Bacteriologic contamination of intravenous infusion delivery systems in an intensive care unit.

Seventy intensive care unit patients were admitted to a double-blind prospective study to determine the level of contamination associated with the admixture and administration of intravenous solutions and whether intravenous filtersets prevented bacteremia. Patients were randomly assigned a 0.22 micron filterset (real filter) or a filter cartridge without a 0.22 micron membrane (blank filter) on all possible intravenous lines. Forty-six (14.1 percent) real filtersets and 38 (11.3 percent) blank filtersets were found to be contaminated, and overall 30 patients (42.4 percent) were found to have extrinsically contaminated intravenous administration systems at least once during the study. Bacterial adherence to the plastic cartridge was demonstrated to be responsible for culture-positive blank filtersets. Staphylococcus epidermidis was the organism most frequently isolated from real and blank filtersets. Epidemiologic surveillance identified 10 patients with blank filtersets and three patients with real filtersets with clinically significant hospital-acquired bacteremias during the study period. It is concluded that a significant level of extrinsic contamination of intravenous infusion delivery systems occurred on the intensive care unit; documented clinically significant nosocomial bacteremias occurred less often in those patients who had a 0.22 micron bacterial retention filter on all possible intravenous lines.