The Traumatic Brain Injury Model Systems National Database: A Review of Published Research

Temkin

Barber

et al. 2023

Annals of Neurology

ObjectiveIt is not currently possible to predict long‐term functional dependency in patients with disorders of consciousness (DoC) after traumatic brain injury (TBI). Our objective was to fit and externally validate a prediction model for 1‐year dependency in patients with DoC 2 weeks after TBI.MethodsWe included adults with TBI enrolled in TBI Model Systems (TBI‐MS) or Transforming Research and Clinical Knowledge in TBI (TRACK‐TBI) studies who were not following commands at rehabilitation admission or 2 weeks post‐injury, respectively. We fit a logistic regression model in TBI‐MS and validated it in TRACK‐TBI. The primary outcome was death or dependency at 1 year post‐injury, defined using the Disability Rating Scale.ResultsIn the TBI‐MS Discovery Sample, 1,960 participants (mean age 40 [18] years, 76% male, 68% white) met inclusion criteria, and 406 (27%) were dependent 1 year post‐injury. In a TBI‐MS held out cohort, the dependency prediction model's area under the receiver operating characteristic curve was 0.79 (95% CI 0.74–0.85), positive predictive value was 53% and negative predictive value was 86%. In the TRACK‐TBI external validation (n = 124, age 40 [16] years, 77% male, 81% white), the area under the receiver operating characteristic curve was 0.66 (0.53, 0.79), equivalent to the standard IMPACTcore + CT score (p = 0.8).InterpretationWe developed a 1‐year dependency prediction model using the largest existing cohort of patients with DoC after TBI. The sensitivity and negative predictive values were greater than specificity and positive predictive values. Accuracy was diminished in an external sample, but equivalent to the IMPACT model. Further research is needed to improve dependency prediction in patients with DoC after TBI. ANN NEUROL 2023

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Predicting Functional Dependency in Patients with Disorders of Consciousness: A TBI‐Model Systems and TRACK‐TBI Study

Temkin

Barber

et al. 2023

Annals of Neurology

“…We conducted a retrospective analysis of a prospective longitudinal cohort study of participants enrolled in the TBIMS NDB 23 . Characteristics of this continuously enrolling database have been described previously 1, 24, 25 . The sample includes participants who survived acute hospitalization and were admitted to one of 16 inpatient rehabilitation centers in the US that participated in the TBIMS program.…”

Section: Methodsmentioning

confidence: 99%

Comparison of common outcome measures for assessing independence in patients diagnosed with disorders of consciousness: A Traumatic Brain Injury Model Systems Study

Kowalski

Izzy

et al. 2022

Preprint

Importance Patients with disorders of consciousness (DoC) after traumatic brain injury (TBI) recover to varying degrees of functional dependency. Dependency is difficult to measure but critical for outcome interpretation and prognostic counseling. Traditional outcome measures, like the Glasgow Outcome Scale-Extended (GOSE), are mandated by the US Food and Drug Administration for evaluating TBI clinical trial efficacy but have an unknown accuracy for measuring dependency. Objective We used the Functional Independence Measure (FIM) as the reference standard to evaluate how accurately the GOSE and Disability Rating Scale (DRS) assess functional dependency in the worlds largest cohort of patients with DoC after TBI. We propose an alternate, data-driven, approach to measuring dependency. Design, Setting, and Participants In this cohort study, we included patients with DoC prospectively enrolled in the longitudinal Traumatic Brain Injury Model Systems National Database (TBIMS NDB). Participants were survivors of moderate/severe TBI with DoC on admission to a US inpatient rehabilitation center between 1988 and 2020, followed 1 year after injury. Exposures We examined the classification performance of common TBI outcome measure cut-points (GOSE <=3 and <=4 [Lower and Upper Severe Disability, respectively], and DRS >=12 [Severe Disability]) in identifying subjects with functional dependency at 1 year. We compared data-derived optimal cut-points on these scales to a novel DRS-based marker of dependency, the DRSDepend. Main Outcome and Measure Total FIM score < 80 (FIM-dependency) at 1 year. Results Of 18,486 TBIMS participants, 1,483 with DoC on arrival to inpatient rehabilitation met inclusion criteria (mean [SD] age=38 [18] years; 76% male). The sensitivity of GOSE cut-points of <=3 and <=4 for identifying FIM-dependency were 97% and 98%, but specificities were 73% and 51%, respectively. The sensitivity of the DRS cut-point of >=12 was 60%, but specificity was 100%. The DRSDepend had a sensitivity of 83% and a specificity of 94% for classifying FIM-dependency, with a greater AUROC than the data-derived optimal GOSE (<=3, p=0.01) and DRS (>=10, p=0.008) cut-points. Conclusions and Relevance Commonly-used GOSE and DRS cut-points have limited sensitivity or specificity for identifying functional dependency. The DRSDepend identifies FIM-dependency more accurately than GOSE and DRS cut-points, but requires further validation.

