Can We Actually Predict Long-Term Patient Satisfaction After Traumatic Brain Injury?

Area of Science:

• Neuroscience
• Clinical Medicine
• Public Health

Background:

• The Traumatic Brain Injury - Patient Reported Outcome (TBI-PRO) model was developed to predict long-term patient satisfaction using Quality of Life After Brain Injury (QOLIBRI) scores.
• Previous research established the TBI-PRO model for predicting patient-reported outcomes post-TBI.

Purpose of the Study:

• To externally and prospectively validate the TBI-PRO model in predicting long-term patient-reported outcomes.
• To develop a new predictive model using a larger dataset specifically for older adults (≥50 years) with TBI.

Main Methods:

• Prospective cohort study of TBI patients followed for 1 year, comparing TBI-PRO predictions with actual QOLIBRI scores at 3, 6, and 12 months.
• Development of a new model using Bayesian additive regression trees on combined datasets (European NeuroTrauma Effectiveness Research in TBI and Transforming Research and Clinical Knowledge in Traumatic Brain Injury) for older adults.

Main Results:

• Prospective validation showed limited predictive performance for the TBI-PRO model (AUCs ranging from 0.57 to 0.65).
• The novel model for older adults identified key predictors including medical history, hospital stay, blood pressure, age, pupil reactivity, and craniectomy.
• The coefficient of determination for the new model was low (0.19-0.28), indicating poor fitness for predicting long-term quality of life.

Conclusions:

• The TBI-PRO model's prospective validation failed to accurately predict long-term outcomes in TBI patients.
• A newly derived model for older adults also demonstrated poor accuracy in predicting long-term health-related quality of life.
• Current targeted approaches in TBI care have limitations; future research requires more targeted datasets for accurate long-term outcome prediction.