A316 ANTIMICROBIAL LOCKS FOR THE PREVENTION OF CATHETER-RELATED BLOOD STREAM INFECTIONS (CRBSI) IN PATIENTS ON PARENTERAL NUTRITION (PN)– A SYSTEMATIC REVIEW

Abstract

Background: CRBSI are serious complications for patients on PN. Antimicrobial lock solutions have been used as a prophylaxis against CRBSI, however their efficacy in the PN population is unclear. Aim(s): To conduct a systematic review on the efficacy of antimicrobial lock solutions to prevent CRBSI in patients on PN. Method(s): We performed a systematic search of EMBASE, MEDLINE, CENTRAL, ISI Web of Knowledge (from earliest date to December 2015) for randomized controlled trials (RCT) and observational studies evaluating the efficacy of antimicrobial locks compared to control solutions (heparin or saline) to prevent CRBSI in adult and pediatric PN patients. Two independent reviewers performed study selection and data extraction. Critical appraisal of possible bias was performed using the Cochrane Risk of bias tool (RCT) and ROBINS-I (observational studies). Result(s): The literature search identified 771 citations, 112 were reviewed in full, and 19 studies were selected, totaling 536 patients, including 9 pediatric (122 patients) and 10 adult (414 patients) studies. Nine studies assessed an ethanol lock solution, 9 taurolidine, and 1 gentamycin. Two open-labeled RCT were included, both using taurolidine (Klek and Bisseling). The remaining studies were observational, and employed a pre-and post-design at moderate to high risk of bias. Methodological and clinical heterogeneity precluded pooling of the data by meta-analysis. All studies, except one by Klek who used taurolidine in low-risk patients, report lower CRBSI rate in the treatment group (ethanol, taurolidine +/- citrate, and gentamycin lock) compared to controls (heparin or saline). The negative trial by Klek et al. documented 1 CRBSI (0.273/1000 catheter-days) in the taurolidine plus citrate arm, and none in the taurolidine or control arms (p=1.000). This finding is in contrast to Bisseling et al. who reported lower CRBSI rates (0.19 versus 2.02 CRBSI/1000 catheter-days, p=0.008), and longer mean infection-free survival in the taurolidine arm (641 (95% CI 556-727) vs 175 (95% CI 85-266) days, log-rank p<0.0001 in the control group (heparin)). Amongst the observational studies, catheter infection rates ranged from 0 to 14.3 CRBSI/1000 catheter-days for antimicrobial locks vs 3.53 to 26.5 CRBSI/1000 catheter-days in the controls. The calculated CRBSI rate ratios of intervention vs control groups are all <1 (range 0-0.54), suggesting a protective effect attributable to antimicrobial lock solutions as prevention of CRBSI, most pronounced in taurolidine studies. Conclusion(s): Pooled results would appear to support the use of antimicrobial lock solution as prophylaxis for CRBSI in PN patients, yet are limited by low study quality and modest sample sizes. High quality studies are needed to confirm these findings, and define optimal parameters for their use.