The Incidence of Central Venous Catheters and Comorbidities in the Emergency Department Is Increasing

Background: Interventions designed to decrease central line associated blood stream infections (CLABSI) focus on hospital locations with high incidence of central venous catheters (CVCs), namely intensive care units. Elimination of CLABSI will require identifying and interceding in all areas with high incidence of CVC insertions. We hypothesize that new protocols, prolonged emergency department (ED) stays, and increasing visit volumes have increased the indications for CVC insertion in the ED.

Objective: The objective of this study was to quantify the number of ED CVC insertions and assess whether the incidence increased over time.

Methods: Retrospective cohort study using the Healthcare and Utilization Project (HCUP) California State Inpatient Database (SID). This study included all records with ED utilization codes that resulted in hospital admission during the years 2003-2006. The primary outcome, ED CVC insertion was defined as records that included the ICD-9CM code for CVC occurring on day 0 of hospital admission. Elixhauser co-morbidity data were categorized as 0,1,2, or > 3 co-morbidities. Descriptive statistics were used to determine the incidence of CVC insertions per 1000 hospital admissions through the ED. Chi-square was used to test for yearly differences controlling for co-morbidity status.

Results: Over 4 years there were a total of 6,334,546 ED visits resulting in hospital admission and 79,038 CVC insertions on day 0. The mean age of patients undergoing CVC insertion in the ED was 57 years (± 20) and 50.6% were female (95% CI 50.0-51.0). The unadjusted overall CVC rate was 12.5/1000 ED hospital admissions (95% CI 12.4-12.5). The incidence of ED CVC increased annually from 10.2/1000 ED hospital admissions (95% CI 10.0-10.4) in 2003 to 14.6/1000 ED hospital admissions (95% CI 14.4-14.8) in 2006. The incidence increased across all categories of co-morbidity status. However, the proportion of those admitted through the ED with ≥3 co-morbidities increased from 35.0% (95% CI 34.9-35.1) in 2003 to 43.1% (95% CI 43.0-43.2) in 2006 and the largest increase in ED CVC incidence was noted in this group (16.7/1000 ED hospital admissions (95% CI 16.4-17.0) in 2003 to 22.6/1000 ED hospital admissions (95%CI 22.3-23.0) in 2006). The proportion of those who expired upon hospitalization after ED CVC insertion was not significantly different in 2003 (23.4%, 95% CI 22.9-24.3) compared to 2006 (22.6%, 95% CI 22.1-23.2). The median length of stay after ED CVC insertion did not change during this period (6 days, IQR 3-11 days).

Conclusions: The incidence of central venous cannulation in the ED is increasing as patients with more co-morbidities present requiring acute stabilization. Interventions aimed at improving outcomes related to CVC should include the ED.