Long-term survival of veterans with traumatic spinal cord injury.

OBJECTIVE: To investigate the long-term survival of veterans with traumatic spinal cord injury (SCI). DESIGN: Survival in a retrospective inception cohort of veterans suffering service-connected traumatic SCI is compared with survival among veterans disabled by other conditions, survival among nondisabled veterans, and a population-based life table. SETTING: Subjects were identified from a national census of veterans with service-connected disabilities, using a selection algorithm based on disability codes. PATIENTS: A retrospective cohort of 5545 male veterans with traumatic SCI, surviving at least 3 months after injury, is compared with a stratified random sample of 7077 disabled veterans without SCI, a stratified random sample of 6967 nondisabled veterans, and a life table formed from similarly aged American males. MAIN OUTCOME MEASURE: Survival curves, extending from 3 months to 40 years after injury. RESULTS: The mean life expectancy of veterans suffering traumatic SCI and surviving at least 3 months is an additional 39 years after injury, 85% that of similarly aged American males. Although survival with traumatic SCI was comparable to that of the disabled control subjects for approximately 20 years after onset, a clear deficit occurred beyond this point. Older age at injury is a stronger predictor of poorer long-term survival than is complete quadriplegia. CONCLUSIONS: Among patients who survive the acute phase of their traumatic SCI, long-term survival is relatively good. Health care planners, providers, and communities should anticipate an increasing number of persons aging with SCI.

Duke Scholars

Author Gregory P. Samsa Biostatistics & Bioinformatics, Division of Biostatistics