LAUSR

OBJECTIVE To assess the rate of bacterial contamination of fluid and ports in intravenous bags in a veterinary emergency room (ER) and intensive care unit (ICU). DESIGN Experimental model. SETTING Ninety intravenous fluid bags of lactated balanced-electrolytes solution (1 L) hung in a university hospital. INTERVENTIONS Bags were hung in 2 different locations in the ER (sink and bins) and one location in the ICU (sink) for 11 days. Bags were punctured 3 times daily with a sterile needle to simulate clinical use. Injection ports were swabbed and 50 mL of fluid were collected in duplicates on days 0, 2, 4, 7, and 10. Aerobic bacterial cultures were performed on the fluid and injection port. Contamination was defined as bacterial growth of a similar phenotype across 2 consecutive times. Increase in the fluid contamination rate from day 0 was tested using an exact binomial test. Port contamination rate between locations was tested using Fisher's exact test. MAIN RESULTS Combined bacterial growth on injection ports reached a mean (95% confidence interval) of 8.1 (0.005-16.2) cfu/port on day 10. The combined port contamination was 3.3%, 11.1%, 17.8%, and 31.1% on days 0, 2, 4, and 7, respectively. Port contamination was similar between ER and ICU. However, port contamination was higher in the sink versus the bins area (38.3% vs 16.7%, P = 0.032). No fluid bag was contaminated at days 0 and 2. The contamination rate of fluid bag was 1.1% and 4.4% on days 4 and 7, respectively. All bags with contaminated fluid were in the ER (6.7%, 95% exact binomial confidence interval 1.9-16.2%). CONCLUSIONS Injection port contamination reached 31.1% on day 7. Contamination was more likely when the bags were hung next to a sink. In our model of bag puncture, fluid contamination occurred between days 2 and 4.