External Validation of a Prognostic Model for Outcome After Mild Traumatic Brain Injury at 6 Months Post Injury.

Area of Science:

• Neuroscience
• Clinical Medicine
• Public Health

Background:

• Mild traumatic brain injury (mTBI) often leads to persistent symptoms in a minority of patients.
• Early identification of at-risk individuals is crucial for timely interventions.
• The UPFRONT-model was developed to predict poor functional outcomes after mTBI.

Purpose of the Study:

• To externally validate the predictive performance of the UPFRONT-model in an independent cohort of mTBI patients.
• To assess the model's calibration and discrimination in a new population.
• To determine the model's suitability for clinical implementation in identifying patients needing targeted interventions.

Main Methods:

• A prospective, longitudinal, multicenter cohort study involving 126 mTBI patients.
• Data collected included demographics, injury characteristics (e.g., Glasgow Coma Scale, Post Traumatic Amnesia), and pre-existing psychological factors.
• Functional recovery was assessed at 6 months using the Glasgow Outcome Scale Extended (GOS-E).

Main Results:

• The UPFRONT-model demonstrated acceptable discriminative ability (AUC=0.74), similar to its development cohort.
• Calibration analysis revealed a systematic underestimation of recovery (predicted 46% vs. observed 75%).
• Psychological variables consistently proved to be robust predictors across both samples.

Conclusions:

• The UPFRONT-model exhibits good discriminative power in an external validation cohort.
• The model's tendency to underestimate recovery necessitates further refinement before widespread clinical adoption.
• The UPFRONT-model shows potential for early risk stratification in mTBI patients, guiding personalized treatment strategies.