Date of Approval:
Multimorbidity, the presence of two or more diseases in the same individual, has major implications for public health and health services management. The Clinical Practice Research Datalink is a large database of routinely collected patient data that provides the opportunity to closely examine the burden of multimorbidity in the United Kingdom. In this study we aim to estimate the number and proportion of people in the United Kingdom with chronic diseases and multimorbidity.
This study examines the burden of chronic conditions, comorbidity and multimorbidity in the UK. Incidence and prevalence rates will be calculated for individual chronic conditions that have previously been well-studied in multimorbidity research. Chronic conditions will be defined by a list derived from: the UK Quality and Outcomes Framework 2015, the Charlson Comorbidity Index and the US Department of Health and Human Services Multiple Chronic Conditions framework. Comorbidity will be defined with reference to 6 clinically important index conditions: ischaemic heart disease, stroke, heart failure, diabetes, chronic kidney disease and chronic obstructive pulmonary disease. Multimorbidity will be defined as having two or more chronic conditions. We will also examine the nature of multimorbidity by subgroups defined by Charlson Comorbidity Index score, and combinations of up to 5 of the most prevalent chronic conditions. Prevalence (by calendar year) and incidence will be calculated for each of these defined subgroups of comorbidity and multimorbidity. We will also examine determinants of multimorbidity by stratification into age groups, sex, socioeconomic status and calendar year; and regression models adjusting for hypertension, cholesterol, smoking, body mass index, anti-hypertensive treatment and lipid-lowering therapy. We will also examine temporal trends in multimorbidity combinations through phenomapping, which is unbiased phenotype mapping using unsupervised machine learning.
Health Outcomes to be Measured:
Each variable will defined by a combination of Read codes (CPRD), ICD-10 codes (HES and ONS) and BNF medication codes (CPRD, HES). Where available, phenotypes defined and previously used by The George Institute will be updated for this study. If this is also not available, a code phenotype will be developed using Read codes, ICD-10 codes and BNF medication codes.
Kazem Rahimi - Chief Investigator - The George Institute for Global Health
Jenny Tran - Corresponding Applicant - The George Institute for Global Health
Amit Kiran - Collaborator - University of Oxford