Abstract

Background: Accurate diagnosis of catheter-related blood stream infection (CRBSI) is necessary to make a decision about removal of the catheter. Differential time to positivity (DTP) and the ratio of quantitative cultures (RQC) between central and peripheral blood cultures have not been evaluated against a strict standard in children, namely catheter tip culture. Objective: Our aim is to compare DTP and RQC in the diagnosis of catheter tip-confirmed catheter-related infection in children. Method: Prospective study performed in 2 large hospitals in Santiago, Chile. Children with clinically suspected CRBSI had 2 peripheral and central vein blood samples obtained for automated culture in Bact/Alert and for quantitative cultures in 5% sheep blood agar plate. The catheter tip was cultured. Sensitivity, specificity, positive predictive value, negative predictive value, likelihood ratios (LR), and accuracy of DTP and RQC were compared against catheter tip-confirmed CRBSI. Results: During a 3-year period, 344 clinically suspected CRBSIs were diagnosed in children of which 124 episodes met study criteria. Catheter tip culture-confirmed CRBSI in 25 (20%) of 124 episodes. A total of 34 microorganisms were cultured from 25 CRBSI; 8 of 25 (32%) episodes were polymicrobial. Staphylococcus aureus followed by coagulase-negative Staphylococcus were the most common microorganisms. For CRBSI, DTP and RQC reached a sensitivity of 75% versus 24% (P < 0.001), specificity of 86 versus 94%, positive predictive value of 58% versus 50%, negative predictive value of 93% versus 82%, LR of 5.48 versus 4.50, and accuracy of 0.84 versus 0.79. Conclusions: In children, DTP was better than RQC for diagnosis of catheter tip-confirmed CRBSI.