(Margaret McCartney is a Radio 4 inside health resident Dr. )
BMJ Editorial: Sex and gender should not be conflated in medical data
BMJ 2025; 389 doi: https://doi.org/10.1136/bmj.r797 (Published 25 April 2025)
Cite this as: BMJ 2025;389:r797
Margaret McCartney, senior lecturer1,
Susan Bewley, emeritus professor (honorary) in obstetrics and women’s health2
The target of any sex question should be sex
Accurate data are essential to clinical care, research, and health service planning. But in some data sets, the NHS and the wider medical profession have conflated the key demographic variables of sex and gender.12 This can decrease the integrity and reliability of data3 and potentially compromise healthcare.4567A government commissioned independent review led by Alice Sullivan, professor of sociology at University College London, was published in March 2025. It concluded that, for good practice, the “default target of any sex question should be sex”; and questions about gender identity should be asked separately.1
Sex (male or female) is determined by gametes at conception with phenotype influenced by chromosomes, hormones, and reproductive organs.8A few people are found to have one of a small number of clinically recognised variations in sex development (VSDs) after expert investigation.9These individuals nonetheless have a sex, either male or female.81011 Gender, by contrast, is a mutable social construct, defined as “the norms, behaviours, and roles associated with being a woman, man, girl, or boy … and varies from society to society and can change over time.”12
The Sullivan review set out to examine how data and statistics are gathered in respect to the recording of sex and gender. It found that sex, gender, and gender identity have become conflated in research and clinical datasets in recent years. This may have occurred because gender has been used as a synonym for sex, with some clinicians, researchers, or administrators misunderstanding the distinctions. But some individuals and organisations have lobbied for gender identity to be collected instead of, not in addition to, sex with deliberate merging of categories.131415161718
Risks of conflation
The effect of this conflation is a decrease in data integrity and reliability for both information gathering (eg, for surveys or studies) and individual patient care in the NHS. The 2021 Census of England and Wales, for example, asked: “Is the gender you identify with the same as your sex registered at birth?” The intention was to capture data about people who, for example, identify as transgender or non-binary. But the survey did not ask a clear and specific question about sex and generated unreliable data about the gender identity in the surveyed population.319
In the NHS, sex and gender have been recorded interchangeably in medical records for over 10 years.20 Recorded sex is used for interpreting test results and screening invitations, but a person can change their NHS record on request and alter the original medical record marker of sex to that of their gender identity.21 The Sullivan review strongly recommends against this.1 Multiple symptoms, differential diagnoses, laboratory reference ranges, imaging, and clinical risk calculators are sex specific, including haemoglobin, renal function, and cardiovascular risk calculations.4 Using the opposite sex to make clinical decisions is a patient safety issue; it produces erroneous test results and risk calculations, causes incorrect issuing of screening invitations, and increases the risk of wrong diagnoses.56
Respectful approach
GPs have in some cases been instructed to alter records to remove information relating to a patient’s sex.22 Such interference with records assumes sex is irrelevant in a medical history, when it is not.4 It should be possible to note gender (among other identity markers) without changing the sex marker. This would maintain fidelity and accuracy, reduce risks, and also afford choice. Trans people might have understandable concerns about privacy but can be reassured that all staff and regulated professionals are bound by strict duties of confidentiality. The Royal College of General Practitioners already recommends recording both sex and gender.23
Conflating sex and gender also has implications for policy making. Programmes intended to tackle health inequalities in LGBTQ+ people have led advocates to inappropriately advise clinicians on prescribing hormones.24 A Royal College of Obstetricians and Gynaecologists guideline that assumes trans women become gynaecological patients remains stalled since consultation in 2022.25Medical literature and patient information materials that miss important information potentially lead to confusion (eg, replacing the word “women” with “people” may skew interpretation of medical research26) and dehumanisation (for example, by referring to women as a function of their anatomy or physiology 26), both of which undermine inclusivity.1927
Disentangling sex from gender identity as recommended by the Sullivan report and using open, respectful discussions with patients would enable good medical practice approaches to solve these problems and put patient safety first. Both sex and gender can be recorded, but sex should always be included.
Footnotes
Competing interests: We have read and understood BMJ policy on declaration of interests and declare that MM is a director of Beira’s Place, Edinburgh, which provides support for women affected by sexual violence. Full declarations of interests can be found at Sunshine UK (https://www.whopaysthisdoctor.org/).