Risk scores for predicting incident heart failure admission in patients with chronic coronary syndromes: Validation...

Area of Science:

• Cardiology
• Clinical Prediction Models
• Public Health

Background:

• Heart failure (HF) admission significantly impacts prognosis in chronic coronary syndrome (CCS) patients.
• Existing stratification schemes for predicting HF admission lack external validation.

Purpose of the Study:

• To develop point scores for predicting incident HF admission using data from prior studies.
• To externally validate these scores in an independent prospective cohort.
• To compare the discriminative ability of different prediction scores for HF admission.

Main Methods:

• Independent predictive variables for HF admission in CCS patients without baseline HF were identified from four prospective studies (CARE, PEACE, CORONOR, CLARIFY).
• Scores were generated based on the relative magnitude of Cox regression coefficients.
• Scores were validated and compared in a monocentric prospective cohort.

Main Results:

• The validation cohort comprised 1212 patients, with 171 experiencing HF admission over up to 17 years.
• All developed scores demonstrated statistically significant discriminative ability for predicting HF admission.
• The CORONOR score showed superior predictive performance compared to the CLARIFY score (C-statistic 0.73 vs. 0.69, p=0.03).

Conclusions:

• All evaluated scores possess significant discriminative ability for predicting incident HF admissions in an independent cohort.
• The predictive performance of the scores was largely comparable.
• Significant differences in discriminative ability were observed only between the highest and lowest performing scores.