Methods: CLABSIs were defined according to CDC/NHSN definitions. CLANCs were defined using a novel classification as non-infectious events resulting in premature removal of the line. 592 oncology patient records (< 24 years; 2006-16) were reviewed. Wilcoxon rank sum tests were used for continuous and ordinal characteristics and Chi-square or Fisher’s exact tests for categorical characteristics.
Results: 656 CVCs were inserted in 368 patients, for a total of 175,941catheter days (9.6% inpatient). Events included: 108 CLABSIs (42 inpatient and 66 outpatient) and 89 CLANCs (44 inpatient and 45 outpatient). The all-harm event rate was 1.1 per 1000 CVC days; the sum of CLABSI (0.61) and CLANC (0.50) rates. Inpatient rates were: all-harm (4.9), CLABSIs (2.4), and CLANCs (2.5). Outpatient event rates were: all-harm (0.72), CLABSIs (0.45), and CLANCs (0.27). For all lines treated independently, risk ratio of an adverse event was strongly correlated with CVC type (tunneled CVCs vs ports; 11.8; <0.001), age at placement per 1 year older (0.89; <0.001), gender (females vs males; 1.6; 0.021), and tumor type (AML vs Non-AML Leukemia/Lymphoma; 4.0; <0.001). Tunneled CVCs carried greater risk for both CLABSI (10.8; <0.001) and CLANC (13.2; <0.001) than ports.
Conclusion: We have developed an all-harm index to quantify the total harm associated with central line use. Among pediatric oncology patients with CVCs, major non-infectious complications occur at rates similar to those reported for CLABSIs. Although event rates per 1000 CVC days were lower among outpatients, the total number of infectious and non-infectious harm events was similar in the inpatient and outpatient settings. Additional quality improvement efforts are required to reduce the total harm associated with CVC use, and modifiable factors such as catheter choice could significantly impact the rate of both CLABSIs and CLANCs.
S. Worley, None
A. Tang, None
V. Arakoni, None
C. B. Foster, None