Effect of blood alcohol level on Glasgow Coma Scale scores following traumatic brain injury

Abstract

Objective: It is a common clinical perception that alcohol intoxication systematically lowers Glasgow Coma Scale (GCS) scores when evaluating traumatic brain injury (TBI). However, the research findings in this area do not uniformly support this notion. The purpose of this study is to examine the effects of blood alcohol level (BAL) on GCS scores following TBI.

Method: Participants were 475 patients (64% male) who presented to a Level 1 trauma centre following a TBI. Patients were selected if they were injured in a motor vehicle accident and had an available day-of-injury GCS, BAL and Computed Tomography (CT) brain scan.

Results: Overall, acute alcohol intoxication did not significantly affect GCS scores, even in patients with BALs of 200 mg dl⁻¹ or higher. When controlling for the effects of injury severity, acute alcohol intoxication affected GCS scores only in those patients with BALs greater than 200 mg dl⁻¹ who also had intracranial abnormalities detected on CT scan.

Conclusions: These findings suggest that GCS scores can be interpreted at face value in the vast majority of patients who are intoxicated. However, GCS scores will likely over-estimate the severity of brain injury in patients with abnormal head CT scans and BALs greater than 200 mg dl⁻¹.

• Traumatic brain injury
• alcohol
• Glasgow Coma Scale score

Notes

†Two supplementary correlation analyses were undertaken to determine whether the high prevalence of patients with GCS scores of 15 (61.9% of the sample) and/or BALs of 0 mg dl⁻¹ (48.6% of the sample) influenced these results. First, patients were excluded if their GCS was 15. The Pearson correlation between GCS and BAL in this group was r = −0.01 (p = 0.905, n = 181). Second, patients were further excluded if their BAL was 0 mg dl⁻¹. The Pearson correlation between GCS and BAL in this group was r = 0.03 (p = 0.759, n = 93). Due to the lack of relationship between BAL and GCS, no further analyses were undertaken using these two sub-groups.

†Small sample sizes precluded the use of Chi-square analyses to examine the prevalence of patients stratified by the four BAL groups and four GCS groups (as previously presented in Table II) in the CT Normal and CT Abnormal groups separately.

†These analyses were considered exploratory because alcohol tolerance was not measured per se. Rather, pre-injury alcohol use was used as a proxy for this variable. Alcohol use history is one of many factors that may contribute to a person's tolerance for alcohol. For example, other contributing factors may also include, but are not limited to, a person's body type, food consumption, digestion, illness and years of drinking. These additional variables were not available and were not considered in these analyses. In addition, sample sizes were too small for many groups to enable meaningful comparisons.