Importance of Bnp changes during the follow-up in elderly outpatients with heart failure

The aim of the present study was to analyze whether absolute BNP and relative BNP change values during follow-up of elderly patients with chronic HF could predict the development of acute decompensated episodes.

A total population of 108 elderly outpatients with the diagnosis of HF was retrospectively reviewed. Available BNP levels after at least one stable visit and one acute decompensated episode were required. Follow-up period was 12 months (343 visits, 42.6% decompensated HF episodes). Relative BNP changes were analyzed: "CC period" (patients who remained stable) and "CD period" (patient who suffered from a decompensated HF episode). Changes in BNP levels were significantly higher in CD than in CC periods (mean and median differences 138% and 85%, vs 16% and 0%, respectively; p<0.001). The clinical model (AUC=0.83) had a sensitivity of 67.06%, and a specificity of 80.36%. Relative BNP change (29%) showed by itself, a similar AUC (0.83) and specificity (79%) and an improved sensitivity (0.80) than the clinical model. When relative BNP change was introduced at the clinical model, a similar specificity was obtained and the diagnostic accuracy, AUC (0.89 vs 0.83, p=0.01) and sensitivity were improved. Absolute BNP changes showed worse AUC than that derived from relative BNP changes or clinical assessment.

Percent BNP change values during the follow-up showed better results than absolute BNP values and improved the clinical assessment for diagnostic of decompensated HF episodes in elderly outpatients.