The harder you look, the more you find: Catheter-associated bloodstream infection surveillance variability

Background: Catheter-related bloodstream infections are an important quality performance measure and remain a significant source of added morbidity, mortality, and medical costs. Objective: Our objectives were to assess variability in catheter-associated bloodstream infections (CA-BSI) surveillance practices, management, and attitudes/beliefs in pediatric intensive care units (PICUs) and to determine whether any correlation exists between surveillance variation and CA-BSI rates. Methods: We used a survey of 5 health care professions at multiple institutions. Results: One hundred forty-six respondents from 5 professions in 16 PICUs completed surveys with a response rate of 40%. All 10 (100%) infection control departments reported inclusion or exclusion of central line types inconsistent with the Centers for Disease Control and Prevention CA-BSI definition, 5 (50%) calculated line-days inconsistently, and only 5 (50%) used a strict, written policy for classifying BSIs. Infection control departments report substantial variation in methods, timing, and resources used to screen and adjudicate BSI cases. Greater than 80% of centers report having a formal, written policy about obtaining blood cultures, although less than 80% of these address obtaining samples from patients with central venous lines, and any such policies are reportedly followed less than half of the time. Substantial variation exists in blood culturing practices, such as temperature thresholds, preemptive antipyretics, and blood sampling (volumes, number, sites, frequencies). A surveillance aggressiveness score was devised to quantify practices likely to increase identification of bloodstream infections, and there was a significant correlation between the surveillance aggressiveness score and CA-BSI rates (r = 0.60, P = .034). In assessing attitudes and beliefs, there was much greater confidence in the validity of CA-BSI as an internal/historical benchmark than as an external/peer benchmark, and the factor most commonly believed to contribute to CA-BSI occurrences was patient risk factors, not central line maintenance or insertion practices. Conclusion: There is substantial variation in reported CA-BSI surveillance practices among PICUs, and more aggressive surveillance correlates to higher CA-BSI rates, which has important implications in pay-for-performance and benchmarking applications. There is a compelling opportunity to improve standardized CA-BSI surveillance to enhance the validity of this metric for interinstitutional comparisons. Health care professionals' attitudes and beliefs about CA-BSI being driven by patient risk factors would benefit from recalibration that emphasized more important drivers-such as the quality of central line insertion and maintenance.