Evaluating the Use of Sex in Gastroenterology Algorithms: A Fairness Assessment in Representing Sex Framework Analysis

Clinical algorithms are integral to decision-making in gastroenterology, yet the inclusion of sex as a variable remains underexamined. We applied the Fairness Assessment in Representing Sex (FAIRS) framework to evaluate the medical, ethical, and equity implications of sex-based variables in adult gastroenterology algorithms.

Of 184 algorithms reviewed, 11 met inclusion criteria by explicitly incorporating sex or gender. The FAIRS framework was applied to these algorithms.

Using FAIRS, 10 algorithms demonstrated appropriate and clinically justified inclusion of sex, typically reflecting biologically meaningful differences that improved prognostic accuracy without exacerbating disparities. Notably, Model for End-Stage Liver Disease 3.0 exemplified how sex inclusion can mitigate inequity in liver transplant allocation. By contrast, the Obesity Surgery Mortality Risk Score failed FAIRS criteria because evidence suggests sex-based risk differences are confounded by comorbidities and access to care.

Our findings highlight that sex inclusion in algorithms must be explicitly justified, continuously reevaluated, and contextualized to avoid perpetuating bias.