Frequent smartphone check-ins predicted 75% of suicide attempts and 87% of suicide-related events like hospitalization in the week before they occurred, according to a Sept. 11 Boston Globe report.
Researchers from Cambridge, Mass.-based Harvard followed 619 adults and adolescents who had sought care for suicidal thoughts or behavior at two Boston-area hospitals. On a sliding scale from 1 to 10, participants were asked how hopeless and agitated they felt, how strong their urge to kill themselves was and how capable they felt of resisting that urge. For three months, participants received 20-question surveys on their smartphones six times a day. Researchers used the responses and information about survey completion in machine-learning models to predict suicide attempts or other suicide-related crises in the following seven days.
The study generated nearly 80,000 surveys. Every one-point increase among adults in agitation on a 0-to-10 scale was associated with an 11% increase in the odds of a suicide attempt in the following week.
“We were looking for common signals like skin conductance, heart rate variability, accelerometer data, which you can get from a cellphone or from a wrist sensor, to try and measure things like how active a person is, how they are sleeping, how much they are moving around during the day,” Harvard psychology professor Matthew Nock, PhD, a researcher for the study, told the Globe. “While depression is an important risk factor, agitation also seems to be a really strong predictor of near-term suicide attempt risk.”
Patient participation presented a hurdle. Participants were paid $1 for each completed survey, but fewer than half of the surveys sent were opened and participation declined over the study period. Despite the participation challenges, researchers are continuing to explore how real-time data could help forecast changes in a person’s mental state.
“We open our phone and we see, is it going to rain where I am in the next day, in the next hour,” he said. “The more data and the more dynamic data we can get, the more accurate our prediction, our forecast. We want to do the same thing for a person’s mental state.”
The study will be published in the October issue of the Journal of Psychopathology and Clinical Science.
At the Becker's Fall Behavioral Health Summit, taking place November 4–5 in Chicago, behavioral health leaders and executives will explore strategies for expanding access to care, integrating services, addressing workforce challenges and leveraging innovation to improve outcomes across the behavioral health continuum. Apply for complimentary registration now.
