The role of intestinal colonization with Gram-negative bacteria as a source for intensive care unit-acquired bacteremia*

Abstract

Selective digestive tract decontamination aims to eradicate gram-negative bacteria in both the intestinal tract and respiratory tract and is combined with a 4-day course of intravenous cefotaxime. Selective oropharyngeal decontamination only aims to eradicate respiratory tract colonization. In a recent study, selective digestive tract decontamination and selective oropharyngeal decontamination were associated with lower day-28 mortality, when compared to standard care. Furthermore, selective digestive tract decontamination was associated with a lower incidence of intensive care unit-acquired bacteremia caused by gram-negative bacteria. We quantified the role of intestinal tract carriage with gram-negative bacteria and intensive care unit-acquired gram-negative bacteremia. Data from a cluster-randomized and a single-center observational study. Intensive care unit in The Netherlands. Patients with intensive care unit stay of >48 hrs that received selective digestive tract decontamination (n = 2,667), selective oropharyngeal decontamination (n = 2,166) or standard care (n = 1,945). Selective digestive tract decontamination or selective oropharyngeal decontamination. Incidence densities (episodes/1000 days) of intensive care unit-acquired gram-negative bacteremia were 4.5, 3.0, and 1.4 during standard care, selective oropharyngeal decontamination, and selective digestive tract decontamination, respectively, and the daily risk for developing intensive care unit-acquired gram-negative bacteria bacteremia increased until days 36, 33, and 31 for selective digestive tract decontamination, standard care, and selective oropharyngeal decontamination and was always lowest during selective digestive tract decontamination. Rectal colonization with gram-negative bacteria was present in 26% and 71% of patient days during selective digestive tract decontamination and selective oropharyngeal decontamination, respectively (p < .01). Irrespective of interventions, incidence densities of intensive care unit-acquired gram-negative bacteremia was 4.5 during patient days with both intestinal and respiratory tract gram-negative bacteria carriage. These incidence densities reduced with 33% (to 3.1) during days with intestinal gram-negative bacteria carriage only and with another 45% (to 1.0) during days without gram-negative bacteria carriage at both sites. Respiratory tract decolonization was associated with a 33% and intestinal tract decolonization was associated with a 45% reduction in the occurrence of intensive care unit-acquired gram-negative bacteremia.