Background: The incidence of heart failure is frequently reported using hospital discharge diagnoses. The specificity of a diagnosis has been shown to be high but the sensitivity of a reported diagnosis is unknown. Purpose: To study the accuracy of a heart failure diagnosis reported to the Danish National Patient Registers during routine clinical work. Methods: The patient population consisted of 3644 consecutive patients admitted to all departments in one hospital. Diagnoses reported to the National Patient Register were recorded. A study team evaluated each patient independently of routine care, performed an echocardiogram and evaluated whether clinical symptoms of heart failure were present. Heart failure was defined in accordance with current ESC guidelines as symptoms of heart failure and evidence of cardiac dysfunction. Results: A registered diagnosis of heart failure (n = 126) carried a specificity of 99% and a sensitivity of 29% for all patients. The positive predictive value was 81%, the negative predictive value 90%. Conclusion: The diagnosis of Heart Failure in the Danish National Registers is underreported, but very specific.

Objective: To evaluate whether measurements of N-terminal pro-brain natriuretic peptide (NT-proBNP) can be used to differentiate patients with normal and reduced left ventricular ejection fraction (LVEF) in an unselected consecutive group of hospital inpatients. Setting: City general hospital, Copenhagen, Denmark. Patients and design: During a 10 month period 2230 admissions to a city general hospital (80% of targeted patients) had an echocardiographic evaluation of left ventricular function, a comprehensive clinical evaluation, and blood analysis of N-terminal-pro-brain natriuretic peptide (NT-proBNP) within 24 hours of admission. Exclusions resulted from lack of informed consent or failure to obtain the required evaluations before death or discharge from hospital. Echocardiography was unsatisfactory in 37 patients, so the final number studied was 2193. Results: A raised NT-proBNP (> 357 pmol/l) identified patients with an LVEF of < 40% (n = 157) with a sensitivity of 73% and a specificity of 82%. The negative predictive value of having an NT-proBNP concentration below 357 pmol/l was 98%. Concentrations of NT-proBNP increased with increasing age and with decreasing LVEF (p < 0.05). A predicted concentration of NT-proBNP (corrected for age, sex, and serum creatinine) was determined for each patient. In patients with an NT-proBNP value less than predicted, the probability of having an LVEF of > 40% was more than 97%. This probability rapidly decreased to 70% as the measured NT-proBNP increased to 150% of the predicted value. Conclusions: A single measurement of NT-proBNP at the time of hospital admission provides important information about LVEF in unselected patients.

Preserved systolic function among heart failure patients is a common finding, a fact that has only recently been fully appreciated. The aim of the present study was to examine the value of NT-proBNP to predict mortality in relation to established risk factors among consecutively hospitalised heart failure patients and secondly to characterise patients in relation to preserved and reduced systolic function. Material: At the time of admission 2230 consecutively hospitalised patients had their cardiac status evaluated through determinations of NT-proBNP, echocardiography, clinical examination and medical history. Follow-up was performed 1 year later in all patients. Results: 161 patients fulfilled strict diagnostic criteria for heart failure (HF). In this subgroup of patients 1-year mortality was approximately 30% and significantly higher as compared to the remaining non-heart failure population (approx. 16%). Using univariate analysis left ventricular ejection fraction (LVEF), New York Heart Association classification (NYHA) and plasma levels of NT-proBNP all predicted mortality independently. However, regardless of systolic function, age and NYHA class, risk-stratification was provided by measurements of NT-proBNP. Having measured plasma levels of NT-proBNP, LVEF did not provide any additional prognostic information on mortality among heart failure patients (multivariate analysis). Conclusion:The results show that independent of LVEF, measurements of NT-proBNP add additional prognostic information. It is concluded that NT-proBNP is a strong predictor of 1-year mortality in consecutively hospitalised patients with heart failure with preserved as well as reduced systolic function.

AimsThe aim of this study was to assess the epidemiological features and prognosis of heart failure with preserved ejection fraction (HFPEF) and to compare these findings with those from patients with reduced ejection fraction. Furthermore the effects of N-terminal pro brain natriuretic peptide (NT-proBNP) requirement in the heart failure diagnosis were assessed by repeating the analyses in the subgroup of patients with elevated NT-proBNP. Methods and resultsIn 1844 patients admitted, a clinical diagnosis of heart failure was made in 433; amongst these 61% had HFPEF. An elevated NT-proBNP applied to the heart failure diagnosis reduced the number of heart failure patients to 191, and amongst these 29% had preserved ejection fraction. Use of NT-proBNP reduced clinical differences between heart failure patients with preserved and reduced ejection fraction. When not using NT-proBNP, patients with reduced ejection fraction had higher mortality [hazard ratio (HR) 1.24, 95% confidence interval (CI) 1.01-1.52; P ¼ 0.04], even after adjustment for other significant predictors of mortality, except NT-proBNP (HR 1.29, 95% CI 1.04-1.59; P ¼ 0.02). However, no difference in mortality was observed when NT-proBNP was adjusted for (HR 0.90, 95% CI 0.71-1.15; P ¼ 0.4), or used for the heart failure diagnosis (HR 0.96; 95% CI 0.71 -1.29; P ¼ 0.8). ConclusionUsing a heart failure diagnosis requiring elevated NT-proBNP reduces the prevalence of HFPEF and results in a survival similar to that of heart failure with reduced ejection fraction. In contrast, when NT-proBNP is not used for the heart failure diagnosis or adjusted for, HFPEF is associated with a lower mortality in both univariate and multivariate analysis.--