Big Pharma Tests How AI Can Lower Bias in Clinical Trials

Pharmaceutical giant Johnson & Johnson (J&J) has started using artificial intelligence (AI) to make clinical trials more diverse. It hopes its technology could make treatments for marginalized communities more effective.

J&J’s AI creates a heatmap depicting which clinics black people frequent in order to recruit some patients for clinical trials. Previously, disadvantaged communities were prevented from taking part in trials because of cost and distance, preventing a more well-rounded development of medication.

J&J Plans to Increase AI Trials

The new technology has helped increase the participation of black cancer patients by 5.2%, a trend J&J hopes will continue. It plans to initially increase diversity in 50 and then 100 trials during 2024.

Trials that don’t test effectiveness across gender, race, and age can widen health disparities and reduce the effectiveness of medicines. Big Pharma often enlists established academic medical centers whose populations may not be diverse to ensure equitable treatments.

Around three-quarters of participants in approved drug trials in 2020 were white. Eleven percent were Hispanic, and 8% Black.

AI Success in Clinical Trials

Smaller drug companies, called “biotech” firms, use AI in other ways. For example, Palo Alto-based Inceptive is testing whether AI can develop messenger-RNA (mRNA) vaccines similar to the COVID-19 shot.

AI In clinical trials | Source: HIT Consultant

COVID vaccine maker Moderna recently affirmed the role of AI in creating its coronavirus vaccine. Recursion Pharmaceuticals has started human trials for five new drugs, one of which treats a neurovascular disease.

However, AI has not, since Moderna’s vaccine, developed a drug whose effectiveness has been proven in the real world. Scientific American pointed out in 2022 that the effectiveness of AI models depends on the data-handling practices of researchers.

Read more: Top 7 Machine Learning Applications in 2023

It recommends that clinically-used AI models be developed with the input of patients and the US Federal Drug Administration. Several companies, including Recursion rival Exscientia, have also pared back their expectations of what AI can do.

Exscientia CEO Professor Andrew Hopkins said the company would “focus on high-value” internal programs” that could lead to “significant therapeutic benefit.” Todd Rudo, the chief medical officer at Clario, the company that develops endpoint technologies for clinical trials, agreed that AI was little more than a tool.

“Ultimately, we need to remember that AI isn’t a magic bullet…AI is a methodology, a tool. And like any tool, deep scientific expertise will always be necessary to find the right tool for the right job.”

Read more: Will AI Replace Humans?

Do you have something to say about J&J using AI in clinical trials or anything else? Please write to us or join the discussion on our Telegram channel. You can also catch us on TikTok, Facebook, or X (Twitter).