The Cardiovascular Literature-Based Risk Algorithm (CALIBRA): Predicting Cardiovascular Events in Patients With...

Area of Science:

• Nephrology
• Cardiology
• Biostatistics

Background:

• Cardiovascular disease (CVD) is the leading cause of mortality in chronic kidney disease (CKD) patients.
• Current risk prediction models are often inadequate for this population.
• Accurate CVD risk assessment is crucial for effective clinical management of CKD.

Purpose of the Study:

• To develop and validate a novel, literature-based risk prediction model for cardiovascular (CV) hospitalizations in CKD patients.
• To assess the performance of the new model, named CArdiovascular, LIterature-Based, Risk Algorithm (CALIBRA), against existing CV risk scores.

Main Methods:

• A literature-based, naïve-bayes model (CALIBRA) was developed incorporating 31 traditional and CKD-specific risk factors.
• CALIBRA was validated in two independent CKD cohorts: the FMC NephroCare (EuCliD®) and the German Chronic Kidney Disease (GCKD) study.
• Model performance was evaluated using c-statistics and calibration, with comparisons to Framingham Heart Study (FHS), ASCVD, and INDANA risk scores.

Main Results:

• CALIBRA demonstrated good discrimination in both validation cohorts (AUC 0.79 in EuCliD®, 0.73 in GCKD).
• CALIBRA showed significantly improved accuracy over FHS, ASCVD, and INDANA in both cohorts (ΔAUC ranging from -0.04 to -0.22).
• The model maintained accuracy even with missing data, indicating robustness in real-world settings.

Conclusions:

• CALIBRA offers accurate and robust cardiovascular risk stratification for CKD patients.
• The model provides superior predictive accuracy compared to established CV risk scores.
• CALIBRA's generalizability across diverse CKD populations and clinical settings is supported by the findings.