Removal of race from MESA score does not affect heart disease risk prediction
Click Here to Manage Email Alerts
Key takeaways:
- Removing race/ethnicity from the MESA risk score did not change its effectiveness at heart disease risk prediction.
- This will enable the score to be used more broadly.
CHICAGO — When race was removed from the MESA risk prediction score, its ability to predict risk for heart disease did not change significantly, researchers reported at the American Heart Association Scientific Sessions.
The MESA risk score, developed in 2015, includes race/ethnicity, coronary artery calcium score and the following risk factors: high total cholesterol, low HDL, high BP, family history of heart disease, smoking and type 2 diabetes status.
“In light of conversations around the potential harms of including race/ethnicity, we decided it would be valuable to look deeper into the role race/ethnicity in the MESA risk score,” Quinn White, BA, a doctoral student at the University of Washington in Seattle, told Healio.
White and colleagues developed a version of the risk score without race/ethnicity and compared its effectiveness at heart disease risk prediction with the original version.
The researchers found the Harrell’s C statistic for both scores were virtually identical: 0.8 for the new version without race/ethnicity and 0.797 for the original version.
“When we don’t include race/ethnicity in the model development process, we see that the model performs comparably to the model that does include race/ethnicity. These models were similar both in their ability to distinguish events from nonevents, and in the extent to which the models’ predicted event rates were very close to the observed event rates,” White told Healio. “We also assessed model performance stratified by race/ethnicity, and we saw that the models performed similarly within each group. That is, including race/ethnicity is not changing the predictions of the MESA score in a meaningful way.”
White said in an interview that the new version “is more inclusive in the sense that it no longer requires one to designate a race/ethnicity. This means that we can still make predictions for individuals that do not fall into a single racial/ethnic group included in MESA, or who do not wish to disclose their race/ethnicity. Additionally, the updated MESA risk score includes only biological patient characteristics, eg, systolic blood pressure, rather than a combination of biological characteristics and a social construct (race).”
In a video commentary posted on the AHA website, Sadiya S. Khan, MD, MSc, FAHA, Magerstadt Professor of Cardiovascular Epidemiology and associate professor of cardiology, medical social sciences and preventive medicine (epidemiology) at Northwestern University Feinberg School of Medicine and chair of the writing group for the AHA’s PREVENT equations, said the study “helps to reinforce the fact that with a diverse population, and with predictors that perhaps are on the causal pathway, so individuals who are developing high blood pressure or diabetes that we know contributes to heart disease risk, that may differ across groups, is then being captured in the risk prediction model. And so, I think because of those reasons, the model performs well when you have the right predictors in place, even without the social construct of race built into it.”
References:
- Dr. S. Khan on SS24 Su3054 on MESA model with-without race. https://newsroom.heart.org/file/dr253817?action=. Published Nov. 11, 2024. Accessed Dec. 10, 2024.
- MESA heart disease risk score worked well with or without race included. https://newsroom.heart.org/news/mesa-heart-disease-risk-score-worked-well-with-or-without-race-included. Published Nov. 11, 2024. Accessed Dec. 10, 2024.
For more information:
Quinn White, BA, can be reached at qwhite@uw.edu.
Collapse
White Q, et al. Abstract Su3054. Presented at: American Heart Association Scientific Sessions; Nov. 16-18, 2024; Chicago.
Read more about
Click Here to Manage Email Alerts