Study: Passive sensor data collected from smartphones and Fitbits linked to severity of chemotherapy symptoms

By Laura Lovett
Share

A small study published in the Journal of Medical Internet Research found that passive sensor data collected from smartphones and Fitbits could be linked to the severity of symptoms cancer patients experienced during their chemotherapy treatment. 

"We found that on days when the patients reported worse-than-average symptoms, they tended to spend more time being sedentary, moved the phone more slowly, and spent more minutes using apps on the phone,” Carissa Low, professor of medicine and psychology at the University of Pittsburgh, and lead author of the study, said in a statement. "Collecting these objective behavioral measures from smartphone sensors requires no additional effort from patients, and they could prove beneficial for long-term monitoring of those undergoing arduous cancer treatments or those with other chronic illnesses."

Symptoms such as fatigue and sleep disturbances are common among chemotherapy patients. Other symptoms like nausea and pain are also common but can fluctuate with each chemo cycle, according to researchers. The study set out to look for links between these symptoms and certain mobile and Fitbit activities. 

The study was made up of 14 patients undergoing chemotherapy for gastrointestinal cancer. Each of the participants carried an Android phone provided by the study and wore a Fitbit for the duration of the 4-week study. Every day patients rated the severity of 12 common chemotherapy symptoms. Severity ratings were totaled to create a symptom burden score for each day. The scores were centered on individual patients means and categorized into low, average and high symptom burden days, according to the study. 

Researchers looked at the sensor and symptom data and discovered a number of mobile and Fitbit features were correlated to patient-reported symptom burden scores, according to the study. Using this algorithm researchers found an 88 percent accuracy between patient-reported symptom burdens and the correlated mobile and Fitbit features.

The number of sedentary bouts and time spent on the smartphone strongly correlated to higher symptom severity. 

Researchers also examined the performance models they built with data from each device. They found that the phone activity had a higher correlation to patients’ symptom burdens than Fitbit recorded activity. 

“Results of device-specific feature selection indicate that features from mobile phone sensors were more valuable in symptom estimation than Fitbit features,” authors in the study wrote. “In particular, features related to mobility and activity and phone usage patterns yielded the most accurate models. This suggests that future passive sensing research focused on symptoms could consider relying only on the features derived from the phone accelerometer and GPS as well as information about duration of phone and app usage and battery charges.”

The study was only four weeks long, which was a limitation, according to the authors. It was also limited by the fact that participants used study-specific mobile phones and the data many not have reflected personal mobile phone use patterns, according to researchers. 

Currently the researchers are running follow-up studies to determine whether the same passive sensing approach can be used to identify complications following cancer surgery.