A new generation of wearable alcohol sensor has potential for use as a self-management tool by social drinkers, as reported in Alcoholism: Clinical and Experimental Research. However, further development will be needed before the devices are suitable for widespread use similar to other health and fitness biosensors. To date, transdermal alcohol content (TAC) sensors − which measure ethanol excreted through the skin − have been used mainly within the criminal justice system for abstinence monitoring. However, there has been increasing interest and research into their potential use with other drinking populations to motivate and support alcohol self-management. For this purpose, alcohol biosensors must be accurate, able to detect alcohol with minimal delay after drinking, and feasible and acceptable to wear.

Traditional (older-generation) TAC sensors are worn on the ankle. However, these sensors are bulky, and their noise and vibration can disrupt daily activities including exercise and sleep. A further limitation is that data upload occurs only when close to a base station. The latest TAC sensors, worn on the wrist, have better implementation potential but have not yet been rigorously tested under real-world conditions. To address this gap, researchers from Yale conducted a field-based study with heavy drinking young adults to assess the accuracy and acceptability of a new generation sensor − known as BACTrack Skyn − compared with a traditional ankle-worn device.

The study included 47 participants aged 18 to 25, who were instructed to wear both the Skyn and a traditional sensor for 1 to 6 days, and to drink as they usually would. Acceptability of Skyn was excellent compared to the traditional device, and also when compared to keeping a mobile drinking diary as an alternative measurement tool. Compared with the traditional sensor, the Skyn device was associated with lower disruption of sleep, exercise and social comfort. Device accuracy (i.e. ability to reliably detect drinking) for Skyn was at least adequate (and for some measures, excellent). However, the researchers were unable to derive automated rules for alcohol detection that worked consistently across different contexts. For example, the level of TAC corresponding to drinking was different among participants that enrolled during the first versus second half of the study.

Overall, the findings suggest that, in time, new-generation alcohol biosensors could enter the behavioral self-management space for use by social drinkers. However, enhanced hardware and analytics are needed to autodetect drinking more consistently.