Objective: To identify the benefits and harms of hyperbaric oxygen therapy (HBOT) to treat traumatic brain injury (TBI).
Data sources: MEDLINE, EMBASE, the Cochrane Library, HealthSTAR, CINAHL, MANTIS, professional society databases, and reference lists. Databases were searched from inception through December 2003.
Study selection: We included English-language studies of patients with TBI given HBOT and evaluating functional health outcomes.
Data extraction: Data were abstracted by 1 reviewer and checked by a second. Study quality was rated as good, fair, or poor.
Data synthesis: Two fair-quality randomized controlled trials of patients with severe brain injury reported conflicting results. One found no difference in mortality (48% HBOT vs 55% control) or morbidity at 1 year. In young patients with brainstem contusion, significantly more regained consciousness at 1 month with HBOT (67%) than control (11%) (P<.03). The other found a significant decrease in mortality in the HBOT group at 1 year (17%) compared with controls (31%) (P=.037). This decrease in mortality was accompanied by an increase in proportion of patients with severe disability. Patients with intracranial pressure (ICP) greater than 20 mmHg or a Glasgow Coma Scale score of 4 to 6 had significantly lower mortality at 1 year than controls. Five observational studies did not provide better evidence of effectiveness or adverse events. Two indicated a potential for initially reducing elevated ICP in some patients. However, rebound elevations higher than pretreatment levels occurred in some patients. Adverse events, including seizures, pulmonary symptoms, and neurologic deterioration, were reported; however, no study systematically assessed adverse events, and none reported adverse events in control groups.
Conclusions: The evidence for HBOT for TBI is insufficient to prove effectiveness or ineffectiveness, and more high-quality studies are needed. The evidence indicates that there is a small chance of a mortality benefit, which may depend on subgroup selection. The effect on functional status and the incidence and clinical significance of adverse effects are unclear.