Diagnostic accuracy of semi-quantitative and quantitative culture techniques for the diagnosis of catheter-related infections in newborns and molecular typing of isolated microorganisms
1 Departamento de Microbiologia e Imunologia, Instituto de Biociências, UNESP - Univ Estadual Paulista, Botucatu, SP, Brasil
2 Departamento de Pediatria, Faculdade de Medicina, UNESP - Univ Estadual Paulista, Botucatu, SP, Brasil
3 Departamento de Bioestatística, Instituto de Biociências, UNESP - Univ Estadual Paulista, Botucatu, SP, Brasil
BMC Infectious Diseases 2014, 14:283 doi:10.1186/1471-2334-14-283Published: 22 May 2014
Catheter-related bloodstream infections (CR-BSIs) have become the most common cause of healthcare-associated bloodstream infections in neonatal intensive care units (ICUs). Microbiological evidence implicating catheters as the source of bloodstream infection is necessary to establish the diagnosis of CR-BSIs. Semi-quantitative culture is used to determine the presence of microorganisms on the external catheter surface, whereas quantitative culture also isolates microorganisms present inside the catheter. The main objective of this study was to determine the sensitivity and specificity of these two techniques for the diagnosis of CR-BSIs in newborns from a neonatal ICU. In addition, PFGE was used for similarity analysis of the microorganisms isolated from catheters and blood cultures.
Semi-quantitative and quantitative methods were used for the culture of catheter tips obtained from newborns. Strains isolated from catheter tips and blood cultures which exhibited the same antimicrobial susceptibility profile were included in the study as positive cases of CR-BSI. PFGE of the microorganisms isolated from catheters and blood cultures was performed for similarity analysis and detection of clones in the ICU.
A total of 584 catheter tips from 399 patients seen between November 2005 and June 2012 were analyzed. Twenty-nine cases of CR-BSI were confirmed. Coagulase-negative staphylococci (CoNS) were the most frequently isolated microorganisms, including S. epidermidis as the most prevalent species (65.5%), followed by S. haemolyticus (10.3%), yeasts (10.3%), K. pneumoniae (6.9%), S. aureus (3.4%), and E. coli (3.4%). The sensitivity of the semi-quantitative and quantitative techniques was 72.7% and 59.3%, respectively, and specificity was 95.7% and 94.4%. The diagnosis of CR-BSIs based on PFGE analysis of similarity between strains isolated from catheter tips and blood cultures showed 82.6% sensitivity and 100% specificity.
The semi-quantitative culture method showed higher sensitivity and specificity for the diagnosis of CR-BSIs in newborns when compared to the quantitative technique. In addition, this method is easier to perform and shows better agreement with the gold standard, and should therefore be recommended for routine clinical laboratory use. PFGE may contribute to the control of CR-BSIs by identifying clusters of microorganisms in neonatal ICUs, providing a means of determining potential cross-infection between patients.