Abstract

DIAZ ALVAREZ, Manuel. A bacterial meningitis predictive model for newborns with cerebrospinal fluid pleocytosis. Rev Cubana Pediatr [online]. 2013, vol.85, n.3, pp.356-370. ISSN 0034-7531.

Introduction: there were developed a prediction tools for bacterial meningitis (BM) in pediatric patients, but those do not include the neonatal period. Objetive: to develop and validate a prediction model for BM in neonatal patients who present pleocytosis of the cerebrospinal fluid (CF) when they were evaluated by clinical and laboratory variables. Patients and method: retrospective and analytic study, which included 320 neonatal patients evaluated for probable infection, with the diagnosis of meningitis, 45 with BM and 275 with aseptic meningitis, and who were admitted in the Neonatology Ward at the Universitary Pediatric Hospital «Juan M. Márquez», from february of 1992 up to december of 2009. There was developed a Predictive Model for BM, owing to some variables identified in uni and multivariate analysis. This model was developed from a «derivation group» and it was contrasted in the «validation group». Results: the variables selected to be included at the predictive model were: 1) Toxic appearance, 2) Seizures in the course of the illness, 3) positive CF Gram stain, 4) Polymorphonuclear leukocytes proportion in CF >80%, and 5) CF glucose concentration <1.0 mMol/l. The values for the variables as a whole revealed to have 100.0% sensitivity and negative predictive values. With this proposed model there were not any patient misclassified with negative prediction for BM, who result bearer of this kind of infection, in the derivation group. In the validation group the model had a similar performance, which accredit it like an internally validate model. Conclusions: the prediction model developed rest on rutinary evaluation variables, and it allowed to identify without errors all neonatal patients with BM. The ejecution of this model would be an usefull tool in the clinical decisions making.

Keywords : newborn infant; bacterial meningitis; aseptic meningitis; prediction models; serious bacterial infection; cerebrospinal fluid.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )