Simon Goldberg, an assistant professor with the School of Education’s Department of Counseling Psychology, appeared in an article in STAT news, which delivers journalism about health, medicine, and the life sciences.
The story, headlined “How Mindstrong’s rush to roll out a ‘smoke alarm’ for mental illness led to its downfall,” discusses the recent failure of the mental health company Mindstrong, and its efforts to use technological data to expedite the mental health diagnoses process.
According to STAT, Mindstrong intended to “develop a ‘biomarker’ that could passively analyze users’ typing speed, typos, and tapping and scrolling patterns for early signs of cognition and mood changes indicating conditions like depression. If effective, the system could help them get therapy or other treatment faster, and save insurers money on more intensive mental health care in the long run.”
The data used would be collected from users of virtual health programs, and a pilot program was launched in 2018 in Los Angeles County. However, despite initial expectation, health workers soon “expressed concern that Mindstrong’s predictive technology didn’t work,” according to STAT.
Mindstrong was also criticized for a lack of transparency with patients about their data collection.
Reflecting on what this suggests for the future of using technology to collect health data, Goldberg said, “I think the lesson is probably more around how tech companies are run, more than anything about the sort of potential in this area.”
“I think it can work, and I think it’s probably a question of when — and not overpromising and not putting the funding before the science,” added Goldberg, who is also an affiliate faculty member with UW–Madison’s Center for Healthy Minds.
“On the one hand, it seems to me that if anyone’s going to crack these hard-to-crack nuts quickly, it’ll be in the private sector because they can raise money faster and they just have more engineers at their disposal,” Goldberg said. “In the academic world, things happen slower, (but) that might ultimately produce science that’s more trustworthy and ultimately more effective.”
To learn more, read the full article at statnews.com.