Incidence and Risk Factors (RF) for Central Line-Associated Bloodstream Infection (CLABSI) in a Hematology-Oncology (Hem-Onc) Unit

Background:  CLABSI is an important complication in Hem-Onc patients (pts).  National Health Care Safety Network (NHSN) data provides a CLABSI rate of 1.86 per 1000 central venous catheter-days (CVC-d) without stratification for RF (e.g., oncologic diagnosis (dx), receipt of antineoplastic drugs (chemo), neutropenia). This stratification is important since the rate of infection can vary greatly with these RF.

Objective:  To determine the incidence of CLABSI on a Hem-Onc unit and to describe the characteristics of and RF for CLABSI in this type of unit.

Methods:  Between 7/1/09 and 6/30/10, all positive blood cultures from pts with CVC housed on a 40 bed adult Hem-Onc unit of a tertiary care medical center were evaluated for CLABSI using NHSN definitions.  Cases were pts housed on the unit with CLABSI and controls were pts housed on the unit during the same time interval without CLABSI.  Medical charts and electronic records were reviewed for age, sex, length of stay (LOS), dx, receipt and duration of chemo, type of CVC and presence and duration of neutropenia.  Neutropenia-days were defined as number of days from first Absolute Neutrophil Count (ANC) <500mm³ or first total White Blood Cell count (WBC) <1000mm³ to either first ANC > 500 mm³ or first total WBC in normal range (5000-10,000mm³).  The RF were collected from the date of admission to the date of discharge of the Hem-Onc unit.  Univariate analyses were performed.

Results:  There were 37 pts with CLABSI and 716 control pts.  The rate of CLABSI was 3.98 infections per 1000 CVC-d.  Organisms causing CLABSI were:  enterococci (55%), coagulase-negative staphylococci (14%), S. aureus (12%), Candida (12%) and enteric gram-negative bacilli (7%).  5 (14%) CLABSI were polymicrobic. The subcutaneous venous access devices had the highest rate of CLABSI (4.67 per 1000 CVC-d), followed by peripheral inserted central catheters (4.13 per 1000 CVC-d, and triple lumen catheters (3.17 per 1000 CVC-d).  The mean (+/- Standard Deviation (SD)) LOS on the unit before CLABSI was 13 +/- 28 days.  Compared with pts without CLABSI, pts with CLABSI were significantly more likely to be older in age, to have a longer LOS on the unit, to have a dx of leukemia, to have received more days of chemo and had a longer duration of neutropenia.  See table.

Conclusions: On a Hem-Onc unit, CLABSI rates varied by type of CVC. Older age, LOS on the unit, a dx of leukemia, receipt and duration of chemo, and duration of neutropenia were significantly associated with CLABSI.   As many of these variables may be correlated, further analyses are needed. NHSN does not provide RF to stratify CLABSI rates.  The lack of stratification makes benchmarking difficult and can be misleading.  Additional studies are needed to further define RF that can influence CLABSI rates.  Understanding of these RF may assist in developing strategies to prevent CLABSI among Hem-Onc pts.