Cohort profile for the creation of the SAIL MELD-B e-cohort (SMC) and SAIL MELD-B children and Young adult e-cohort (SMYC)

Abstract

AbstractPurposeWe have established the SAIL MELD-B electronic cohort (e-cohort SMC) and the SAIL MELD-B children and Young adults e-cohort (SMYC) as a part of the Multidisciplinary Ecosystem to study Lifecourse Determinants and Prevention of Early-onset Burdensome Multimorbidity (MELD-B) project. Each cohort has been created to investigate and develop a deeper understanding of the lived experience of the ‘burdensomeness’ of multimorbidity by identifying new clusters of burdensomeness indicators, exploring early life risk factors of multimorbidity and modelling hypothetical prevention scenarios.ParticipantsThe SMC and SMYC are longitudinal e-cohorts created from routinely-collected individual-level population-scale anonymised data sources available within the Secure Anonymised Information Linkage (SAIL) Databank. They include individuals with available records from linked health and demographic data sources in SAIL at any time between 1stJanuary 2000 and 31stDecember 2022. The SMYC e-cohort is a subset of the SMC, including only individuals born on or after the cohort start date.Findings to dateThe SMC and SMYC cohorts include 5,180,602 (50.3% female and 49.7% male) and 896,155 (48.7% female and 51.3% male) individuals respectively. Considering both primary and secondary care health data, the five most common long-term conditions for individuals in SMC are ‘Depression’, affecting 21.6% of the cohort, ‘Anxiety’ (21.1%), ‘Asthma’ (17.5%), ‘Hypertension’ (16.2%) and ‘Atopic Eczema’ (14.1%), and the five most common conditions for individuals in SMYC are ‘Atopic Eczema’ (21.2%), ‘Asthma’ (11.6%), ‘Anxiety’ (6.0%), ‘Deafness’ (4.6%) and ‘Depression’ (4.3%).Future plansThe SMC and SMYC e-cohorts have been developed using a reproducible, maintainable concept curation pipeline, which allows for the cohorts to be updated dynamically over time and manages for the request and processing of further approved long-term conditions and burdensomeness indicators extraction. Best practices from the MELD-B project can be utilised across other projects, accessing similar data with population-scale data sources and trusted research environments.STRENGTHS AND LIMITATIONS OF THIS STUDY-SMC and SMYC are representative of the Welsh population.-Anonymised cohorts serve as an effective strategy for overcoming consent-related barriers, enabling seamless data aggregation and analysis.-The creation of a reproducible concept curation pipeline to manage and process data extraction for the e-cohorts enables efficient delivery of datasets in support of multiple research questions and outcomes.-Routine data does not capture data on important aspects such as quality of life.-Routine data can be subject to missing data or errors.-Lack of coverage of burdensomeness indicators in routine data.