Data Summaries

The TBIMS Centers collect data for the first prospective, longitudinal, multicenter study ever conducted which examines the course of recovery and outcomes following the delivery of a coordinated system of acute neurotrauma and inpatient rehabilitation. The NDSC stores the data in the NDB and makes it available for researchers to conduct research that contributes to evidence-based rehabilitation interventions and clinical and practice guidelines that improve the lives of individuals with TBI.

The objectives for the NDB are to:

The data is collected via repeated surveys of individuals at regular intervals post-injury. This includes an interview administered at inpatient rehabilitation discharge, called Form 1, which includes <336> variables. It further includes follow-up interviews administered via telephone, in person, or mailout questionnaire, called Form 2, which includes <315> variables. Sources of data include abstraction of medical records, specialized data collection forms, patient interview and testing, and family interview. The patient is the primary source of follow-up information; if the patient cannot be interviewed, follow-up is attempted with a proxy.

The NDSC performs multiple data quality checks to ensure the integrity of the data in the NDB. These include checks at the point of data entry, such as checks for valid codes and correct ranges, logical checks between variables, and consistency checks between variables over time. The data entry system will also identify cases with errors or blanks, notify of follow-ups coming due, warn about overdue follow-ups, calculate missing data rates, and calculate follow-up rates.

The TBIMS NDB has been used in <266> studies from its inception through the end of calendar year <2023>. These studies have focused on the epidemiology of moderate-to-severe TBI, the natural history of TBI outcomes and comorbidities, predictors of TBI outcomes and comorbidities, validation of severity and outcome measurement, and longitudinal change over time.

Data categories include demographic characteristics of the population, causes and severity of injury, pre-injury conditions and limitations, disability, health measurements, mood and behavior measurements, and participation outcomes. The charts and tables presented in the next section comprise a descriptive summary of a subset of the variables in the TBIMS NDB. All variable definitions and histories can be found in the [data dictionary]https://www.tbindsc.org/DataDictionary.aspx.

Note

The following chapters are devoted to reviewing data elements that are currently captured in the TBIMS. Overall the dataset is reduced to persons who were admitted to inpatient rehabilitation between 1989-10-01 and 2025-12-31 and whose followup window have closed as of 2025-12-31. Furthermore, the data set has been restricted to follow-up years 1,2,5,10 and every 5 years after. There exists 5282 follow-ups that were conduced in the “off” years that are not represented in this dataset.

Date Added

The data set for each measure has been filtered for when the variable was added to the database. For example if a variable was added October 1, 2017 then all cases with a rehab admission date or followup date prior would be filtered out. In the rare instance where there were multiple add dates this report includes data only where the variable was asked.

Tables and graphs

Where appropriate, calculated fields are used for exploration. This will help reveal a “true missingness” of the variable. Each Data element will have a table representing the data as well as a graph visualizing the data. Further, each table will have a missingness column or row depending on the type of variable. There will be a sentence describing the overall missingness of the variable (either by Form 1 or Form 2). If there are harmonized variables between Form 1 and 2. the data will be stacked. Form 1 variables referencing “at the time of injury” will be labeled “INJ” others will be “ADM” or “DIS” representing Admission or Discharge Scores. Form2 Variables will be designated by the followup year.