Clusters of Long-Term Conditions and Adverse Health Outcomes in People with Multimorbidity

Conference:

SAPC ASM 2023 - Brighton

Talk Code:

5C.3

Presenter:

Stefanie J. Krauth

Co-authors:

Stefanie J. Krauth*, Lewis Steel, Sayem Ahmed, Grace Dibben, Peter Hanlon, Jim Lewsey, Barbara Nicholl, David McAllister, Emma McIntosh, Rod S. Taylor, Sally J. Singh, Frances S Mair, Bhautesh D. Jani

Author institutions:

School of Health and Wellbeing, College of Medical, Veterinary and Life Sciences, University of Glasgow ; Department of Respiratory Sciences, University of Leicester

Problem

Multimorbidity, the presence of two or more long-term conditions (LTC) in patients, is becoming the norm rather than the exception in healthcare practice. Yet, it is currently unclear how best to classify multimorbidity beyond counting the number of long-term conditions (LTCs), nor do we understand the relationship of multimorbidity classification (based on types of LTC combinations) with health-care resource use, hospitalisations, or mortality in the UK. A better understanding of clusters of LTCs and their association with health outcomes may be important for the development of effective interventions targeted at patients with multimorbidity.Objectives: 1. To analyse how clusters of LTCs in different age groups relate to adverse health outcomes, including primary care use, number and duration of hospitalisations, and mortality in multimorbid patients. 2. To investigate what additional information, if any, can be gained from using clusters over counts of LTCs in understanding the risk of adverse health outcomes.

Approach

Latent Class Analysis was used to identify clusters of LTCs in different age groups (18-36, 37-55, 56-73, >73) in two large community cohorts: UK Biobank (n= 498,936), and the Secure Anonymised Information Linkage Databank (SAIL) (n= 1,552,084). Latent class membership was assigned using posterior probabilities. Incident rate ratios were computed for the number and duration of hospitalisations, primary care use, and all-cause mortality over a 10-year period, using negative binomial and Poisson regression modelling.

Findings

The clusters that were associated with the most severe adverse health outcomes differ in their LTC-profile between young & middle-aged patients to older & elderly patients. In the two younger age groups, clusters including mental health & pain disorders and discordant multimorbidity as main components had the highest rates of adverse health outcomes whereas clusters including cancer, and complex multimorbidity were more strongly associated with adverse health outcomes in the older age groups. These findings were largely consistent across the two cohorts. After adjusting for number of LTCs, sex, age, deprivation, and lifestyle factors, different clusters of LTCs showed distinct associations with adverse health outcomes in all age groups with significantly differing incident rate ratios (IRR) for hospitalisations, primary care use, and all-cause mortality. IRRs for all-cause mortality in young adults aged 18-36 in SAIL increased with increasing number of LTCs. After adjusting for number of LTC counts, young adults in the cluster “depression & substance abuse” showed an all-cause mortality IRR of 6.42 (p=<0.01) compared to participants without multimorbidity while the cluster “Asthma & Depression” had an all-cause mortality IRR of 1.40 (p<0.01) in the same age group.

Consequences

Our findings suggest that the specific combinations of LTCs may offer additional, clinically, and economically important information over LTC counts alone in the risk stratification of patients with multimorbidity.

Submitted by:

Stefanie J. Krauth

Funding acknowledgement:

This work was funded by the National Institute for Health Research, UK