Charles H Bombardier1, Darren C Lee2, Debbie L Tan2, Jason K Barber3, Jeanne M Hoffman2. 1. Department of Rehabilitation Medicine, University of Washington School of Medicine, Seattle, WA. Electronic address: chb@uw.edu. 2. Department of Rehabilitation Medicine, University of Washington School of Medicine, Seattle, WA. 3. Department of Neurological Surgery, University of Washington, Seattle, WA.
Abstract
OBJECTIVES: To determine the diagnostic accuracy of a self-report measure of traumatic brain injury (TBI) in people with spinal cord injury (SCI), and to assess the potential effect of TBI on acute and postacute outcomes. DESIGN: Incident TBI as determined by self-report was compared with systematic medical records review by physicians blinded to self-reported TBI. SETTING: Inpatient rehabilitation unit. PARTICIPANTS: From 155 consecutive admissions, participants (N=105; 73% men; mean age, 45.9y; 73% white; 58% with tetraplegia) who met inclusion criteria and were admitted on the day of injury were recruited and assessed. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURE: Step 2 of the Traumatic Brain Injury-4 (TBI-4) interview was used to estimate the presence and severity of TBI. RESULTS: Estimated incidence of TBI was 33% by chart review and 60% based on Step 2 of the TBI-4. Ninety-one percent of those with TBI based on chart review sustained mild injuries. At the optimal cutoff of "possible TBI," the TBI-4 had a sensitivity of 83% (95% confidence interval, 66%-93%), a specificity of 51% (95% confidence interval, 39%-64%), and a Youden Index of 0.3 (95% confidence interval, 0.2-0.5). Compared with those without TBI, those with chart review-determined TBI had nonsignificant trends toward a longer acute care length of stay and a higher functional status at rehabilitation discharge, but equivalent 1-year outcomes. CONCLUSIONS: Step 2 of theTBI-4 did not meet Youden's criteria (≥0.8) for good diagnostic accuracy in the context of recent SCI. Comorbid TBIs were almost exclusively mild and not associated with poorer outcomes 1 year after SCI.
OBJECTIVES: To determine the diagnostic accuracy of a self-report measure of traumatic brain injury (TBI) in people with spinal cord injury (SCI), and to assess the potential effect of TBI on acute and postacute outcomes. DESIGN: Incident TBI as determined by self-report was compared with systematic medical records review by physicians blinded to self-reported TBI. SETTING: Inpatient rehabilitation unit. PARTICIPANTS: From 155 consecutive admissions, participants (N=105; 73% men; mean age, 45.9y; 73% white; 58% with tetraplegia) who met inclusion criteria and were admitted on the day of injury were recruited and assessed. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURE: Step 2 of the Traumatic Brain Injury-4 (TBI-4) interview was used to estimate the presence and severity of TBI. RESULTS: Estimated incidence of TBI was 33% by chart review and 60% based on Step 2 of the TBI-4. Ninety-one percent of those with TBI based on chart review sustained mild injuries. At the optimal cutoff of "possible TBI," the TBI-4 had a sensitivity of 83% (95% confidence interval, 66%-93%), a specificity of 51% (95% confidence interval, 39%-64%), and a Youden Index of 0.3 (95% confidence interval, 0.2-0.5). Compared with those without TBI, those with chart review-determined TBI had nonsignificant trends toward a longer acute care length of stay and a higher functional status at rehabilitation discharge, but equivalent 1-year outcomes. CONCLUSIONS: Step 2 of theTBI-4 did not meet Youden's criteria (≥0.8) for good diagnostic accuracy in the context of recent SCI. Comorbid TBIs were almost exclusively mild and not associated with poorer outcomes 1 year after SCI.