Last year Roche, a Swiss pharmaceutical firm, published a review of the clinical trials on neurological drugs it had held between 2016 and 2021. It found that black people were under-represented in all but one. Surprisingly, that news represents progress, because it shows that trial organisers are becoming more aware of a dangerous bias that sets back the safety and efficacy of medical treatments.
Many trials exclude certain groups, and do so deliberately—children, for example, or people with physical or learning disabilities, pregnant women and the elderly. For such groups, participation has stalled or even reversed. There are good explanations for the exclusion, such as the difficulty of getting informed consent or the potential harm to unborn children.
Yet the consequences can be absurd. A recent review found that half of trials around the world testing hip-fracture interventions excluded people who lived in nursing homes, were old or had some level of cognitive impairment. Though these groups make up almost a third of all patients suffering hip fractures, it is unclear if the interventions will work as safely or as effectively on them. Their doctors face an invidious choice: prescribe anyway, with uncertain results; or deny their patients new treatments.
A shocking example of such exclusion is of people with Down’s syndrome. They have long been left out of clinical trials, including recent trials of lecanemab and donanemab, the first drugs against Alzheimer’s that seem to slow the progress of the disease. This is despite the fact that those with Down’s are highly likely to develop it. Yet without data from trials, doctors will not prescribe them the drugs, for fear of unknown side-effects.
Obtaining informed consent for trials is not always easy, especially from people with learning disabilities or dementia. Accounting for different groups’ risks of side-effects can complicate the analysis of the data. And some groups mistrust doctors because of a history of mistreatment, which makes recruiting them harder.
Even so, broadening the range of trials’ participants can be practically useful, because they may lead to new medical insights. Running trials on people who are more likely to develop Alzheimer’s, such as those with Down’s syndrome, might help researchers test whether their drugs work preventively.
Something like that happened with the Dallas Heart Study in the 2000s. As a large piece of epidemiological research, it included an ethnically representative sample of people and found a genetic variant in some African-Americans which was correlated with 40% lower bad cholesterol. That gene is now one of the foremost drug targets in the fight against cardiovascular disease.
Fortunately, the bias of clinical testing may be changing. Almost half of trial participants in America are now women (in the rest of the world it is still only 40%). America and Britain look likely to publish regulations that require trial organisers to explain whom they ought to include and how they plan to recruit them.
The drug and medical-device industries are likely to object. Companies may fret about the speed and cost of broad-based trials. During the covid-19 pandemic, Moderna slowed down its vaccine trial because its recruiters, a private contractor, had not enrolled enough subjects from ethnic minorities. In that time, millions contracted the virus.
The trade-off almost always favours efficacy. A pandemic on such a scale is very rare. As broad-based trials become the norm, they will be easier and faster to set up. Firms are rightly granted valuable monopolies as a reward for financing the research needed to discover successful drugs and bring them to market. A quid pro quo should be that the trials which lead to those monopolies reveal who will benefit and by how much.
© 2025, The Economist Newspaper Limited. All rights reserved. From The Economist, published under licence. The original content can be found on www.economist.com