In Reply Dr Selvarajah and colleagues note that patients with traumatic spinal cord injury may die at the scene of the incident or be discharged from the emergency department without being admitted as inpatients. As discussed in our article, these patients were not included in our calculations because the NIS includes only those who were admitted as inpatients.
The 2 regional studies referenced by Selvarajah and colleagues in which 38% died before admission included 619 cases from California in 1970-1971 based in part on the identification of spinal cord injury using autopsy records1 and 154 cases from Olmsted County, Minnesota, from 1935-1981.2 In later studies based on state surveillance, 9% of patients in Utah in 1989-19913 died before hospital admission, as did approximately 23% in Mississippi in 1992-1994.4 These numbers suggest that older prehospital mortality data may not be able to be extrapolated to contemporary estimates of spinal cord injury given changes in seatbelt laws, traffic safety, motor vehicle designs, public awareness, drunk driving laws, changes in etiology with an aging population and increasing numbers of falls, and improvements in acute management of trauma.
Jain NB, Harris MB, Garshick E. Trends in Traumatic Spinal Cord Injury—Reply. JAMA. 2015;314(15):1643-1644. doi:10.1001/jama.2015.11200