Are there better ways to evaluate how prescribed therapies are working while patients are living their everyday lives, when doctors aren’t there to check? New answers may come from the digital devices we carry and wear, says Donna Coffman, a biostatistician in the College of Public Health.
Coffman is leading a National Cancer Institute-funded study of smoking cessation treatments as a way to develop models that could help researchers improve patient responses to all sorts of ongoing therapies, from managing addictive behaviors to compliance with medication-taking.
“We now have all these devices, from mobile phones to wearables, and they can collect a lot of data,” she says. “In the old days, you’d say to a smoker who’s trying to quit, ‘how have your withdrawal symptoms been over the last six months?’ Now you can send little surveys multiple times per day. That allows us to look at very fine grain associations between, for example, withdrawal symptoms and relapse.”
It’s called ecological momentary assessment (EMA), which—put more simply—is about measuring how people are doing (the assessment) at points in time (the momentary) while they are out in the world (the ecological). By providing new, real-time glimpses of patient behavior and feelings, EMA can help researchers and clinicians understand how medicine and other treatments work over the course of time. That can foster everything from development of new therapies to individually tailored modifications of existing treatments. It can enable real-time interventions as well.
“If you knew there was a moment when someone was likely to relapse, you could send an intervention through their phone,” Coffman says.
It falls on biostatisticians like Coffman to create mathematical models that will help researchers work with this new flood of data, using it to establish meaningful associations and develop their responses.
In her smoking cessation study, patients trying to quit were given either nicotine patches, the medication varenicline (brand name Chantix), or a combination therapy including the patch plus optional nicotine lozenges. Each participant was sent a mini-questionnaire at three different moments during the day, asking them to rate the level of their cravings and mood and indicate whether they had smoked or been exposed to triggers like drinking alcohol.
The frequent digital check-ins enable analysis of not only each treatment’s effectiveness but also how its effectiveness changed over time.
“Is the effect of the nicotine patch on withdrawal symptoms stronger shortly after quitting, and then diminishes as time goes on? Does it change over the course of the day?,” Coffman asks. “If you know effectiveness of something diminishes, you can more easily decide where you want to come back in, and maybe give an additional intervention or do something to boost the effectiveness.”
Other wearable devices such as heart rate monitors and pedometers could increasingly become part of real-time data gathering in consenting patients to make health interventions more timely and effective.
As part of the grant, Coffman also is creating software that can help researchers use data gathered in various kinds of EMA studies. “I really want people to use the methods I’ve developed,” she says. “These models are new. Before, there were no ways to answer these questions.”