Central line-associated blood stream infection (CLABSI) surveillance is increasingly utilized as an objective measure of quality of care provided by individual hospitals. CLABSI is defined by the National Healthcare Safety Network (NHSN) as a primary bloodstream infection (BSI) in a patient with a central line within the 48-hour period before the development of the BSI (NHSN CLABSI). This traditional definition of CLABSI includes pathogens better described as hospital-acquired blood stream infections (HABSI), such as enteric gram-negative bacilli (GNB) and streptococcus viridans - pathogens inherently more common in patients undergoing hematopoietic cell transplantation (HCT) due to the resultant neutropenia and disruption of mucosal barriers, and unlikely to be line-related. To avoid this misclassification, we have developed a modified CLABSI definition (MCLABSI) which excludes HABSI (DiGiorgio, Infect Control Hosp Epidemiol. 33: 865–8, 2012). MCLABSI includes all of the pathogens under the NHSN definition of CLABSI except Viridans group streptococci species in patients with mucositis, and Enterococcus, Enterobacteriaceae, or Candida species in patients with neutropenia or graft-vs-host disease of the gut. We compared the incidence of CLABSI and its impact on survival using both definitions in acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) patients undergoing SCT.
AML and MDS patients undergoing HCT between August 2009 and December 2011 were identified from the Cleveland Clinic Unified Transplant Database, and NHSN CLABSI and MCLABSI rates were obtained from the infection control database. CLABSI incidence was estimated using Kaplan-Meier method, and risk factors for mortality were identified using stepwise Cox proportional hazards analyses. Of the 73 patients identified (median age 52, range 16–70), 48 were male, 44 had AML, and 29 MDS. Patients received a median of 2 prior chemotherapy regimens (range 0–6), 3 had prior radiation, and 6 had prior transplant. 54 underwent myeloablative and 19 reduced-intensity preparative regimens; stem cell source included bone marrow (BM, n=34), peripheral stem cells (PSC, n=24), and cord blood cells (CBC, n=15). The median CD34+ count was 2.42 × 106/kg and median time to neutrophil recovery (absolute neutrophil count > 500/μL) was 14 days (range 6–24) with BM/PSC and 28 days (range 19–77) with CBC. Most (88%) had mucositis, including 17 (28%) with grade 3 or 4. Twenty-three (31.5%) developed NHSN CLABSI, compared to 8 (11.0%) who developed MCLABSI following HCT, of whom 16 (69.6%) and 7 (87.5%) died, respectively. The majority (16/23) of NHSN CLABSI occurred within 14 days (median 9 days, range 2–211 days) of HCT (Figure ), varying from a median of 5 days (range 2–12 days) for CBC and 78 days for BM/PSC (range 7–211 days, p<.001). Pathogens in NHSN CLABSI included 11 enteric Gram-negative bacilli, 7 Streptococcus viridans group, 6 enterococcus (3 vancomycin resistant), 5 Staphylococcus (3 methicillin resistant), 2 fungal species, 2 Gram-positive bacilli, 1 Pseudomonas, 1 other Streptococcus species, and 1 Stenotrophomonas. MCLABSI occurred a median of 12 days (range 5–176 days) from HCT (Figure ), 7 days for CBC (range 5–12 days) compared to 77 days (range 13–176 days) for BM/PSC (p<.001). Pathogens isolated in MCLABSI included 5 Staphylococcal species (3 MRSE), 2 Streptococcus viridans group, 2 GPB, 1 VRE, and 1 Pseudomonas. 4 NHSN CLABSI and 2 MCLABSI were polymicrobial, and 4 patients had recurrent CLABSI (all of whom died, including 3 MCLABSI). When NHSN CLABSI was analyzed as a time-varying covariate in univariable analysis, it was associated with an increased risk of mortality (HR 3.72, 95% CI 1.88 – 7.36, p<.001), as was MCLABSI (HR 2.96, CI 1.27–6.89, p=.012). CLABSI remained a significant risk factor for mortality in multivariable analysis, by both the NHSN (HR 7.14, CI 3.31 – 15.31, p<.001) and MCLABSI (HR 6.44, CI 2.28–18.18, p<.001) definitions.
CLABSI is a common complication in AML and MDS patients undergoing SCT, and is associated with decreased survival. CLABSI is identified less commonly with the exclusion of HABSI in the modified definition, which more precisely identifies patients with BSI related to their central lines. The distinction between these definitions is important to guide preventative infectious control measures, particularly given CLABSI's role as a quality measure influencing reimbursement.
Hill:Teva Pharmaceuticals: Honoraria, Speakers Bureau.
Asterisk with author names denotes non-ASH members.