Smartphone Breath Alcohol Testing Devices Vary Widely in Accuracy

Newswise — Alcohol breath testing devices that pair with smartphones are marketed as safety tools for general use, but their accuracy is highly variable, a new laboratory study shows. While some of these devices potentially help people avoid driving while impaired, others may mislead users into thinking falsely that they are fit to drive. Alcohol-impaired driving kills 29 people a day and costs $121 billion a year in the US, amplifying interest in personal breath-testing devices marketed at consumers. Such devices paired with smartphone apps are widely available via online marketplaces though information about their origins and accuracy is scarce. The study in Alcoholism: Clinical & Experimental Research compares the accuracy of six such devices with that of two validated alcohol-consumption tests.

Researchers worked with 20 moderate drinkers aged 21–39. The participants were given three doses of vodka over 70 minutes with the goal of reaching a peak blood alcohol concentration (BAC) of around 0.1%. After each dose, participants’ breath alcohol concentration (BrAC) was measured using smartphone-paired devices and a police-grade handheld device, a validated means of estimating BAC. After the third dose, their blood was drawn and tested for BAC, the most accurate way of measuring alcohol consumption. Researchers also explored the devices’ ability to detect breath alcohol concentration above common legal driving limits (0.05% and 0.08%). They used statistical analysis to explore differences between the measurements.

The participants’ peak blood alcohol concentration ranged from 0.06% to 0.14%. All the breath-testing devices, including the police-grade device, underestimated BAC, consistent with previous research — in this study by a mean of more than 0.01%. The accuracy of smartphone-paired devices varied widely. The most accurate — the BACtrack Mobile Pro and the police-grade device — underestimated BAC by no more than 0.02%. Other devices yielded wider margins. Drinkmate and DRIVESAFE Evoc generated average estimates of 0.04% below peak BAC; BACtrack Vio and Floome differed significantly from the police-grade device at certain points in the study. The devices also varied in detecting driving-limit thresholds. BACtrack Mobile Pro and Alcohoot were the most sensitive, while Drinkmate and DRIVESAFE failed to detect BAC limit thresholds of 0.08% more than half the time.

The researchers concluded that such devices are potentially useful for remotely monitoring alcohol consumption and may help reduce risky driving behavior. The BACtrack Mobile Pro, for example, was suitable for personal, clinical, and research use. Other devices, however, dangerously underestimated BAC and frequently failed to detect risky breath-alcohol levels. The researchers recommend closer government regulation of such devices and research into their effects on users’ decisions  to drive. (Since the study, Drinkmate has been discontinued, and other models have been updated.)

Accuracy of consumer marketed smartphone-paired alcohol breath testing devices: A laboratory validation study. M. Delgado, F. Shofer, R. Weatherill, B. Curtis, J. Hemmons, E. Spencer, C. Branas, D. Wiebe, H. Kranzler. (p xxx)

ACER-20-4566.R2