Illustration(s) pertain to the topic addressed in this publication, not the specific research or data presented in the publication

Increased bloodstream infection rates in surgical patients associated with variation from recommended use and care following implementation of a needleless device

Abstract

OBJECTIVE: To determine if an apparent increase in bloodstream infections (BSIs) in patients with central venous catheters (CVCs) was associated with the implementation of a needleless access device.

DESIGN: Retrospective cohort study using a derived CVC-days factor for estimating appropriate denominator data.

SETTING: A 350-bed urban, acute, tertiary-care hospital.

METHODS: BSI surveillance data were obtained, and high-risk areas for BSIs were determined. A random 5% sample of medical records was used to estimate CVC days, and a cohort study was conducted to compare BSI rates before and during needleless device use. A survey was conducted of nursing needleless-device practices.

RESULTS: The surgical intensive-care unit (SICU), the medical intensive-care unit, and the solid organ transplant unit (OTU) were identified as high-risk units. Using existing surveillance BSI data and the estimated CVC days, the catheter-related BSI rates in the high-risk surgical patients were significantly higher during the needleless-device period compared with the preneedleless-device period (SICU, 9.4 vs 5.0/1,000 CVC days; OTU, 13.6 vs 2.2/1,000 CVC days). A survey of the nurses revealed that 60% to 70% were maintaining the needleless devices correctly.

CONCLUSION: We observed a significant increase in the BSI rate in two surgical units, SICU and OTU, associated with introduction of a needleless device. This increase occurred shortly after the needleless device was implemented and was associated with nurses’ unfamiliarity with the device, and needless-device use and care practices different from the manufacturer’s recommendations.

Cookson ST, Ihrig M, O’Mara EM, Denny M, Volk H, Banerjee SN, Hartstein AI, Jarvis WR

Infect Control Hosp Epidemiol 1998 Jan;19(1):23-7

PMID: 9475345