Evaluation of the source of infection in patients with severe sepsis

Introduction

The growing frequency of patients with severe infection in the ICU, resulting in persistent high mortality associated with high costs, is a concern that calls for attention in critical care medicine. It is important to amplify knowledge about severe sepsis and septic shock, in an attempt to prevent it, to identify it early and to reduce mortality. The objective of this study is to evaluate the source of infection and the evolution of patients with severe infection in the ICU.

Methods

All patients admitted to the ICU of a public university hospital in the period January–June 2004 were included. The variables collected were demographic data, admission diagnostic, SOFA and APACHE II scores, definition of sepsis and sepsis-related conditions were in accordance with the ACCP/SCCM definitions, and the source and site of infection were recorded for each of first sepsis event. The length of stay and mortality were also recorded. For statistical analysis, the program Epi Info version 3.3.2 was used.

Results

During the study period 316 patients were admitted to the ICU, the male sex was more frequent (65.8%), and the mean age was 56.5 ± 20.4 years. At admission 141 patients (44.6%) had a diagnostic of severe infection, 86 (28.5%) being severe sepsis and 55 (18.2%) septic shock. The most frequent admission diagnoses of these patients were sepsis, gastrointestinal surgery and intracranial hemorrhage. When comparing the group of patients with severe infection with the other patients we found a higher APACHE II score (25.09 ± 8.7 and 17.93 ± 6.7, respectively; P < 0.0001), and a higher SOFA score (9.4 ± 4.3 and 5.5 ± 3.3, respectively; P < 0.0001). The sites of infection more frequently observed were pulmonary (63.8%), abdominal (11.3%) and urinary (7.8%). The source of infection was in the community in 46.1% of the cases of severe infection and nosocomial in 53.2% (P = 0.23). The mortality stratified by the source of infection did not differ among patients (60% community and 62.6% in the nosocomial infection group, P = 0.52).

Conclusion

Severe infection was a common cause of admission to the ICU in this study. The patients with severe infection had a higher severity of disease and more organ failure when compared with the other patients admitted to the ICU. The frequency of community and nosocomial infection was similar in the group of patients with severe infection, as was the associated mortality.