Identifying Structural Factors Associated with Central Line-Associated Bloodstream Infections (CLABSI) Risk in a Single-Center Pediatric Academic Hospital

BACKGROUND: Efforts to prevent CLABSIs have primarily focused on individual compliance with established care bundles, however little is known to what extent organizational structure and process influence CLABSI outcomes. METHODS: To expose associations within clinical units and identify CLABSI risk, aggregate monthly data from July 2014 to June 2016 was collected on admitted patients aged 0 months to 18 years. Nine clinically disparate units were categorized into four clinical paradigms to analyze factor effect: Intensive Care Nursery (ICN), Pediatric Intensive Care Unit (PICU), Hematology/Oncology (Heme/Onc) and Medical Surgical (Med/Surg). This retrospective analysis evaluates three structure-related factors: acuity-based nurse workload/rate of hours per patient day (HPPD), number of full and part-time staff (FTPT), and number of separations/nurse turnover. Four process-related factors include average length of stay and rates of central line entry stratified by type: laboratory collection, medications, and flush. Multivariable Poisson regression was used to produce incidence rate ratios (IRR) and account for central line days. Factors by unit type were standardized to represent one standard deviation change. RESULTS: A total of 104 CLABSIs were identified with rates ranging from 0 to 9.5 CLABSI per 1,000 central line days. Nurse turnover was a significant risk in ICN (IRR: 1.41; P = .018) and Med/Surg (IRR: 1.36; P = .046) yet non-significant for the PICU (IRR: 0.90; P = .341) and Heme/Onc (IRR: 1.01; P = .871) floors. FTPT was associated with increased CLABSI in PICU (IRR: 1.49; P = 0.005) yet protective for the ICN (IRR: 0.34; P < .001). Length of stay was a risk for Heme/Onc (IRR: 1.43; P = 0.088), yet protective for the ICN (IRR: 0.65; P = .002) and PICU (IRR: 0.62; P = .016). Central line access reasons were not statistically significant with the exception of lab entries on the Med/Surg floor (IRR:1.67; P = 0.030). CONCLUSION: CLABSI mitigation requires more than individual competence with task-related practices. The unique work ecology of each clinical area may broadly influence CLABSI as an outcome. Extending analyses to consider organizational structure can inform resource allocation and recalibrate traditional prevention strategies. DISCLOSURES: All authors: No reported disclosures.