Michael G. McDonell, Jordan Skalisky, Emily Leickly, Sterling McPherson, Samuel Battalio, Jenny R. Nepom, Debra Srebnik, John Roll, Richard K. Ries

This study investigated which ethyl glucuronide immunoassay (EtG-I) cutoff best detects heavy versus light drinking over five days in alcohol dependent outpatients. A total of 121 adults with alcohol use disorders and co-occurring psychiatric disorders took part in an alcohol treatment study. Participants provided self-reported drinking data and urine samples three times per week for 16-weeks (total samples=2761). Agreement between low (100ng/mL, 200ng/mL), and moderate (500ng/mL) EtG-I cutoffs and light (women ≤3 standard drinks, men ≤4 standard drinks) and heavy drinking (women >3, men >4 standard drinks) were calculated over one to five days. The 100ng/mL cutoff detected >76% of light drinking for two days, and 66% at five days. The 100ng/mL cutoff detected 84% (1 day) to 79% (5 days) of heavy drinking. The 200ng/mL cutoff detected >55% of light drinking across five days and >66% of heavy drinking across five days. A 500ng/mL cutoff identified 68% of light drinking and 78% of heavy drinking for one day, with detection of light (2-5 days <58%) and heavy drinking (2-5 days <71%) decreasing thereafter. Relative to 100ng/mL, the 200ng/mL and 500ng/mL cutoffs were less likely to result in false positives. An EtG-I cutoff of 100ng/mL is most likely to detect heavy drinking for up to five days and any drinking during the previous two days. Cutoffs of ≥500ng/mL are likely to only detect heavy drinking during the previous day.