Third heart sound in hospitalised patients with acute heart failure: insights from the ATTEND study.

BACKGROUND: Several previous studies have suggested that detection of a third heart sound (S3) in patients with chronic congestive heart failure is associated with adverse long-term outcomes. However, the short-term prognostic value of identifying an S3 on admission in patients with acute heart failure (AHF) is not well established. We therefore analysed the in-hospital prognostic value of detecting an S3 on admission in hospitalised patients with AHF. METHODS: The Acute Decompensated Heart Failure Syndromes (ATTEND) study investigators enrolled 4107 patients hospitalised with AHF. Investigators evaluated the presence or absence of an S3 during routine physical examination. RESULTS: On admission to hospital, 1673 patients (41%) had an S3. Patients with an S3 had a higher heart rate, higher serum level of B-type natriuretic peptide and higher creatinine levels than patients without an S3. However, there were no significant differences of systolic blood pressure, serum sodium, haemoglobin, C-reactive protein and total bilirubin between the two groups. Multivariate analysis adjusted for various markers of disease severity revealed that only the presence of an S3 was independently associated with an increase of in-hospital all cause death [adjusted odds ratio (OR), 1.69; 95% confidence interval (CI), 1.19-2.41; p = 0.003] and cardiac death (adjusted OR, 1.66; 95% CI, 1.08-2.54; p = 0.020) among the congestive physical findings related to heart failure (S3, rales, jugular venous distension and peripheral oedema). CONCLUSIONS: Detecting an S3 on admission was independently associated with adverse in-hospital outcomes in patients with AHF. Our findings suggest that careful bedside assessment is clinically meaningful.

Relationship between sleep apnoea and mortality in patients with ischaemic heart failure.

OBJECTIVE: To determine whether the influence of sleep apnoea (SA) on the risk of death differs in patients with ischaemic and in those with non-ischaemic heart failure (HF). DESIGN: Prospective observational study. PATIENTS: Consecutive patients with HF with left ventricular ejection fraction < or =45% newly referred to the HF clinic between 1 September 1997 and 1 December 2004. MAIN OUTCOME MEASURES: Patients underwent sleep studies and were divided into those with moderate to severe SA (apnoea-hypopnoea index > or =15/h of sleep) and those with mild to no SA (apnoea-hypopnoea index <15/h of sleep). They were followed up for a mean of 32 months to determine all-cause mortality rate. RESULTS: Of 193 patients, 34 (18%) died. In the ischaemic group, mortality risk adjusted for confounding factors was significantly higher in those with SA than in those without it (18.9 vs 4.6 deaths/100 patient-years, hazards ratio (HR) = 3.03, 95% CI 1.04 to 8.84, p = 0.043). In contrast, in the non-ischaemic HF group, there was no difference in adjusted mortality risk between those with, and those without, SA (3.9 vs 4.0 deaths/100 patient-years, p = 0.929). CONCLUSIONS: In patients with HF, the presence of SA is independently associated with an increased risk of death in those with ischaemic, but not in those with non-ischaemic, aetiology. These findings suggest that patients with ischaemic cardiomyopathy are more susceptible to the adverse haemodynamic, autonomic and inflammatory consequences of SA than are those with non-ischaemic cardiomyopathy.