Initial management of and outcome in patients with pneumococcal bacteremia: a retrospective study at a Swiss university hospital, 2003-2009

Abstract

PURPOSE:

The aim of this quality control study was to assess the time to initial diagnostic procedures and the time to the first dose of antibiotics in patients with pneumococcal bacteremia, and to investigate whether the timeliness of these interventions influenced outcome.
METHODS:

We retrospectively studied patient characteristics, chronological sequence of diagnostic and therapeutic steps, and the course of disease of all patients with pneumococcal bacteremia at a Swiss university hospital between 2003 and 2009, and we analyzed associations between these factors and the length of hospital stay (LOS) and mortality.
RESULTS:

A total of 102 episodes of pneumococcal bacteremia in 98 patients were analyzed, of whom 15.7% died during hospitalization. The median time (interquartile range [IQR]) to the first antibiotic dose was 4.0 (2.0-5.9) h, and the median times (IQR]) to blood cultures, chest radiograph, lumbar puncture, and brain computed tomography (CT) scan or magnetic resonance imaging (MRI) were 1.4 (0.5-3.3), 2.5 (1.2-4.2), 4.2 (2.7-7.2), and 2.3 (0.6-6.2) h, respectively. The time to diagnostic procedures and therapy were not associated with LOS or death. Risk factors for death in the univariable analysis were: Charlson comorbidity index [odds ratio [OR] (95% confidence interval) per unit increase, 1.3 (1.1-1.6)], neutropenia [OR 10.1 (2.0-51.0)], human immunodeficiency virus (HIV) infection [OR 3.9 (1.1-13.8)], chronic respiratory disease [OR 4.4 (1.2-16.0)], chronic liver disease [OR 3.2 (1.0-9.7)], smoking [OR 3.8 (1.1-13.5)], injection drug use [OR 9.7 (1.5-63.7)], and antibiotic therapy within 6 months before admission [OR 4.0 (1.3-12.5)]. The multivariable analysis revealed age >60 years (P = 0.048) and alcoholism (P = 0.009) as risks for prolonged LOS.
CONCLUSIONS:

The outcome of pneumococcal bacteremia may be more influenced by patient characteristics than by minor differences in the timeliness of initial diagnostic and therapeutic measures within the first several hours after hospital admission.