Alterations in US Health Data Replacement of "Gender" with "Sex" Ignite Concerns over Public Confidence
In a recent publication in The Lancet, researchers have raised concerns about the alteration of over 100 health datasets from US government agencies, with a shift from collecting data on "gender" to focusing solely on "sex." This change, it is argued, moves away from meaningful transparency and could have significant implications for the accuracy and completeness of health data, trust in science, and international research collaboration.
Context: Under the current administration, there has been a significant rollback of federal efforts to collect data on sexual orientation and gender identity (SOGI). This includes rescinding guidance on adding SOGI questions to Medicaid and CHIP applications and removing gender identity variables from major public health data systems, such as the CDC’s Behavioral Risk Factor Surveillance System and the National Violent Death Reporting System. These actions align with an administration focus on limiting recognition and support for LGBTQ+ health issues.
Potential Impact: The elimination of gender identity data undermines the ability to identify and address specific health disparities among LGBTQ+ populations, such as elevated suicide risk, mental health challenges, and barriers to access. These hidden changes also erode public trust by obscuring transparent and inclusive data collection practices, especially as they appear politically motivated rather than scientifically justified.
Moreover, US public health data often contributes to global health research. Alterations that exclude gender identity variables could hinder comprehensive international studies on health disparities, skew comparative data analyses, and weaken efforts to address global LGBTQ+ health inequities.
The authors of the analysis, led by Janet Freilich, a professor at the Boston University School of Law, and Aaron S Kesselheim, a professor at Harvard Medical School, suggest that these changes can degrade the quality of the underlying information and can undermine the interpretation of the results of studies, potentially making healthcare recommendations ineffective or counterproductive.
The datasets were modified between January 20 and March 25, 2025, and weren't routinely updated monthly or more frequently. The change to "gender" to "sex" was observed in 106 of the 114 datasets (93 percent) analysed. Only 15 of these changed datasets had indications that alterations had been made.
The authors argue that transparency can ensure that datasets are still trusted and useful. They also suggest that these changes could affect trust amongst the international research community, with researchers from abroad less willing to put their own data in repositories hosted by the US Government.
The authors reference Secretary Robert F Kennedy Jr's calls for 'radical transparency', but argue that unlogged data manipulation is not consistent with this. The solution proposed is for databases to use terms that accurately describe the data collected and for any changes to be clearly logged.
- The modification of over 100 health datasets from US government agencies, altering "gender" to "sex," could have significant repercussions for public health research, particularly in identifying and addressing specific health disparities among LGBTQ+ populations.
- The shift in focus from collecting data on sexual orientation and gender identity (SOGI) to primarily collecting data on "sex" may undermine the quality of health data, potentially making healthcare recommendations ineffective or counterproductive.
- The authors of a recent publication in The Lancet argue that these changes could erode trust in science, both domestically and internationally, as researchers from abroad may become less willing to collaborate with or put their own data in repositories hosted by the US Government.
- To maintain trust in science and ensure accuracy and completeness of health data, the authors suggest that databases should use terms that accurately describe the data collected and any changes to these databases should be clearly logged.