Clinical coding in multiple long-term conditions research: challenges and best practices

Access to health data, of good enough volume and quality, is central to our understanding of multiple long-term conditions (MLTCs) and how to treat them. There is a rise in the use of health data for cohort selection using clinical codes, with the potential to advance areas of MLTC treatment and polypharmacy research. We have reviewed literature with published methodologies to discuss current understanding and potential challenges around coding of clinical data, selection of data sources (e.g. electronic health records) and validity of the decision making process. We include a suggested framework  to curate drug lists with clinician input and provide an extensive list of resources to assist MLTC researchers to access or create code lists, prioritising re-usability and transparency.  Sharing best practice guidelines and agreeing on standards can help this research field move towards having a unified approach to structuring and coding clinical information, with the ultimate goal being more accurate and effective clinical decision making.