Objective: Elaborate and assess the degree of validity of a prognostic model for evaluating patients admitted to the Neonatal Intensive Care Unit (NICU).
Design: Cases and controls nested in a cohort.
Setting: NICU of two tertiary hospitals and another second level hospital.
Patients: The study was carried out in two phases. During the first phase (elaboration of a model), 336 newborns were studied; 112 belonged to the case group (dead patients) and 224 to the control group (live patients discharged). For the second phase (model validation), 300 patients were included that did not participate in the first phase, 100 cases and 200 controls.
Measurements: For each patient admitted to the study, clinical, paraclinical, perinatal and comorbidity factors were determined within the first 12 hours. Variables of statistical significance in the bivariate analysis were included in a logistic regression model with the objective of identifying a prognostic model.
Results: The variables that constituted the prognostic index were gestational age x birth weight, the paO2/FiO2 ratio x O2 saturation, arrest cardiac, major congenital malformations, septicemia and base excess. The model showed to have a sensitivity of 70% and a specificity of 91% during the elaboration cohort. In the validation cohort, sensitivity was 68% and specificity was 92%, a positive predictive value of 80%, negative predictive value of 85% and a correct classification rate was 84%.
Conclusions: The Neonatal Mortality Prognostic Index (NMPI) developed in this study showed to be useful for the evaluation of hospital mortality for severely ill newborns admitted to NICU.