Predictors of hospital mortality and mechanical ventilation in patients with cervical spinal cord injury.

PURPOSE The objective of this study was to identify predictors of death and mechanical ventilation in patients with traumatic cervical spinal cord injury. METHODS From 1981 to 1994, 72 patients with traumatic cervical spinal cord injury resulting in neurological deficits were identified in this retrospective study. For each patient, neurological and associated injuries, physiological variables, complications, hospital mortality and the need for mechanical ventilation were recorded. Univariate and multivariate logistic regression analyses were done to identify predictors of mortality and the need for mechanical ventilation. RESULTS Fifteen patients (21%) died in the first three months after injury. Univariate analyses identified age, heart disease, neurological level at C4 and above, GCS < or = 13, forced vital capacity and cough, to be associated with mortality. Multivariate logistic regression identified age (P = 0.01), neurological level (P = 0.03) and GCS (P = 0.05) as independent predictors of mortality. In 41 patients (57%), the lungs were mechanically ventilated. Univariate analyses identified. The following predictors of the need for mechanical ventilation: neurological level at C5 and above, complete cord lesions, copious sputum, pneumonia and lung collapse. Multivariate logistic regression identified copious sputum (P = 0.01) and pneumonia (P = 0.01) as independent predictors of the need for mechanical ventilation. CONCLUSION Age, neurological level and GCS are independent predictors of mortality in patients with traumatic cervical spinal cord injury. Copious sputum and pneumonia are independent predictors of the need for mechanical ventilation.