Predictors and outcomes of infection-related hospital admissions of heart failure patients

Abstract

Background: Infections are one of the most common causes for hospitalization of patients with heart failure (HF). Yet, little is known regarding the prevalence and predictors of different types of acute infections as well as their impact on outcome among this growing population.

Methods and results: We identified all patients aged 50 or older with a major diagnosis of HF and at least one echocardiography examination who had been hospitalized over a 10-year period (January 2000 and December 2009). Infection-associated admissions were identified according to discharge diagnoses. Among 9,335 HF patients, 3530 (38%) were hospitalized at least once due to infections. The most frequent diagnoses were respiratory infection (52.6%) and sepsis/bacteremia (23.6%) followed by urinary (15.7%) and skin and soft tissue infections (7.8%). Hospitalizations due to infections compared to other indications were associated with increased 30-day mortality (13% vs. 8%, p<0.0001). These higher mortality rates were predominately related to respiratory infections (OR 1.28 [95% CI 1.09, 1.5]) and sepsis\bacteremia (OR 3.13 [95% CI 2.6, 3.7]). Important predictors for these serious infections included female gender, chronic obstructive pulmonary disease, past myocardial infarction and echocardiography-defined significant right (RV) but not left ventricular dysfunction.

Conclusions: Major infection-related hospitalizations are frequent among patients with HF and are associated with increased mortality rates. Elderly female patients with multiple comorbidities and those with severe RV dysfunction are at higher risk for these infections.

Figure 1. 30-day mortality rates per infection type.

Thirty-day mortality was compared between each of the HF patient cohorts admitted with distinct infections and HF patients admitted for non-infectious causes. HF patients admitted with respiratory infections, bacteremia or sepsis had significantly higher mortality rates, while HF patients admitted with skin and urinary tract infections had similar and lower 30-day mortality rates, respectively.