Current use and timing of spinal surgery for management of acute spinal cord injury in North America: results of a retrospective multicenter study

Object. A multicenter retrospective study was performed in 36 North American centers to examine the use and timing of surgery in patients who have sustained acute spinal cord injury (SCI). The study was performed to obtain information required for the planning of a randomized controlled trial in which early and late decompressive surgery are compared. Methods. The records of all patients aged 16 to 75 years with acute SCI admitted to 36 centers within 24 hours of injury over a 9-month period in 1994 and 1995 were examined to obtain data on admission variables, methods of diagnosis, use of traction, and surgical variables including type and timing of surgery. A total of 585 patients with acute SCI or cauda equina injury were admitted to participating centers, although approximately half were ultimately excluded because they did not meet inclusion criteria. Common causes for exclusion were late admission, age, gunshot wound, and absence of signs of compression on imaging studies. Thus, only approximately 50% of patients with acute SCI would be eligible for inclusion in a study of acute decompressive surgery. Although all patients underwent computerized tomography (CT) scanning, only 54% underwent magnetic resonance imaging, and CT myelography was performed in only 6%. Complete neurological injuries (American Spinal Injury Association Grade A) were present in 57.8%. Traction was applied in only 47% of patients who sustained cervical injury, in whom decompressive traction was successful in only 42% of cases. Neurological deterioration occurred in 8.1% of cases after traction. Surgery was performed in 65.4% of patients. The timing of surgery varied widely: less than 24 hours postinjury in 23.5%, between 25 and 48 hours postinjury in 15.8%, between 48 and 96 hours in 19%, and more than 5 days postinjury in 41.7% of patients. Conclusions. These data indicate that although surgery is commonly performed in patients with acute SCI, one third of cases are managed nonoperatively, and there is very little agreement on the optimum timing of surgical treatment. The results of this study confirm the need for a randomized controlled trial to assess the optimum timing of decompressive surgery in SCI.