Central nervous system infections after military missile head wounds.

Abstract

To evaluate variables instrumental in central nervous system infections after military missile head wounds, using uni- and multivariate analysis in 964 patients during the 8-year Iran-Iraq War. Factors considered in this retrospective study were: the types of projectile, mode of injury, paranasal sinus involvement, number of lobes involved, transventricular injuries, place of exploration (base hospital or Nemazee Hospital), cerebrospinal fluid (CSF) fistulas, Glasgow Coma Scale (GCS) score, retained bone, and retained shell fragments. During the study period, 105 patients (11%) developed central nervous system infections, including 20 abscesses, 1 case of cerebritis, 2 cases of fungus cerebri, and 82 cases of meningitis. gram-negative organisms, especially Klebsiella pneumoniae, were the most frequent offending organisms. Forty-one percent of the 133 deaths were due to infections, but the death rate from infection was only 4.4%. Univariate analysis showed mode of injury, number of lobes involved, ventricular penetration, paranasal sinus involvement, CSF fistulas, place of exploration, GCS score, and retained bone fragments to have significant bearing on the incidence of central nervous system infections. On the other hand, multivariate regression analysis disclosed the following factors each enhancing infection: CSF fistulas (chi2 = 46.526), transventricular injuries (chi2 = 13.4790), and paranasal petrous sinuses involvement (chi2 = 4.2221). When compared with primary exploration at the Nemazee Hospital, both exploration at a base hospital and no exploration at all were associated with increased chances of infection (chi2 = 4.7629 and 8.3220, respectively). Additionally, when tangential, crossed penetrating, and uncrossed penetrating injuries were compared with through-and-through injuries, the uncrossed penetrating mode was associated with less infection (chi2 = 0.1652, 2.6353, and 5.0817, respectively). Only two patients were readmitted for new evidence of infection 3 and 5 months after missile head wounds, one definitely due to and the other on suspicion of CSF fistulas. One hundred and thirty-seven of 587 patients with retained bone fragments were followed a mean of 42 months with no evidence of delayed infection. In this study, CSF fistulas and transventricular and paranasal sinus injuries all were associated with increased chances of central nervous system infections after military missile head wounds. Infection rate was lower in penetrating injuries not crossing into another dural compartment. Exploration at the Nemazee Hospital, despite delays in evacuation, had less incidence of infection than surgery at a base hospital within the first 24 hours of injury. Retained bone and metal fragments, a lower GCS score at the time of admission, secondary exploration at the Nemazee Hospital, and number of lobes involved were less important when evaluated in a multivariate regression model.