Urine Ethyl Glucuronide Unraveling the Reality of Abstinence Monitoring in a Routine Outpatient Setting: A Cross-Sectional Comparison with Ethanol, Self Report and Clinical Judgment

Abstract

Aims: To test the screening performance of urinary ethyl glucuronide (EtG) under routine clinical conditions in a sample of alcohol-dependent outpatients, comparing it against urinary ethanol, self reports and clinical judgment. Methods: A cross-sectional study under routine conditions was conducted in February 2015, where 613 consecutive urinary samples, provided by 188 outpatients with alcohol use disorders, were analyzed for ethanol and EtG (cut-off level = 500 ng/ml). Clinical variables such as the presence of aversive medication, comorbidities and clinician judgment were also collected. The discrepancy between the number of alcohol and EtG positives was recorded. A logistic regression analysis including clinical variables was conducted to assess for predictors of EtG positivity. Results: Urinary alcohol yielded 9 positives (1.5% of all urine samples) belonging to 8 patients. EtG yielded 136 positives (22% of all urine samples) belonging to 74 patients. Of these, 93.4% (127 of 136) were negative for alcohol. All urinary alcohol positives resulted in EtG positives. The clinician judged 48 samples from 26 patients as belonging to not abstinent patients and 550 samples from 178 patients as belonging to abstinent patients. She was unsure in 15 samples from 15 patients. When comparing it against EtG as the gold standard, the area under the curve was 0.592. Self reports were extremely unreliable in this study, with only 5 patients reporting drinking in a total of 6 urine samples. In the logistic regression model, only aversive medications (OR 2.1, 95% CI 1.3-3.3) and clinician judgment (OR 2, 95% CI 1.4-2.9) resulted in significant effects. Conclusions: EtG performed largely better than ethanol for urine screening in alcohol outpatients, detecting an extra 20.4% (125 out of 613) of positives. It means that for each alcohol-positive sample, there were 15 EtG-positive samples. Although better than ethanol, clinician judgment was also not performed efficiently. If routinely implemented in the screening of alcohol outpatients, EtG might bring relevant changes that merit further research.