Understanding the medical determinants and health service needs of older people who experience loneliness in Sydney, Australia

Introduction: In Australia there has been a sharp increase in the proportion of older people who live alone. Living alone increases the risk of being socially isolated or experiencing loneliness. Older people who are lonely have an increased risk of dying sooner and are more likely to experience a decline in their mobility [1]. Understanding if and how social isolation/loneliness impacts on managing health conditions and use of health services is important in providing quality care and preventing premature mortality. Living alone is not necessarily a predictor of social isolation/loneliness, rather lack of time spent with family or friend may be a better indicator. More supportive social relationships are related to a decreased mortality risk [2]. Research has been undertaken on the determinants of isolation/loneliness (individual, social, community and environment), however less emphasis has occurred on the medical determinants and how these might be mitigated. Using our existing data linkage resource, the Central and Eastern Sydney Primary and Community Health Cohort/Resource (CES-P&CH), which includes questionnaire data, primary care records, prescribing information, hospital records, emergency department records, cancer registry, and vital statistics on over 30,000 participants in CES aged 45 years and over we explored patterns of service use in people who are socially isolated. Methods: A record linkage study using 45 and Up Study questionnaire data, MBS claims, hospitalisations and deaths was undertaken. Social isolation was defined using a combination of baseline questionnaire data on living arrangements, family and friend support, and health issues that impacted on work/daily activities. Participant characteristics and health conditions at baseline (2006-2009) and health service use (including GP presentations, care plan use, and number and frequency of hospitalisations over the next 10 years) were compared for those who were defined as socially isolated versus those who were not using multivariate/time series models. Results and Discussion: At baseline 20% of participant lived alone; 30.8% had no partner; 18.5% had no children; 44.3% were not working; 4.9% regularly need help with daily tasks; 12.0% had severe physical limitations; 8.8% did not have support from family or friends; 9.6% were urinary incontinent, and 11.3% had depression or anxiety. These participants were included in the socially isolated group for the analysis if they had multiple risks. This paper will discuss the different algorithms that were developed to describe social isolation. This paper will also provide the results from the multivariate/time series models and how this information is being/can be used to better understand and provide person-centred quality care in CES. Limitations and suggestions for future research: Because the research study used an existing record linkage resource we were limited to the items that were included in the questionnaire to define social isolations. This research study would benefit from sensitivity testing of the resultant social isolation algorithm. References: 1- Perissinotto CM, Stijacic CI, Covinsky KE. Loneliness in older persons: a predictor of functional decline and death. Arch Intern Med 2012;172(14):1078-83. 2- Holt-Lunstad J, Smith TB, Layton JB. Social Relationships and Mortality Risk: A Meta-analytic Review. PLoS Med 2010;7(7): e1000316. doi:10.1371/journal.pmed.1000316