Abstract

CRUZ PORTELLES, Alain et al. Mortality Predictive Factors in Severe Traumatic Brain Injury. CCM [online]. 2014, vol.18, n.3, pp. 415-429. ISSN 1560-4381.

Introduction: to know the prognosis of patients with some disease is one of the bases of the current clinical medicine. Objectives: to determine the main predictor of mortality at 28 days after severe traumatic brain injury (Glasgow ≤8) patients admitted at intensive care of V. I. Lenin Hospital. Methods: a case-control (1:1) study pared by age, Injury Severity Score, and the need of neurosurgery was performed. Discrete variables were compared using Chi squared test or Fisher exact test as needed and continuous one were compare with T-Student for α=0.05. A binary logistic regression model was established including statistically different variables to determine predictors of mortality according with Hosmer-Lemeshow. Results: of 253 included patients in the study, 84.6 % were males. Mortality was higher among older patients (48.5±19.3 vs. 55.9±18.7 years, p=0.002), >45 year-old (p=0.026), and higher APACHE II score (31±20.1 vs. 23±12.1, p<0.001); Glasgow coma scale on admission and at third day (p<0.001), and mean blood pressure were lower (65.3±10.4 vs. 84.2±12.5, p<0.001). Subarachnoid hemorrhage (24.6% vs. 37%, p=0.04) was associated with mortality. Multivariate analysis showed that a Marshall tomographic score of V-VI (OR=2.14, IC:1.28-3.57, p=0,001), one or both pupils unreactive (OR=5.46, IC: 2.99-9.95), >45 years (OR=1.92, IC:1.01-4.12), and subdural hematoma (OR=3.25, IC:1.18-9.00, p=0,023) were predictors of mortality. Conclusions: tomographic grade V-VI of Marshall, pupillary alterations, age >45 years, and the presence of subdural hematoma were predictors of mortality among the patients included in this research

Keywords : severe traumatic brain injury; odds ratio; mortality; logistic regression.

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