Epidemiology of incident spinal fracture in a complete population

Study Design. Cross-sectional observational study of incident spinal fractures using an administrative database. Objectives. To identify and define all patients who have spinal fractures within a complete population. Summary of Background Data. The true incidence of spinal column and cord injury is not known. Previous studies have been institutional or practice based. Accurate information concerning the magnitude of the spinal injury population and their characteristics may provide a more rational basis for public health decision making and resource allocation. Methods. The study dates were April 1, 1981 to March 31, 1984. Using the Manitoba Health Services Insurance Plan database, all patients with ICD-9-CM coding of 805.x and 806.x (spinal column fracture with and without spinal cord injury) were identified. Incidence rates, age and gender distribution, and ambulatory and hospital contacts were identified. Hospital discharge abstracts were used to classify mechanisms of injury, associated injuries, and length of stay. Results. The annual incidence rate of spinal fracture was 64 per 100,000. Two thousand sixty-three patients were identified, with .944 being admitted to the hospital. There were two peaks of incidence occurring in young men and elderly women. Of the hospitalized patients, 182 had cervical injury, 286 had thoracic fracture, and 403 had injury in the lumbosacral spine. Associated injuries occurred in 38% of hospitalized patients. Length of stay was an average of 38.5 days. Overall mortality was 4.1%. Neurologic injury occurred in 122 patients. Conclusions. Ambulatory care of spine injuries is more common than hospital care. Two peaks of incidence occur-in young men and elderly women. Future decisions for research, public health policy, and resource allocation can be based on these data.