Design: Cohort study.

Setting: Model spinal cord injury care systems and Shriner's Hospitals spinal cord injury units throughout the United States.

Patients: A total of 28,239 consecutive persons admitted to the model system or to a Shriner's Hospital within 1 year of injury.

Main Outcome Measure: Length of survival and cause of death.

Results: Among persons who were admitted to the model system within 1 day of injury, the odds of dying during the first postinjury year were reduced by 67% for persons injured between 1993 and 1998 relative to persons injured between 1973 and 1977 after adjusting for trends in age, gender, race, neurologic level of injury, Frankel grade, ventilator status, etiology of injury, sponsor of care, and model system where treatment occurred. However, mortality rates after the first anniversary of injury, which had also been declining from 1973 to 1992, increased 33% for persons injured between 1993 and 1998 relative to persons injured between 1988 and 1992. Respiratory disease was the only cause of death after the first anniversary of injury for which the relative odds increased meaningfully during the latest time period (76% increase over 1988-1992 compared to all other causes).

Conclusion: While great improvements in life expectancy have been achieved since the Model SCI Systems program began, current data support the need for renewed efforts to improve the prevention and treatment of the complications of spinal cord injury. (C) 1999 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation.