The extent of gender/sex variables included in clinical algorithms across medical specialties

Gender/sex variables are often included in clinical algorithms, typically without articulating which aspects of gender or sex-linked biologies are hypothesized to contribute to the tools’ output. To date, there exists little systematic evaluation of how gender/sex inputs are used in clinical algorithms. We analyzed all clinical algorithms available in the online database MDCalc, assessing whether and how they include gender/sex inputs. For tools with gender/sex inputs, we evaluated the consequences of, justification for, and timeline of including such inputs. Of 602 tools reviewed, 112 (18.6 percent) included gender/sex. These inputs were used to adjust the algorithmic outcome by directly allocating points according to a change in the gender/sex variable or to normalize the outcome by using gender/sex to interpret other tool inputs. Fewer than half of tools included text justifying the inclusion of gender/sex in initial tool development, and only five tools had been updated since their initial development. Systematic reconsideration and regular reevaluation of gender/sex inputs in clinical algorithms may be needed to determine which inputs, if any, are warranted to support clinicians in providing better care for patients.