5 Questions: Kit Delgado on Breathalyzers and smartphones

hs1interview25z-b
DrinkMate is one of numerous breathalyzers on the market that work with users’ smart phones to determine blood alcohol levels. University of Pennsylvania researchers plan to test its accuracy in a study of several apps.

On each of his shifts as an emergency room physician at Penn Presbyterian Medical Center, Kit Delgado sees the tragic consequences of people making shortsighted decisions.

So for the last several years, Delgado, an assistant professor of emergency medicine and epidemiology, has turned his research efforts toward preventing the injuries that bring people to trauma centers such as his.

He’s been focusing on cellphone use while driving, and – especially pertinent to the holiday party season – drinking and driving.

According to the National Highway Traffic Safety Administration, more than 10,000 people were killed in 2015 in vehicle crashes involving drivers impaired by alcohol. That was up 3.2 percent from the previous year and accounted for nearly a third of all traffic-related deaths.

Dr. Delgado spoke to us recently about what he sees as “a huge amount of potential to intervene using devices that are in people’s pockets every day” — their smartphone, linked to a personal Breathalyzer. 

Can you tell us about these devices?
Personal Breathalyzers have been on the market for more than 20 years. Basically, they are a miniaturized version of the same technology that police officers can use on the side of the road. What’s different — and it is creating a whole new world of opportunity — is that some of these devices now have their data linked to your smartphone.

There are several available for iPhone and Android, including: Alcohoot, BACtrack, DrinkMate, DriveSafe, and Floome. They retail online from $35 to $140. You get the device, then download an app to your smartphone. You turn on the app and the device, blow into the device, and your breath alcohol concentration, or BAC, which estimates your blood-alcohol level, appears on your phone.

The advantage is that you can quantify and track your blood-alcohol levels over time. You can also share those readings. For example, suppose you’re a parent and you want to know how your teen is doing. BACtrack has an app called BACtrack View that the parent can download. The teen will get a notification on the phone to submit a breath-alcohol measurement. The teen turns on the app and blows into the device while holding the phone, as if taking a selfie. The device will record the breath alcohol and also will take a picture of your face — so you’re not giving it to a friend to game the system. It also gives a time stamp and your location.

But I gather the main use for now is personal monitoring?
Yes, to take out the guesswork on figuring out how intoxicated you are. While you should always plan to have a designated driver, people are using personal Breathalyzers to check whether they are too impaired to drive. Although the national DUI limit is a blood-alcohol level of 0.08, driving is impaired at levels of 0.04 and above. Many Breathalyzer apps provide an estimate of when your blood-alcohol level is expected to decline to below this point and prompt alternatives to driving such as calling rideshare service like Uber or Lyft, and many also show a map of nearby food options.

The Colorado Department of Transportation distributed smartphone Breathalyzers to 225 randomly chosen bar patrons. The average breath-alcohol level from the 4,823 measurements taken was 0.087 (above the 0.08 legal limit); 84 percent of participants agreed owning a smartphone Breathalyzer lowered their risk of a DUI.

How accurate are the smartphone Breathalyzers?
When they first came out, I bought the devices and tested them myself. At least one of them, Breathometer, had a much lower measure than the others. That was disconcerting to me, from a public health standpoint. It has since been pulled off the market. Many of the devices on the market are not Food and Drug Administration-approved, and thus have not submitted or published data on accuracy.

That’s what motivated me to do my first study in this area. We are bringing people into the hospital in a lab setting. After a rigorous screening process, we provide them with a weight-based dose of alcohol to get them just above the legal limit, and compare the accuracy of the devices on the market to a police-grade device. To my knowledge, this study, supported by the Penn Injury Science Center, will be the first objective, rigorous assessment of the comparative accuracy of smartphone Breathalyzers. For right now, I would not trust devices that have not received FDA approval.

Based on your research, what do you see as the potential for smartphone Breathalyzers?
Like exercise trackers and other connected devices, smartphone-paired Breathalyzers can act as a facilitator of healthy behavior change. However, without a behavioral strategy to promote continued engagement aimed at promoting better health or safety, people will stop using these devices over time. I’m interested in testing behavioral strategies that leverage smartphone Breathalyzers with the ultimate goal of reducing drunk-driving crashes. 

In the spring, we’ll start a study with seed money from Penn’s Center for Health Incentives and Behavioral Economics. We will pilot test behavioral strategies that use smartphone Breathalyzers, text message reminders, rideshare app credits, and behavioral incentives for reducing alcohol-impaired driving and excessive drinking. We will be recruiting bar patrons in the Philadelphia suburbs who admit to drinking and driving in the past. We also just received additional funds to conduct focus groups with college students and college administrators. That will enable us to develop a promising behavioral intervention for reducing heavy binge drinking on college campuses. 

What next?
There are some limitations to personal Breathalyzers. It’s just one more thing to carry around. Also, if you want to get an accurate measure of your breath alcohol, you have to wait at least 15 minutes after your last drink. If you take a reading just after drinking a shot of tequila, you’re going to get an inaccurately high reading. It can take about 20 to 30 seconds for the devices to warm up. There also might be a stigma with using these devices since you don’t see them every day. If someone pulls one out at a bar, it makes people wonder, "Is this someone with a drinking problem or who is in a treatment program?"

But there is already a lot of innovation. For example, there are prototypes in development for wearable wristband devices that will accomplish some of the same goals. Basically, they get an estimate of your blood alcohol through a sensor on your skin, typically through your sweat. Imagine you have a bracelet that does this continuously and sends it to your phone. It’s much more discreet and will likely be on the market in the next two years. You can imagine that as the technology gets better, these can be integrated into smartwatches.

Where things are going to be headed is in improving the behavioral strategies behind these apps, to use these devices to get people to not drink, or drink more moderately, and not drink and drive. I think there is a lot of potential for better using the shareable data from these devices to reward responsible behavior and leverage friends and family to encourage more responsible drinking, or recognize when someone may be getting into trouble. All that functionality is out there. It will help these devices to not just be a novelty, but to make a difference and reduce drunk-driving crashes.

sandybauers10@gmail.com