Background: Two types of transdermal electrochemical sensors that detect alcohol at the skin surface were evaluated. One, the AMS SCRAM™ device, is locked onto the ankle and is based on a fuel cell sensor; the other, a Giner WrisTAS™ device, worn on the wrist, is based on a proton exchange membrane. SCRAM is used by several court systems in the United States to monitor alcohol offenders, WrisTAS, a research prototype, is not commercially available.
Methods: The 2 devices were worn concurrently by 22 paid research subjects (15 men, 7 women), for a combined total of 96 weeks. Subjects participated in both laboratory-dosed drinking to a target of 0.08 g/dl blood alcohol concentration (BAC), and normal drinking on their own; all subjects were trained to use and carry a portable fuel-cell breath tester for BAC determinations. Overall 271 drinking episodes with BAC ≥ 0.02 g/dl formed the signal for detection—60 from laboratory dosing, and 211 from self-dosed drinking, with BAC ranging from 0.02 to 0.230 g/dl (mean 0.077 g/dl).
Results: False negatives were defined as a transdermal alcohol concentration response equivalent <0.02 g/dl when BAC ≥ 0.02 g/dl. The overall true-positive hit rate detected by WrisTAS was 24%. The low detection rate was due to erratic output and not recording during nearly 67% of all episodes; reportedly a chipset, not a sensor problem. SCRAM correctly detected 57% across all BAC events, with another 22% (total 79%) detected, but as <0.02 g/dl. When subjects dosed themselves to BAC ≥ 0.08 g/dl, SCRAM correctly detected 88% of these events. SCRAM devices lost accuracy over time likely due to water accumulation in the sensor housing. Neither unit had false-positive problems when true BAC was <0.02 g/dl.
Conclusions: Each device had peculiarities that reduced performance, but both types are able to detect alcohol at the skin surface. With product improvements, transdermal sensing may become a valuable way to monitor the alcohol consumption of those who should be abstaining.