Using electronic health records to understand multimorbidity in older people: a scoping review

The increasing availability of electronic health records (EHR), encompassing routinely collected general practice, hospital, and linked national census data, presents significant opportunities to enhance our understanding of multimorbidity at scale. However, the utility and clinical impact of EHR in advancing multimorbidity research, particularly concerning older adults who are disproportionately affected, remain unclear. This study aims to synthesise the literature on the use of EHR to investigate multimorbidity in individuals aged 65 years and older.

We conducted a scoping review, performing literature searches across Medline, Embase, CINAHL, and the Cochrane Central Register of Controlled Trials. Studies published from the inception of these databases until September 2024 were included if they investigated multimorbidity using EHR in individuals aged 65 years and older. The extracted data were synthesised using an iterative descriptive process.

A total of 753 studies were identified, with 46 meeting the eligibility criteria for inclusion. All studies originated from high-income countries, primarily the USA and Spain. Research focused on areas such as healthcare utilisation and costs, predictive modelling based on patient disease profiles, specific clustering models, relationships between multimorbidity and particular disease categories, aetiology, polypharmacy, and the accuracy of EHR. Reporting on data completeness, sociodemographic characteristics, and the inclusion of ethnic and social diversity within EHR datasets was notably limited.

The use of EHR to elucidate multimorbidity is expanding and holds substantial promise. However, enhanced reporting of data completeness and sociodemographic characteristics is essential to facilitate the effective translation of findings into real-world populations. Neglecting these aspects, risks exacerbating inequality gaps for individuals 65 years and older with multimorbidity.