Of 299 heart failure patients followed for up to 285 days, 96 deaths were recorded (mortality rate: 32.1%). Overall survival rate was 67.9% and median survival was Not reached. The single strongest independent mortality predictor after multivariate adjustment was High Blood Pressure (HR=1.61).

The Kaplan-Meier curve shows the probability of surviving at each follow-up day across 299 patients. Survival drops below 75% around day 100. Overall survival at the end of follow-up was 57.6%. The shaded band shows the 95% pointwise confidence interval.

Patients are grouped into Low (<30%), Borderline (30-49%), and Normal >=50% ejection fraction categories. Stratified Kaplan-Meier curves show that lower ejection fraction is associated with worse survival. The log-rank test comparing all three groups gives p < 0.001, confirming statistically significant differences. Shaded bands are 95% confidence intervals per group.

Kaplan-Meier curves compare survival between anaemic and non-anaemic patients. The log-rank test gives p = 0.099. Anaemia does not appear to be a statistically significant prognostic factor in this cohort. Shaded bands show 95% confidence intervals.

The Cox proportional hazards model was fitted with all available clinical covariates simultaneously. Hazard ratios > 1 indicate increased mortality risk; HR < 1 indicates a protective association. The strongest independent predictor is High Blood Pressure with HR = 1.61, meaning each unit increase roughly multiplies the death hazard by that factor.

Each point represents one patient, coloured by outcome. Patients who died had median serum creatinine 1.3 vs. 1 in survivors. Median ejection fraction was 30% in non-survivors vs. 38% in survivors. Clustering of non-survivors at high creatinine combined with low ejection fraction visually confirms these two features as the dominant risk separators.

Mean clinical values are compared between patients who died and those who survived, with two-sample t-test p-values. Variables with statistically significant differences (p < 0.05): Age, Ejection Fraction, Serum Creatinine, Serum Sodium. Lower p-values indicate that a variable is more strongly differentiated by outcome.