Inclusion of under-served groups in trials: an audit at a UK primary care clinical trials unit.

backgroundClinical trials need to include patients who are representative of the population who may receive the tested interventions in the future. The importance of inclusivity is recognised by ethical and funding bodies and has public support. Appropriate inclusion is required to provide equitable evidence-based healthcare and to comply with ethical principles for research. However, there is little information about the inclusivity of most under-served groups in UK clinical trials.

methodsThis audit assesses the inclusion of under-served groups in trials run by the Oxford Primary Care Clinical Trials Unit (PC-CTU). We included trials with ethical approval between 2017 and 2023. We checked protocols, patient-facing information and selected data collection tools for information on the under-served groups in the INCLUDE guidance and protected characteristics in the UK Equality Act 2010, to identify explicit exclusions and data collection.

resultsWe included 19 trials. They were in a variety of clinical conditions, testing different types of interventions, both Clinical Trial of an Investigational Medicinal Product (CTIMP) and non-CTIMP. Most were non-commercially funded. We reviewed 21 protocols, 29 Patient Information Sheets/Leaflets and 40 data collection tools. Common exclusions were based on age (19), sex or gender (11), language (8), capacity to consent (14), pregnancy (11), multiple health conditions (10) and severity of illness (17). Trials most often collected data on age (19), sex or gender (15), ethnicity (16), education (11), address (13), mental health conditions (6), who gave consent (19), addiction (6), multiple health conditions (10), severity of illness (17), smoking status (12) and obesity (13).

conclusionsOften, exclusions were due to the focusing of the trial for a specific group, such as older people, women, or people being treated for a specific severity of condition. However, many explicit exclusions may not have been essential, may have reduced the inclusivity of the trials and might limit the applicability of the trial's findings to people to whom the tested interventions might be relevant. These include the exclusion of people aged under 18, people without English language fluency and people without capacity to consent. All trials could have collected more informative data on under-served group status.