Impact of infection on admission and of the process of care on mortality of patients admitted to the Intensive Care Unit

Abstract

A prospective, cohort, clinical, observational study was performed in 14 Intensive Care Units (ICUs) to evaluate the contemporary epidemiology, morbi-mortality and determinants of outcome of the population with an infection on admission. All 3766 patients admitted during a consecutive 12-month period were screened. Their median age was 63 [26-83], 61.1% were male and 69.8% had significant comorbidities. On admission to the ICU 1652 patients (43.9%) had an infection, which was community acquired in 68.2% (one-fifth with healthcare-associated criteria) and ward-acquired in the others. Roughly half presented to the ICU with septic shock. As much as 488 patients with community-acquired infections were deemed stable enough to be first admitted to the ward, but had similar mortality to unstable patients directly admitted to the ICU (35.9% vs. 35.1%, p 0.78). Only 48.3% of this infected population had microbiological documentation and almost one-quarter received inappropriate initial antibiotic therapy. This, along with comorbidities, was a main determinant of mortality. Overall, infected patients on admission had higher mortality both in the ICU (28.0% vs. 19.9%, p <0.001) and in the hospital (38.2% vs. 27.5%, p <0.001) and even after being discharged to the ward (14.2% vs. 9.6%, p <0.001). Also, patients not infected on admission who acquired an infection in the ICU, had an increased risk of dying in the hospital (odds ratio 1.41 [1.12-1.83]). Consequently, infection, regardless of its place of acquisition, was associated with increased mortality. Improving the process of care, especially first-line antibiotic appropriateness, and preventing ICU-acquired infections, may lead to better outcomes.