“…) develop and internally test a dependency (DRS Depend ) prediction model at 1-year post-injury in a large national database (TBI Model Systems [TBI-MS] 27 ); 2) evaluate model performance in an independent sample of patients enrolled in the Transforming Research and Clinical Knowledge in TBI (TRACK-TBI) 28 study; and 3) compare our model's performance to the International Mission for Prognosis and Analysis of Clinical Trials in TBI (IMPACT) 18 score, a frequently-used, validated 20 prognostic model.METHODSWe conducted a secondary analysis of two prospective, longitudinal cohorts: TBI-MS National Database27 and TRACK-TBI 28 (clinicaltrials.gov NCT02119182; https://tracktbi.ucsf.edu/). Characteristics of each cohort have been previously described1,2,[28][29][30][31] . Briefly, TBI-MS enrolls adults (≥ 16 years) with moderate or severe TBI who survive acute hospitalization and are admitted to inpatient rehabilitation centers in the US.…”

mentioning

confidence: 99%

Predicting Functional Dependency in Patients with Disorders of Consciousness: A TBI-Model Systems and TRACK-TBI Study

Temkin

Barber

et al. 2023

Preprint

Importance: There are currently no models that predict long-term functional dependency in patients with disorders of consciousness (DoC) after traumatic brain injury (TBI). Objective: Fit, test, and externally validate a prediction model for 1-year dependency in patients with DoC 2 or more weeks after TBI. Design: Secondary analysis of patients enrolled in TBI Model Systems (TBI-MS, 1988-2020, Discovery Sample) or Transforming Research and Clinical Knowledge in TBI (TRACK-TBI, 2013-2018, Validation Sample) and followed 1-year post-injury. Setting: Multi-center study at USA rehabilitation hospitals (TBI-MS) and acute care hospitals (TRACK-TBI). Participants: Adults with TBI who were not following commands at rehabilitation admission (TBI-MS; days post-injury vary) or 2-weeks post-injury (TRACK-TBI). Exposures: In the TBI-MS database (model fitting and testing), we screened demographic, radiological, clinical variables, and Disability Rating Scale (DRS) item scores for association with the primary outcome. Main Outcome: The primary outcome was death or complete functional dependency at 1-year post-injury, defined using a DRS-based binary measure (DRSDepend), indicating need for assistance with all activities and concomitant cognitive impairment. Results: In the TBI-MS Discovery Sample, 1,960 subjects (mean age 40 [18] years, 76% male, 68% white) met inclusion criteria and 406 (27%) were dependent at 1-year post-injury. A dependency prediction model had an area under the receiver operating characteristic curve (AUROC) of 0.79 [0.74, 0.85], positive predictive value of 53%, and negative predictive value of 86% for dependency in a held-out TBI-MS Testing cohort. Within the TRACK-TBI external validation sample (N=124, age 40 [16], 77% male, 81% white), a model modified to remove variables not collected in TRACK-TBI, had an AUROC of 0.66 [0.53, 0.79], equivalent to the gold-standard IMPACTcore+CT score (0.68; 95% AUROC difference CI: -0.2 to 0.2, p=0.8). Conclusions and Relevance: We used the largest existing cohort of patients with DoC after TBI to develop, test and externally validate a prediction model of 1-year dependency. Model sensitivity and negative predictive value were greater than specificity and positive predictive value. Accuracy was diminished in an external sample, but equivalent to the best-available models. Further research is needed to improve dependency prediction in patients with DoC after TBI.