Background: Predictive factors for mortality are different among countries; knowledge of risk factors is important for planning strategies to reduce mortality in trauma. The objective of this study was to identify predictors of death in traumatic cervical spinal cord injury (TSCI) in Brazil.
Methods: From 2001 to 2005, 84 patients with clinically detectable TSCI were identified in this retrospective study. For each patient, neurologic and associated injuries, physiologic variables, complications, treatment, and hospital mortality were recorded. Bivariate and multivariate logistic regression analyses were done to identify predictors of mortality.
Results: Twenty-two (26.2%) patients died in hospital. Car crash (39%) and falls (37.85) were the most frequent causes of trauma. The causes of death were as follows: neurologic, 8 (36.4%); respiratory, 4 (18.2%); septic complications, 2 (9.1%); venous thrombosis and embolism, 2 (9.1%); and undetermined, 4 (18.2%). Bivariate analysis identified absent neurologic function (risk ratios [RR] = 4.5; 95% confidence intervals [CI], 1.6-12.7), high injury severity score (p = 0.001) and low revised trauma score (p = 0.001); Glasgow Coma Scale (GCS) score <9 (RR = 47.4; 95% CI, 5.4-413.2); shock at admission (RR = 2.5; 95% CI, 0.8-7.9); vasopressor use (RR = 25.8; 95% CI, 6.1-109.6); mechanical ventilation (RR = 31.9; 95% CI, 6.6-154.0); acute renal insufficiency (RR = 10.0; 95% CI, 0.98-102.1) as associated with mortality. The mainly independents predictors for mortality were GCS score <9 and vasopressor use. Lowest mortality rate (5.2%) was observed for patients with TSCI alone.
Conclusions: GCS score <9, mechanical ventilation, and vasopressor use were predictors of mortality with TCSI, and if these risk factors were absent, we observed low mortality rate.