Conception and bicentric validation of the proSCANNED score, a simplified bedside prognostic score for Heart Failure patients

Figures

• 

  Figure 1. Integer risk score and probability of dying within 1 and 4.4 years in the derivation sample. The histogram shows the number of patients for each score. The points and error bars show the mean predicted probabilities and their Wald’s 95% Confidence Interval. In bold, the predicted probabilities of dying within 1 year; in light, the predicted probabilities of dying within 4.4 years.

• 

  Figure 2. Predicted mortality over 4.4 years for theoretical patients in the scores 0 to 4. Cumulative mortality risk over 4.4 years for theoretical patients with scores of 0 to 4. Scores ≥ 5 were not represented because they need larger scale of the vertical axis, limiting analysis of lower scores.

• 

  Figure 3. ROC curves of the integer score for the prediction of 1 and 4.4-years mortality in the derivation sample. (A) 1 year mortality. (B) 4.4 years mortality. The grey shape displays the 95% confidence interval of sensitivity. ROC receiver operating characteristics. AUC area under the curve. CI confidence interval. CI were computed using a boot-strap method.

• 

  Figure 4. ROC curves of the integer score for the prediction of 1 and 4.4-years mortality in the validation sample. The grey shape displays the 95% confidence interval of sensitivity. ROC receiver operating characteristics. AUC area under the curve. CI confidence interval. CI were computed using a boot-strap method.

• 

  Figure 5. Calibration of the score in the validation sample, for the mortality risk at 1 year.