Prospective Analysis of Catheter Associated Blood Stream Infection in ED-Placed Central Venous Catheters

Background:    The placement of central venous catheters (CVCs) is a daily occurrence in many emergency departments (ED). Although some complications of these devices are well studied (e.g., pneumothorax), the risk of infection is sparsely described in the literature. 

Objective: We prospectively determined the duration of placement and the rate of infection in CVCs placed in a tertiary university-based ED.

Methods: 50 consecutive patients receiving a CVC were enrolled.  Each subsequent day, a study team member visited the patient to determine (1) if the line was still in place, and (2) if the patient met CDC criteria for nosocomial line sepsis.  Additional measurements included initial indication for the device, site of admission (ICU vs. general care), admission diagnosis and, at the conclusion of the study, whether removal of the catheter had been unequivocally documented in the medical record.

Results: 39 catheters (78%) were placed for resuscitation and the remainder for routine IV access.  39 (78%) of patients were admitted to the ICU, 7 (14%) were admitted to the floor, and 4 (8%) died in the ED.  The cumulative number of catheter days studied was 165 and the median duration of catheter placement was 48 hr.  Total in-hospital mortality was 14%.  No device became infected per CDC criteria; 0% (CI 0-6%), 0/1000 catheter days (CI 0-20/1000 catheter days).  2 devices were tip-cultured following removal; both were culture positive, but review of each case failed to indicate an indication for this procedure or any other evidence to suggest infection.  Neither location of admission nor indication for placement were associated with line duration by Cox proportional hazards modeling (p=.25).  In 21 of 50 cases (42%), removal of the line was not clearly documented in the chart, raising concerns for the validity of retrospective analysis for study of catheter related complications.

Conclusions: In a prospective consecutive series of 50 catheters followed for 165 days, no case of catheter sepsis related to an ED-placed device was detected, consistent with previous retrospective studies.  As national guidelines for reporting infection rates support normalizing rates per 1,000 catheter days, future studies of ED-related catheter associated blood stream infections may face difficulty accumulating sufficient numbers of devices to evaluate any benefit of targeted infection control strategies.