Data Insights: Multiple health conditions and social care
Written by 3 October 2019
ADR Scotland's ‘Data Insights’ briefings share snapshots of findings from research projects using linked administrative data. This week, Scottish Centre for Administrative Data Research (SCADR) Research Fellow David Henderson shares findings from his work on ‘multimorbidity’ and receipt of social care. You can read excerpts below, or the full two-page briefing.
Multimorbidity – the presence of two or more long term conditions – is increasing in prevalence in Scotland. It has several negative outcomes including higher mortality, decreased quality-of-life, decreased functional status, and higher usage of healthcare. Little is known, however, about the relationship between multimorbidity and social care and how sociodemographic factors influence receipt of care.
Our research shows that for those aged over 65 in Scotland multimorbidity is associated with receipt of care, and that the probability of receiving social care increases with severity of multimorbidity. We also found that those living in more deprived areas were more likely to receive social care.
What we did and what we found
We created a cohort of all individuals with a valid Community Health Index (CHI) number aged over 65 in Scotland for the years 2011-2016 (~ 1.1 million people). We linked administrative data records from across the NHS in Scotland and Scottish Government (detail on the specific datasets used is set out in the full briefing).
- Almost all of those receiving social care had multimorbidity (measured by the number of body systems individuals were prescribed medicines for – indicated by chapter codes from the British National Formulary (BNF)). For example, 93% had repeat medicines prescribed from two-or-more BNF chapters and over two-thirds were prescribed repeat medicines from four-or-more BNF chapters.
- After adjusting for age, sex, and deprivation status, the probability of receiving social care increased with severity of multimorbidity. For example, those prescribed medicines from two-or-more BNF chapters had a 4% increased probability and those with medicines from six-or-more BNF chapters had a 19% increased probability of receiving social care compared to those with no prescribed medicines.
- Unsurprisingly, increasing age is associated with receipt of social care. Those aged over 95 had an increased probability of almost 50% of receiving care compared to those aged 65-69.
- Despite accounting for two-thirds of all social care receipt, after adjustment for other variables in our models women were only marginally more likely to receive care social care than men. This difference could itself be explained by the possibility that women are more likely to survive their partners and are therefore more likely to be living alone. We were unable to account for this fact in our models.
- The probability of receiving social care increased with deprivation status (measured by Scottish Index of Multiple Deprivation decile of residence) although the magnitude of this effect was lower than seen for multimorbidity and age factors after adjustment. A lack of a good measure of need means we cannot accurately ascertain if provision matches need.
Why this matters
This is one of the first projects to utilise individual-level data on social care for research purposes. Cross-sectoral linkage to health records offers a wealth of opportunity to learn how health and social care services interact. This study provides previously unavailable information about those that utilise social care services. We have also been able to identify where gaps in data coverage exist (e.g. in measures of social care ‘need’).
The relationship between multimorbidity and social care described by our research is important as it provides the first empirical evidence of this association. Much has been written about the increasing prevalence of multimorbidity and the effect this is having on healthcare services. Our results provide a reminder that there are also large implications for non-health care organisations with knock-on effects for funding and policy.
The second phase of research for this project is now being scoped.