A practical method for dealing with missing Glasgow Coma Scale verbal component scores

Area of Science:

• Neuroscience
• Clinical Neurology
• Medical Informatics

Background:

• The Glasgow Coma Scale (GCS) is crucial for assessing impaired consciousness.
• Missing verbal component scores can hinder GCS sum score derivation and prognostic modeling.
• Accurate GCS scores are vital for patient outcome prediction.

Purpose of the Study:

• To develop and assess a simple imputation tool for missing GCS verbal component scores.
• To evaluate the utility of this tool within prognostic models like GCS-Pupils plus age plus CT findings (GCS-PA CT).

Main Methods:

• Analysis of the International Mission for Prognosis and Analysis of Clinical Trials in Traumatic Brain Injury (IMPACT) cohort to determine missing verbal score frequency.
• Development of a single verbal score imputation method based on eye and motor combined (EM) scores.
• Validation using IMPACT and Corticosteroid Randomisation After Significant Head Injury (CRASH) databases, comparing model performance with and without imputed scores.

Main Results:

• Verbal scores were most frequently missing in patients with no eye opening and low motor scores (≤4).
• The developed simple imputation model performed comparably to more complex methods.
• Imputing verbal scores minimally impacted prognostic model accuracy (R2 reduction from 32.1% to 31.4% for death prediction).

Conclusions:

• The developed imputation strategy offers a practical solution for obtaining GCS sum scores when verbal assessment is not possible.
• This facilitates clinical communication, decision-making, and prognosis estimation in head injury patients.
• External validation in diverse clinical populations is recommended for the imputation strategy and GCS-PA charts.