Abstract
BACKGROUND: In heart failure (HF) patients, both natriuretic peptides (NP) and previous HF hospitalization (pHFH) have been used to predict prognosis. HYPOTHESIS: In a large real-world population, both NP levels and pHFH have independent and interdependent predictive value for clinical outcomes of HFH and all-cause mortality. METHODS: Linked electronic health records and insurance claims data from Decision Resource Group were used to identify HF patients that had a BNP or NT-proBNP result between January 2012 and December 2016. NT-proBNP was converted into BNP equivalents by dividing by 4. Index event was defined as most recent NP on or after 1 January 2012. Patients with incomplete records or age < 18 years were excluded. During one-year follow up, HFH and mortality rates stratified by index BNP levels and pHFH are reported. RESULTS: Of 64 355 patients (74 ± 12 years old, 49% female) with available values, median BNP was 259 [IQR 101-642] pg/ml. The risk of both HFH and mortality was higher with increasing BNP levels. At each level of BNP, mortality was only slightly higher in patients with pHFH vs those without pHFH (RR 1.2 [95%CI 1.2,1.3], P < .001); however, at each BNP, HFH was markedly increased in patients with pHFH vs those without pHFH (RR 2.0 [95%CI 1.9,2.1], P < .001). CONCLUSION: In this large real-world heart failure population, higher BNP levels were associated with increased risk for both HFH and mortality. At any given level of BNP, pHFH added greater prognostic value for prediction of future HFH than for mortality.