You are visiting us from Virginia. You are located in HHS Region 3. Your Center is Central East ATTC.

Smartwatch-Type Sensors Accurately Detect Cocaine Use

January 7, 2021
Meg Brunner, MLIS
Ertin E, et al. An examination of the feasibility of detecting cocaine use using smartwatches. Frontiers in Psychiatry 2021 (in press).
Return to ASME Catalog

Addiction Science Made Easy
July 2021
CTN Dissemination Library


Digital health technologies like smartwatches are changing the way research is conducted. They allow for better and more accurate data collection in natural settings, without disruption to participants’ normal daily lives.

Hands Touching Watch on WristCocaine use in clinical trials is usually measured via self-report and/or urine drug screens. Urine drug screens in particular are considered the “gold standard” of use detection for many substances but are intrusive and time-consuming. Researchers have also used ECG sensors to detect cocaine use, because cocaine use typically causes measurable heart rate elevation. However, these sensors require the use of a chest band that participants have not found convenient or comfortable.

Because of that, researchers in the NIDA Clinical Trials Network wanted to test out the use of smartwatches as a tool to collect information about cocaine use in clinical trials. The study, CTN-0073-Ot, measured the amount and quality of data yielded by a wrist-worn measurement device, as well as whether participants found it comfortable and less obtrusive than the chest-based devices worn in previous studies. 

Results of the study found that the quality of data collected was better for the wrist sensors compared to the chest sensors. Participants also reported that the wrist sensors were easier to put on, more comfortable, less self-conscious to wear, and interfered less with daily activities. 

The study also revealed some challenges with the devices that could be addressed in future models, including the fact that different wear styles can impact contact of the sensor with the skin, participant skin color can interfere with certain LED sensors, and participants can’t tell when they haven’t put the devices on correctly, so they don’ t know when they need to adjust the wristband to improve data quality.

Conclusions: This study of the use of a smartwatch-style tool for collecting real-time, natural setting cocaine use found the devices to be as effective as chest-based sensors and more comfortable for participants to wear. Based on these promising results, the next step would be to conduct a study with a larger sample size to increase the precision of cocaine use detection.

Find it in the CTN Dissemination Library