Abstract: ObjectiveTo explore the effect of different antibacterial drugs on reducing the risk of catheter-related bloodstream infections for different populations of pediatrics with a central venous access device (CVAD). MethodsPediatric patients applicable to CVADs were included from inpatient, outpatient and home settings. With antimicrobial lock and/or systemic intravenous antibiotic therapy in the intervention group, the control group was treated with non-antimicrobial lock and/or systemic intravenous antibiotic therapy. Suspected, possible and confirmed catheter-related bloodstream infections were the primary outcome. Literature was searched in both of English and Chinese databases including PubMed, Ovid-EMBASE and SinoMed from the establishment to September 24, 2020. Search terms were antimicrobial lock, central venous catheter, catheter-related bloodstream infection and children. Cochrane risk-of-bias tool for randomized trials and ROBINS-I were used to evaluate the risk of bias for RCTs and non-RCTs respectively. ResultsA total of 30 studies were included into the analysis with 17 RCTs and 13 non-RCTs. According to whether there was a clear catheter-related bloodstream infection at the beginning of the intervention, they were divided into the category of prevention (n=30) and treatment (n=1). In terms of study populations, there were 9 studies for children with intestinal failure, 13 for children with tumors, 4 for PICU and/or NICU, and 4 for others. Based on 29 studies focusing on different locking fluids (ethanol, taurolidine, urokinase, and antibiotics) to prevent catheter-related bloodstream infections, the overall risk density was RR=0.32(95%CI: 0.24-0.43) with statistically significant differences (P<0.01). Random effect model was used with I2=75%. Among them, there were 8 RCTs taking antibiotics as the locking fluid (RR=0.38, 95%CI: 0.18-0.82, P<0.05, I2=70%). Clinical heterogeneity was likely to come from patients with tumors, immunosuppression and vancomycin. Three RCTs were about urokinase (RR=0.71, 95%CI: 0.53-0.95, I2=0). Ten studies were about ethanol (RR=0.31, 95%CI: 0.22-0.43, P< 0.01, I2=57%) and clinical heterogeneity was likely to come from patients with intestinal failure. Eight studies were about (RR=0.23, 95%CI: 0.12-0.42, P<0.01,I2=72%) and clinical heterogeneity was likely to come from patients with intestinal failure, surgery and customized high-risk patients. One study was about catheter-related bloodstream infections treated with hydrochloric acid as the locking fluid (RR=0.84, 95%CI: 0.52-1.37, P=0.49). Among the secondary outcomes, there were statistically significant differences between the intervention group of catheter bacterial colonization and the control group only when taurolidine was used as the locking fluid. Adverse reactions after antimicrobial lock therapy included abnormal taste, vomiting, nausea, dizziness, and flushing. ConclusionThe use of locking fluids can reduce the risk density of CVAD-related bloodstream infections by 68%. Taurolidine has the best preventive effect, followed by ethanol, antibiotics and urokinase in a descending order. However, the effect of antimicrobial lock therapy may be overstated, and further research is likely to affect the credibility of the results of this efficacy evaluation.