Validation of the Seattle heart failure model [SHFM] in heart failure population

ABSTRACT

To determine the effectiveness of Seattle Heart Failure Model [SHFM] in a Pakistani systolic heart failure cohort in predicting mortality in this population. Cohort study. The Armed Forces Institute of Cardiology - National Institute of Heart Diseases, Rawalpindi, from March 2011 to March 2012. One hundred and eighteen patients with heart failure [HF] from the registry were followed for one year. Their 1-year mortality was calculated using the SHFM software on their enrollment into the registry. After 1-year predicted 1-year mortality was compared with the actual 1-year mortality of these patients. The mean age was 41.6 A +/- 14.9 years [16 - 78 years]. There were 73.7% males and 26.3% females. One hundred and fifteen patients were in NYHA class III or IV. Mean ejection fraction in these patients was 23 A +/- 9.3%. Mean brain natriuretic peptide levels were 1230 A +/- 1214 pg/mL. Sensitivity of the model was 89.3% with 71.1% specificity, 49% positive predictive value and 95.5% negative predictive value. The accuracy of the model was 75.4%. In ROC analysis, AUC for the SHFM was 0.802 [p < 0.001]. SHFM was found to be reliable in predicting one-year mortality among patients with heart failure in the Pakistani patients