Purpose: This study compared the Glasgow Prognostic Score (GPS) with brain natriuretic peptide (BNP) level as prognostic factors in patients with acute decompensated heart failure (ADHF).
Subjects and Methods: We investigated 362 patients with ADHF (75±13 years, 55% male) in this retrospective study. Serum C-reactive protein (CRP), albumin, and BNP concentrations were measured on admission. Patients with both an elevated CRP concentration (>1.0 mg/dL) and hypoalbuminemia (<3.5 g/dL) were assigned a GPS score of 2, those with only one of these abnormalities were assigned a score of 1, and patients with neither were assigned a score of 0. GPS scores of 0, 1, and 2 were observed in 47% (n=170), 30% (n=110), and 23% (n=82) of patients, respectively. The patients were also classified into three groups according to BNP tertiles (T1: 28–462 pg/mL, T2: 474–1,130 pg/mL, T3: 1,131–9,830 pg/mL).
Results: During the median follow-up period of 349 days (interquartile range: 72–825), 67 (18.5%) patients died. Relative to a GPS of 0, the hazard ratios for all-cause death were 4.94 (95% confidence interval 2.26–11.0) for a GPS of 2 and 2.31 (95% confidence interval 1.18–4.61) for a GPS of 1, as determined using adjusted Cox proportional-hazards analysis. In contrast, there were no significant differences among BNP groups.
Conclusions: The GPS is more useful than the BNP level for determination of prognosis in patients with ADHF.
View full abstract