Extension of the representativeness of the Traumatic Brain Injury Model Systems National Database: 2001 to 2010.

OBJECTIVE: To extend the representativeness of the Traumatic Brain Injury Model Systems National Database (TBIMS-NDB) for individuals 16 years and older, admitted for acute, inpatient rehabilitation in the United States with a primary diagnosis of traumatic brain injury (TBI) analyses completed by Corrigan and colleagues by comparing this data set to national data for patients admitted to inpatient rehabilitation with identical inclusion criteria that included 3 additional years of data and 2 new demographic variables.
DESIGN: Secondary analysis of existing data sets and extension of previously published analyses.
SETTING: Acute inpatient rehabilitation facilities. PARTICIPANTS: Patients 16 years and older with a primary rehabilitation diagnosis of TBI; the US TBI Rehabilitation population, n = 156 447; and the TBIMS-NDB population, n = 7373.
INTERVENTIONS: None. MAIN OUTCOME MEASURES: Demographics, functional status, and length of stay in hospital.
RESULTS: The TBIMS-NDB was largely representative of patients 16 years and older, admitted for rehabilitation in the United States with a primary diagnosis of TBI on or after October 1, 2001, and discharged as of December 31, 2010. The results of the extended analyses were similar to those reported by Corrigan and colleagues. Age accounted for the largest difference between the samples, with the TBIMS-NDB including a smaller proportion of patients 65 years and older than all those admitted for rehabilitation with a primary diagnosis of TBI in the United States. After partitioning each data set at age 65, most distributional differences found between samples were markedly reduced; however, differences in the preinjury vocational status of the employed and rehabilitation lengths of stay between 1 and 9 days remained robust. The subsamples of patients 64 years and younger were found to differ only slightly on all remaining variables, whereas those 65 years and older were found to have meaningful differences in insurance type and age distribution.
CONCLUSIONS: These results reconfirm that the TBIMS-NDB is largely representative of patients with TBI receiving inpatient rehabilitation in the United States. Differences between the 2 data sets were found to be stable across the 3 additional years of data, and new differences were limited to those involving newly introduced variables. To use these data for population-based research, it is strongly recommended that statistical adjustment be conducted to account for the lower percentage of patients older than 65 years, inpatient rehabilitation stays less than 10 days, and preinjury vocational status in the TBIMS-NDB.