Female health researchers face systemic bias that makes them less likely to win government research grants than their male counterparts, a Canadian study shows.

When judged on scientific merit alone, female researchers were just 0.9 percentage points behind male researchers in grant application success rates, an analysis of almost 24,000 applications to the national health research funding body found.

But when applications were judged on the calibre of the lead researcher rather than the proposed science, females trailed males by 4 percentage points, according to a study published in the Lancet.

Study author Dr Holly Witteman said the analysis, which covered research funding applications research grant programmes between 2011 and 2016, was possible because of a natural experiment at a federal health research funding agency. In 2014, the Canadian Institutes of Health Research (CIHR) introduced two streams of research funding, one in which application reviews focused on the project scientific content and the other based on a review of the principal investigator.

With project-based reviews, the grant application success rates were 12.9% for males and 12.1% for female applicants. But in the Foundation program based on investigator reviews, the success rates were 8.8% for females vs 12.7% for males.

“Our study provides empirical evidence that gender gaps in grant funding stem from women being evaluated less favourably as principal investigators … not on the quality of their proposed research,” writes Dr Witteman and colleagues.

Women might be being rated less favourably for research grants because of conscious or unconscious bias on the part of reviewers, they postulated.

There could also be systemic bias against women in the form of application review criteria that reflect cumulative disadvantage for women, they added. This would create a vicious circle as female researchers received less grant funding, had articles accepted less readily by journals, were less frequently invited to speak at conferences or selected for awards, and not be perceived as leaders, Dr Witteman and colleagues said.

“These discrepancies have made it easier for the average male researcher to build his CV compared with his average female peer,” they said.

“We recommend funders focus assessment on the science rather than the scientist, measure and report funding by relevant applicant characteristics, and develop policies to mitigate effects of all forms of bias,” the researchers concluded.