Operative treatment of spontaneous spinal epidural hematomas: a study of the factors determining postoperative outcome.

OBJECTIVE: We clarify the factors affecting postoperative outcomes in patients who have suffered spontaneous spinal epidural hematomas. METHODS: We review 330 cases of spontaneous spinal epidural hematomas from the international literature and three unpublished cases of our own. Attention was focused on sex, age, medical history, mortality, size and position of the hematoma, vertebral level of the hematoma, preoperative neurological condition, operative interval, and postoperative result. RESULTS: Sex, age, and size and position of the hematoma did not correlate with postoperative outcome. Mortality correlated highly with cervical or cervicothoracic hematomas, especially in patients with cardiovascular disease and those undergoing anticoagulant therapy. Incomplete preoperative sensorimotor deficit correlated highly with favorable outcomes (P < 0.0005), and recovery was significantly better when decompression was performed in < or = 36 hours in patients with complete sensorimotor loss (P < 0.05) and in < or = 48 hours in patients with incomplete sensorimotor deficit (P < 0.005). CONCLUSION: The critical factors for recovery after spontaneous spinal epidural hematoma are the level of preoperative neurological deficit and the operative interval. The vertebral level of the hematoma did not correlate with postoperative results, which suggests that local compression, rather than vascular obstruction, is the main factor in producing neurological deficit.