To ascertain the effect of an infection control program including process control on intensive care unit (ICU) rates of intravascular device (IVD)–associated bloodstream infection (BSI).
Two level III adult ICUs in one public university hospital in Mexico: one medical surgical ICU and one neurosurgical ICU.
All adult patients admitted to study units who had a central venous catheter (CVC) in place for at least 24 hrs.
A prospective before/after trial in which rates of IVD-associated BSI are determined during a period of active surveillance without process control (phase 1) were compared with rates of IVD-associated BSI after implementing an infection control program applying process control (phase 2).
Six hundred five IVD-days were accumulated in phase 1, and 2824 IVD-days were accumulated during phase 2. Compliance with CVC site care and hand hygiene improved significantly from baseline during the study period: placing a gauze dressing over the catheter insertion site (99.24% vs. 86.69%, respectively; relative risk [RR] = 1.14; 95% confidence interval [CI] = 1.07–1.22; p = .0000), proper use of gauze for vascular catheter insertion site (97.87% vs. 84.21%, respectively; RR = 1.16; 95% CI = 1.09–1.24; p = .0000), documentation of the duration of the administration set of the vascular catheter (93.85% vs. 40.69%, respectively; RR = 2.34; 95% CI = 2.14–2.56; p = .0000), and hand hygiene before contact with the patient (84.9% vs. 62%, respectively; RR = 1.37; 95% CI = 1.21–1.51; p = .0000). Overall rates of IVD-associated BSI were lowered significantly from baseline rates after implementation of process control (19.5 vs. 46.3 BSIs per 1000 IVD-days, respectively; RR = 0.42; 95% CI = 0.27–0.66; p = .0001). Overall rates of crude unadjusted mortality were lowered significantly from baseline rates (48.5% vs. 32.8% per 100 discharges, respectively; RR = 0.68; 95% CI = 0.50–0.31; p = .01).
Implementation of an infection control program utilizing education, process control, and performance feedback was associated with significant reductions in rates of IVD-associated BSI and mortality.
From General Hospital (FH, PD, JR, GF), Mexico City, Mexico; Medical College of Buenos Aires (VDR), Buenos Aires, Argentina; and Section of Infectious Diseases, Department of Medicine, University of Wisconsin Medical School (NS), Madison, WI.
Supported by a grant from Baxter Health Care International (to FH, PD, JR, GF).
Address requests for reprints to: Victor D. Rosenthal, MD, Arengreen 1366, Buenos Aires 1405, Argentina. E-mail: firstname.lastname@example.org