Clinician assessment of blood alcohol levels among emergency department patients

Introduction: Alcohol intoxication is a significant public health concern and is commonly seen among emergency department (ED) patients. This study was undertaken to identify the accuracy of clinician assessment of blood alcohol levels among emergency department patients.

Methods: This prospective survey study was conducted at a Level 1 Trauma Center. Eligible study participants included physicians, nurses, and medical students involved in the care of patients who had BAC. Clinicians estimated the BAC prior to results availability.

Results: Among 243 clinicians, the mean difference between the estimated BAC and actual BAC was 17.4 (95% CI: 4.7 to 30.1). Providers tended to overestimate the actual BAC level. The accuracy between roles (attendings, residents, RNs, students) was not significant (ANOVA p-value 0.90). Accuracy was not correlated with age of the patient (Pearson correlation 0.04, p-value 0.54). Accuracy was not associated with the patient's gender (Student's t-test two-tailed p-value 0.90), ethnicity (White versus all others, t-test p-value 0.31), nor insurance (government versus not government, t-test p-value 0.81). The average accuracy value was associated with mode of arrival (t-test p-value 0.003). The average accuracy for walk-in subjects was -14.9 (CI: -32.8 to 3.1) compared to ambulance arrivals 28.3 (CI: 12.7 to 44.0). Providers underestimated BAC for walk-ins and overestimated BAC for ambulance arrivals. Among 107 patients with a BAC of 0, clinician estimates ranged from 0 to 350. Clinicians estimated non-zero BAC levels in 17% of patients with BAC of 0 (N = 18).

Conclusions: Clinicians' estimates of BAC were often inaccurate, and often overestimated the BAC.