Clinical Scorecard: Influence of Racial and Ethnic Factors on the Efficacy of Advanced Breast Cancer Risk Prediction Models
At a Glance
Category
Detail
Condition
Advanced breast cancer risk prediction
Key Mechanisms
Inclusion or exclusion of race and ethnicity in risk models affects calibration and discrimination of advanced breast cancer risk predictions
Target Population
Women aged 40–74 years undergoing breast cancer screening
Care Setting
Breast cancer screening programs and clinical decision-making for screening intervals and supplemental imaging
Key Highlights
Excluding race and ethnicity from the Breast Cancer Surveillance Consortium (BCSC) advanced breast cancer risk model leads to overestimation of risk in Asian women and underestimation in Black women.
The model including race and ethnicity demonstrates better calibration and slightly higher discrimination overall compared to the race-naive model.
Removing race and ethnicity may result in suboptimal identification of high-risk individuals, potentially impacting tailored screening strategies and breast cancer mortality reduction.
Guideline-Based Recommendations
Diagnosis
Use the BCSC advanced breast cancer risk model including race and ethnicity for risk assessment to guide screening decisions.
Management
Apply risk stratification to determine screening intervals: biennial screening for low/average risk, annual screening for intermediate risk, and consideration of supplemental imaging for high risk.
Incorporate race and ethnicity in risk models to improve accuracy and equity in screening recommendations.
Monitoring & Follow-up
Monitor model calibration and discrimination across racial and ethnic groups to ensure equitable risk prediction.
Evaluate screening outcomes to assess effectiveness of risk-based screening intervals and supplemental imaging.
Risks
Excluding race and ethnicity may lead to misclassification of risk, particularly underestimating risk in Black women and overestimating risk in Asian women.
Misclassification can result in inappropriate screening intervals and missed opportunities for early detection.
Patient & Prescribing Data
Women aged 40–74 years undergoing annual or biennial breast cancer screening
Risk prediction models including race and ethnicity better identify women at intermediate or high risk who may benefit from tailored screening intervals and supplemental imaging, potentially reducing advanced breast cancer diagnoses.
Clinical Best Practices
Incorporate race and ethnicity variables in advanced breast cancer risk prediction models to enhance calibration and discrimination.
Use validated risk models like the BCSC advanced breast cancer risk model to guide personalized screening strategies.
Consider federal guidelines recommending biennial screening mammography for women aged 40–74 and inform women with dense breasts about supplemental imaging options.
Continuously evaluate model performance across diverse populations to minimize algorithmic bias and improve health equity.
