more vulnerable to them, with detriment to overall health.…”