P1641What is the best BNP cutoff value to rule out or rule in the diagnosis of heart failure in the community?

Abstract

Abstract Background There is no consensus on the cutoff value of B-type natriuretic peptide (BNP) to rule in or rule out the diagnosis of heart failure (HF) in the community. For instance, the ESC guidelines propose a cutoff of 35 pg/mL and the Canadian Guidelines propose 50 pg/mL. Objectives To evaluate the performance of several BNP cutoffs to rule in or rule out the diagnosis of HF in the community. Methods A total of 633 randomly selected individuals, aged 45 to 99 years, of both sexes, enrolled in a primary care program in several regions of a medium-sized city with 487,562 inhabitants were evaluated. A cross-sectional study, in which one-day clinical data collection, laboratory tests, BNP tests and tissue Doppler echocardiogram (TDE) were performed. The final diagnosis of HF was adjudicated by two independent cardiologists. Sensitivity (SEN), specificity (SPE), negative predictive value (NPV) and positive predictive value (PPV) were evaluated for different BNP cutoffs. A ROC curve was used to determine the best cutoff value. Results The mean age was 59.6±10.4 years and 62% were women. The incidence for ACC/AHA HF stages Zero, A, B, C and D were, respectively, 11.8%, 36.3%, 42.6%, 9.3% and 0%. There was a predominance of HF with preserved versus reduced ejection fraction (59% vs 41%). For the identification of the 59 patients with symptomatic HF, the cutoff of 35pg/mL presented SEN 98%, SPE 87%, NPV 100% and PPV 44%. For cutoff of 50pg/mL these values were SEN 78%, SPE 94%, NPV 98% and PPV 58%. The best combination of SEN and SPE was with a cutoff of 42pg/mL (SEN 92% and SPE 91%). Only one patient with HF had BNP<35pg/mL. With the cutoff of 50pg/mL, 13 (22%) of the 59 pts with symptomatic HF would not have been diagnosed. Conclusions The cutoff with higher specificity to rule in the diagnosis of HF was 50pg/mL. However, with this cutoff an expressive number of patients with HF would have been missed. For screening purpose in the community, the best cutoff to rule out HF was 35pg/mL, as proposed in the ESC guidelines