William C Walker1, Scott D McDonald. 1. Hunter Holmes McGuire Veterans Affairs Medical Center , Richmond, VA, USA. wwalker@mcvh-vcu.edu
Abstract
OBJECTIVE: To determine whether combining basic neurologic examination measurements with posttraumatic amnesia (PTA) duration can enhance prediction accuracy in moderate-to-severe traumatic brain injury (TBI). DESIGN: A descriptive between-subjects study. SETTING: Multicenter: 4 Veterans Affairs Brain Injury Rehabilitation Centers within the Defense and Veterans Brain Injury Center network. PARTICIPANTS: A total of 210 patients (active duty, veteran, or military dependent) with moderate-to-severe nonpenetrating TBI who consented during acute rehabilitation for data collection and completed relevant baseline and 12-month follow-up evaluations. METHODS: Multivariate analysis on neurologic examination predictor variables (upper extremity paresis, lower extremity paresis, Romberg test, and Kurtzke Expanded Disability Status Scale [EDSS]) was performed with block logistic regression modeling that controlled for duration of PTA. MAIN OUTCOME MEASURE: Glasgow Outcome Score at 1 year. RESULTS: Glasgow Outcome Score prediction accuracy by using PTA duration was modestly improved with the included neurologic examination measurements. The most influential predictor among them was EDSS, a composite measurement of neurologic impairment. When EDSS was excluded, the simple measurements of upper limb paresis and the Romberg test also provided independent prognostic value. CONCLUSIONS: This study demonstrated that simple clinical measurements of neurologic impairment (limb paresis, imbalance, other neurologic deficits) are of value in refining the prediction of long-term global outcome from TBI. These measurements may serve as markers of focal injury not captured by PTA duration, a severity marker weighted toward diffuse injury. Copyright Â
OBJECTIVE: To determine whether combining basic neurologic examination measurements with posttraumatic amnesia (PTA) duration can enhance prediction accuracy in moderate-to-severe traumatic brain injury (TBI). DESIGN: A descriptive between-subjects study. SETTING: Multicenter: 4 Veterans Affairs Brain Injury Rehabilitation Centers within the Defense and Veterans Brain Injury Center network. PARTICIPANTS: A total of 210 patients (active duty, veteran, or military dependent) with moderate-to-severe nonpenetrating TBI who consented during acute rehabilitation for data collection and completed relevant baseline and 12-month follow-up evaluations. METHODS: Multivariate analysis on neurologic examination predictor variables (upper extremity paresis, lower extremity paresis, Romberg test, and Kurtzke Expanded Disability Status Scale [EDSS]) was performed with block logistic regression modeling that controlled for duration of PTA. MAIN OUTCOME MEASURE: Glasgow Outcome Score at 1 year. RESULTS: Glasgow Outcome Score prediction accuracy by using PTA duration was modestly improved with the included neurologic examination measurements. The most influential predictor among them was EDSS, a composite measurement of neurologic impairment. When EDSS was excluded, the simple measurements of upper limb paresis and the Romberg test also provided independent prognostic value. CONCLUSIONS: This study demonstrated that simple clinical measurements of neurologic impairment (limb paresis, imbalance, other neurologic deficits) are of value in refining the prediction of long-term global outcome from TBI. These measurements may serve as markers of focal injury not captured by PTA duration, a severity marker weighted toward diffuse injury. Copyright Â