Clinical algorithms may improve if race variables are replaced with causal factors associated with race, such as healthcare access, education and economic position, according to a study published Jan. 1 in The Lancet Digital Health.
The use of race variables in clinical algorithms has been scrutinized in recent years, as race is a social construct, not biological. Researchers from Boston University examined the use of race in a risk calculator for atherosclerotic cardiovascular disease and found that the predicted risk differences for Black and white patients "seems biologically implausible."
The risk calculator generated "substantially divergent" cardiovascular disease risk estimates for Black versus white patients with identical risk profiles, which could lead to race-related disparities in treatment recommendations, according to the authors.
The authors said clinical algorithms should be improved to prevent differential treatment among Black and white patients with identical risk factor profiles. They said replacing race with social determinants of health could rectify the risk differences.