Abstract 360: Evaluating the Challenges of Functional Outcomes in Traumatic Brain Injury Research: Timing of Follow-Up, Prognostic Models, and Missing Data

Abstract

Background: Traumatic brain injury (TBI) is common and debilitating. Randomized trials of interventions for TBI usually assess effectiveness by using long-term functional neurological outcomes but this is costly and difficult. If patient characteristics available at hospital discharge are predictive of 6-month functional outcome, then shorter-term outcomes may be adequate for use in future clinical trials. We evaluated models to predict long-term outcomes after TBI from short-term functional measures and easily obtainable demographic and injury characteristics as covariates, using data from a previously published randomized clinical trial. Methods: The Hypertonic Saline TBI trial of the Resuscitation Outcomes Consortium (ROC) enrolled 1282 TBI patients but had 15% missing data for the primary outcome of 6-month Glasgow Outcome Score Extended (GOSE). We evaluated patterns of missing data, whether functional outcome obtained earlier than six months would adequately reflect 6-month GOSE, and three prognostic models that predict 6-month severe disability (GOSE ≤ 4) via logistic regression using covariates and outcomes at discharge. Results: Patients with missing 6-month GOSE had less severe injuries, higher neurological function at discharge (GOSE), and shorter hospital stays than patients whose GOSE was obtained. Of 1066 (83%) patients with available discharge and 6-month outcomes, 71.2% of patients had the same functional status (severe disability/death vs. moderate /no disability) after 6 months as at discharge, 28% had an improved functional status, and 1% had worsened. Performance was excellent (AUC between 0.88 and 0.91) for all three prognostic models and calibration adequate for two models (p-values 0.22 and 0.85). Conclusion: Missing data was more common in healthier patients suggesting an ascertainment bias if the missing data were ignored during analysis. Shorter duration follow-up appears inadequate in representing long-term functional neurological outcome following TBI; however, all three prognostic models were highly predictive of long-term outcome. Our results support the more widespread use of multiple imputation of the standard 6-month GOSE when the primary outcome cannot be obtained through other means.