Characteristics, service use and mortality of clusters of multimorbid patients in England: a population-based study

Talk Code: 
3C.1
Presenter: 
Rupert Payne
Co-authors: 
Yajing Zhu, Duncan Edwards, Jonathan Mant, Steven Kiddle
Author institutions: 
University of Bristol, University of Cambridge

Problem

Multimorbidity is one of the principal challenges facing health systems worldwide. To help understand the changes to services and policies that are required to deliver better care, we investigated which diseases co-occur and how combinations are associated with mortality and service use.

Approach

We used linked primary and secondary care electronic health records contributed by 382 English general practices to the Clinical Practice Research Datalink (CPRD). The study includes a representative set of multimorbid adults (18+ years old) with two or more long-term conditions (N=113,211). Clinical diagnoses were based on a list of 38 conditions (Barnett, 2012). A random set of 80% of the multimorbid patients (N=90,571) were stratified by age and clustered using latent class analysis, with consistency of results checked in the remaining 20% of multimorbid patients (N=22,640). Associations between multimorbidity clusters, demographics and outcomes were quantified using generalised linear models.

Findings

Multimorbid patients aged 18-44 years old with psychoactive substance misuse were found to have a mortality rate 18 times higher than their non-multimorbid peers. Three distinct types of cardiovascular-related clusters were identified in patients 65+ years old, related to cardiovascular risk factors, established cardiovascular disease, or cardiovascular disease with chronic pain and mental illness; the latter cluster had the highest rates of primary care consultations. The majority of 85+ year old multimorbid patients belonged to a cluster with low service use and mortality in that age range. Clusters characterised by physical-mental health co-morbidity, and respiratory disease with co-morbidity, were common across all age strata. A physical-mental health cluster of 65-85 year olds had a higher 2-year mortality rate (8.4%) than a cardiovascular risk factor cluster (4.7%).

Consequences

This work has highlighted major targets for public health and healthcare, including younger patients with psychoactive substance misuse whose risk factors make them potentially amenable to intervention. The association between older age, multimorbidity and mortality seems to be related to cardiovascular disease, pain, respiratory disease and their comorbidities.

Submitted by: 
Rupert Payne
Funding acknowledgement: 
SJK and YZ are supported by SJK’s MRC Career Development Award (MR/P021573/1)