Central line‐associated bloodstream infections (CLABSIs) cause significant morbidity, mortality, and cost to patients and the health care system. Because duration of use is a major risk factor for infection, the rate of CLABSI could be reduced if central venous catheters (CVCs) were removed when no longer necessary. This study's primary aims included: (1) identifying the prevalence of unnecessary CVCs in ward patients and (2) reducing unnecessary CVC days utilizing a quality improvement (Ql) intervention that improves situational awareness.


We conducted a before‐and‐after study with concurrent controls. A random sample of patients with a CVC was generated. The intervention included physician and staff education, a decision support tool (DST), and organizational change protocols for automatic peripheral intravenous (PIV) catheter placement and prompt removal of idle CVCs. CVC utilization ratio (UR; catheter‐days/patient‐days), CLABSI rates, CVC dwell times, idle CVC days per patient, and CVC reinsertion rates were measured on intervention and control wards. The Student t test and, when necessary, the exact procedure were applied to compare each group's average estimates.


A total of 531 and 312 CVC days were analyzed in the pre‐ and postintervention groups, respectively. Preintervention, 33.9% of ward CVC days were unnecessary. Twenty‐six of the 34 preintervention compared with 18 of the 30 postintervention patients studied had at least 1 idle day. Nursing and physician compliance was 91% and 82%, respectively, during the trial, and 17 CVCs were removed in the first 11 weeks specifically because of the intervention. The average CVC utilization ratio on the intervention floor decreased from 0.32 to 0.25 (P < 0.01) and was unchanged on the control floor (pre, 0.26; post, 0.28, P = 0.57). Although the study was not powered for CLABSI reduction, the average CLABSI rate decreased from 3.29 CLABSI/1000 catheter days to 0.00 on the intervention ward and increased on the control floor. There were no CLABSIs on the intervention floor in the 5 months of the study. On the intervention ward, the mean number of total CVC days per patient was reduced from 14.8 to 10.0 days (P = 0.19), and the mean number of idle CVC days per patient was reduced from 5.1 to 3.6 days (P = 0.36). CVC reinsertion events were not increased after the intervention.


This study demonstrates that a composite Ql intervention can reduce CVC utilization and may reduce CVC dwell times and CLABSIs. Further study will better delineate the magnitude of these outcomes and the transferability of the strategy.

Statistical Process Control Chart demonstrating there were no CLABSIs in the 5‐month study period.

Statistical Process Control Chart demonstrating decrease in central venous catheter utilization after implementation of the intervention.

Author Disclosure:

S. Chemetsky Tejedor, Emory Healthcare, employment; J. Stein, Emory Healthcare, employment; C. Payne, Emory University, employment; S. Ray, Emory University, employment; M. Ido, Emory Healthcare, employment; D. Dressier, Emory Healthcare, employment; K. Rask, Emory University, employment; J. Stein, Ingenious Med, Inc., Patent Owner; SHM, Sanofi, Speakers Bureau Grant; Sanofi, honoraria.