Whether during the office break, drunk at a party, waiting at the bus stop or simply out of habit after a meal: Many people reach for a cigarette, especially when social pressure is high or a routine has established itself. Such trigger situations are especially challenging for people who want to quit smoking when they have to fight the inner urge. A new app could help them in this, which uses artificial intelligence (AI) to automatically recognize relevant situations and intervene to support them.
“We know that attempts to quit smoking often fail because the urge to smoke is triggered by being in places where people used to smoke,” says Felix Naughton, a health psychologist at the University of East Anglia in Britain. He and a team developed the “Quit Sense” smartphone app, which has been reviewed by independent experts The study was presented in the specialized journal “Nicotine & Tobacco Research”.
Quit Sense is an app that addresses the still relatively new psychological principle of Just-In-Time Adaptive Interventions (JITAI) make use of. This is what interventions are called which are intended to enable tailored support in trigger situations. This is possible because JITAIs access real-time data to draw conclusions about a person’s emotional, social, and physical condition. In combination with artificial intelligence, interventions can be even more targeted.
The AI learns the behavior of smoking
In the case of Quit Sense, users must first train the AI with their personal smoking behavior. Every time they light a cigarette, they have to make a note of it in the app, including clues about their mood (“stressed”) and context (“at work”). The app also uses GPS to record the respective location in the background. Within a few days, the AI learns which places and situations can lead users to smoke.
If users then declare in the app that they want to quit smoking, they are notified whenever a suspected trigger is imminent. The app then responds with support messages and contextual questions, intended to alert users to the situation without being intrusive or annoying.
To test the effectiveness of the procedure, the team conducted a controlled study with 209 smokers recruited via social media. All participants received a link to an NHS smoking cessation project. Half of the subjects also received access to the Quit Sense app.
Quit Sense, the AI-assisted smoking cessation app, recognizes when people enter a place where they used to smoke.
(Image: University of East Anglia)
Tailored intervention via app
After six weeks and after six months, the scientists asked the subjects to report on their experiences. Anyone who indicated they had quit smoking in the meantime was asked to provide a saliva sample to confirm their abstinence. While not everyone did, the result seems to speak for itself: four times as many people who accessed Quit Sense quit smoking than those who used NHS support alone. The tailored and timely intervention apparently worked.
For Felix Naughton, this is proof of how smartphone technology and machine learning can be used successfully for interventions in medicine and psychology: “By helping people who are trying to quit smoking to learn and deal with these situations, we can increase the success rate,” he says.
(jle)
