News & Activities
Methodology Monday July Roundup
In this month's #MethodologyMonday, the recent article by Abbidi and Sinha proposes a conceptual framework for using artificial intelligence (AI) and machine learning (ML) to improve equity, diversity, and inclusion throughout the clinical trial lifecycle.
In this #MethodologyMonday focus, the authors argue that the persistent underrepresentation of certain patient groups limits the generalisability of trial findings and can exacerbate health disparities. They highlight how AI/ML methods could support more representative participant recruitment, retention, and analysis strategies. The paper concludes that, while AI-driven approaches offer substantial potential to reduce biases and enhance inclusivity in clinical research, their implementation must be accompanied by careful attention to algorithmic fairness, transparency, ethical oversight, and equitable access to trial participation. The framework proposed uses predictive modelling, adaptive designs, and continuous bias monitoring, providing a structured pathway for researchers to enhance both the scientific validity and ethical integrity of clinical trials
A recent study on participation by sex and race in amyotrophic lateral sclerosis (ALS) clinical trials in the US showed that non-White participants are significantly under-enrolled and that there is a trend to underrepresentation of women. This study highlights the continued disparities that exist in the clinical trial population when compared to the disease prevalence population. This paper will serve as a basis to build solutions for the development of more effective enrolment across all groups in people with ALS.
Next we look at the use of a Delphi study to address equity, diversity, and inclusion improvement in multiple sclerosis studies. Consensus was reached on 45 priorities spanning diversity measurement and reporting, inclusive recruitment practices, the roles of funders and publishers, and researcher training, with most priorities considered achievable rather than merely aspirational. The findings provide a stakeholder-driven framework to improve the inclusiveness, representativeness, and relevance of future MS research in Canada.
The following review, focused on language-based eligibility in UK trials, was led by Issacs with contributions from the SENSITISE consortium members. It exposes methodological practices that may impede diverse participation.
Half of the 32 included RCTs explicitly reported a language-based eligibility criterion, including 63% of trials evaluating talking therapies for depression. These trials had a median proportion of white participants of 97%. Systematic underrepresentation of ethnically diverse patients undermines external validity and is likely to perpetuate health inequities.
