Diagnóstico de sepse neonatal tardia associada ao cateter venoso central em unidade neonatal segundo critérios nacionais e internacionais

Abstract

Objective: To investigate the diagnostic criteria of the late-onset (LOS) infections associated with the central venous catheter (CVC) in a Neonatal unit of a university Hospital of Minas Gerais. Method: prospective cohort study, where the clinical and laboratory data of the neonates with suspicion of late-onset neonatal sepsis associated with CVC were collected of the newborns admitted in the neonatal unit of the Hospital das Clínicas of the Federal University of Minas Gerais from December 2015 to May 2017. All newborns submitted to the CVC were admitted to the study and those, in which the catheter was inserted in an external environment to the unit, were excluded. The diagnosis of LOS neonatal sepsis associated with the CVC followed the ANVISA and CDC/NHSN criteria. For both criteria, univariate models were adjusted and variables whose p-value was 0,20, were included for the multivariate model. In the multivariate analysis of the ANVISA criterion, separate adjustments were made, with group 1 consisting of clinical signs, the group 2 of the laboratory tests and in the group 3, were made up of the adjustment variables of the groups 1 and 2. For both criteria, the variables with value-p 0.05 were kept in the final model. The quality of adjustment of the models was verified by the test of Hosmer & Lemeshow and the statistical analysis was performed in SPSS version 20.0. Results: Of the 191 newborns with CVC evaluated, 92 newborns had suspicion of sepsis. According to ANVISA criteria, 47 newborns had an episode of LOS sepsis associated with CVC, and three presented two episodes each, totaling 53 episodes (33 laboratory confirmed bloodstream infection and 20 clinical bloodstream). According to the CDC / NHSN criteria, 20 episodes were classified as LOS neonatal sepsis associated with the CVC. According to ANVISA criteria, the clinical data candidates for the multivariate model were thermal instability, glucose intolerance, hemodynamic instability and food intolerance. For laboratory variables, C-reactive protein, platelet count, immature neutrophil / total neutrophil ratio and neutropenia remained. In the clinical data and laboratory data groups, the variables significantly associated with CVC-associated late-onset neonatal sepsis were hemodynamic instability (p = 0.004), C-reactive protein (p = 0.047), and immature neutrophils / total neutrophils ratio (p = 0.011). After readjustment of the analysis, the C reactive protein (p = 0.047) and the total immature neutrophils / neutrophils ratio (p =0.011) remained which were significant at the 0.05 level. According to the CDC/NHSN criteria, the variables that were candidates for the multivariate model were hypothermia (p = 0.018) and hyperthermia (p = 0.012). After multivariate adjustment, hypothermia (p = 0.008) and hyperthermia (p = 0.006) remained. Conclusion: Hemodynamic instability, PCR and I/T ratio showed significance in the diagnosis of LOS neonatal sepsis associated with the CVC in infants according to ANVISA criteria and hypothermia and hyperthermia according to CDC/NHSN criteria.