Home › Cardiovascular Scores, Indexes, And Algorithms › CATCH-HF Hospitalization Predictor

CATCH-HF Hospitalization Predictor

  • Age (years)
  • NYHA class (1-4)
  • Prior HF hospitalization in last 12 months (0 = no, 1 = yes)
  • NT-proBNP (pg/mL)
  • eGFR (mL/min/1.73m²)
  • Systolic BP (mmHg)
  • Heart rate (bpm)
  • Hemoglobin (g/dL)
  • Model coefficients (intercept + betas)
    Enter as JSON mapping variable name → numeric coefficient. Example (do not modify variable names):
    {"intercept":-3.2,"age":0.02,"nyha":0.6,"prior_hf_adm":1.1,"ntprobnp":0.0008,"egfr":-0.01,"sbp":-0.005,"hr":0.01,"hgb":-0.12}
  • CATCH-HF Hospitalization Predictor: Explanation and Clinical Context
    This page implements a configurable logistic risk model intended to predict the probability of heart failure (HF) hospitalization using readily available clinical variables (age, NYHA class, prior HF admission, NT-proBNP, eGFR, systolic blood pressure, heart rate, hemoglobin). The calculator itself does not assume or embed any particular published coefficients; instead it accepts a JSON set of model coefficients (intercept + variable betas) so the exact derivation or validation coefficients from the original CATCH-HF publication can be entered to reproduce the original model precisely.

    Clinical use and interpretation: A validated HF-hospitalization risk model should be used as an adjunct to clinical judgement. Predicted probabilities depend entirely on the derived coefficients and on appropriate variable measurement (for example, NT-proBNP assay units and timing). Before clinical use, ensure the model coefficients are from a peer-reviewed derivation and external validation, confirm the population and endpoint match your intended setting, and consider local recalibration if necessary.

    References:
    1) On commonly used HF risk scores and their variables: Get With The Guidelines–Heart Failure (GWTG-HF) risk score (example resource: MDCalc and original papers describe commonly used clinical predictors for short-term mortality and outcomes).
    2) Hospitalization/readmission risk in heart failure and performance of prediction models — contemporary reviews and comparative studies discussing model performance and limitations.
    3) Search notes: At the time of creating this tool, I was unable to locate a peer-reviewed derivation/validation publication explicitly named “CATCH-HF” describing a hospitalization predictor with published coefficients; therefore this template is built to accept the original coefficients when available. (If you have the publication or DOI, paste the intercept and betas into the coefficients box above and the page will compute exact predicted probabilities using that model.)