Cumulative Evidence of Randomized Controlled and Observational Studies on Catheter-Related Infection Risk of Central Venous Catheter Insertion Site in ICU Patients: A Pairwise and Network Meta-Analysis

Background: Selection of central venous catheter insertion site in ICU patients could help reduce catheter-related infections. Although subclavian was considered the most appropriate site, its preferential use in ICU patients is not generalized and questioned by contradicted meta-analysis results. In addition, conflicting data exist on alternative site selection whenever subclavian is contraindicated. Objective: To compare catheter-related bloodstream infection and colonization risk between the three sites (subclavian, internal jugular, and femoral) in adult ICU patients. Data Source: We searched MEDLINE, EMBASE, Cochrane Central Register of Controlled trials, CINAHL, and clinicaltrials.gov. Study Selection: Eligible studies were randomized controlled trials and observational ones. Data Extraction: Extracted data were analyzed by pairwise and network meta-analysis. Data Synthesis: Twenty studies were included; 11 were observational, seven were randomized controlled trials for other outcomes, and two were randomized controlled trials for sites. We evaluated 18,554 central venous catheters: 9,331 from observational studies, 5,482 from randomized controlled trials for other outcomes, and 3,741 from randomized controlled trials for sites. Colonization risk was higher for internal jugular (relative risk, 2.25 [95% CI, 1.84-2.75]; I-2 = 0%) and femoral (relative risk, 2.92 [95% CI, 2.11-4.04]; I-2 = 24%), compared with subclavian. Catheter-related bloodstream infection risk was comparable for internal jugular and subclavian, higher for femoral than subclavian (relative risk, 2.44 [95% CI, 1.25-4.75]; I-2 = 61%), and lower for internal jugular than femoral (relative risk, 0.55 [95% CI, 0.34-0.89]; I-2 = 61%). When observational studies that did not control for baseline characteristics were excluded, catheter-related bloodstream infection risk was comparable between the sites. Conclusions: In ICU patients, internal jugular and subclavian may, similarly, decrease catheter-related bloodstream infection risk, when compared with femoral. Subclavian could be suggested as the most appropriate site, whenever colonization risk is considered and not, otherwise, contraindicated. Current evidence on catheter-related bloodstream infection femoral risk, compared with the other sites, is inconclusive.