Instant Blood Pressure, an app from Aura Labs, has been removed from the app store for more than two years. The makers of the app, which purported to report users' blood pressure using the phone's camera, had to pay $600,000 settlement to the FTC after it was found to be inaccurate, first in an investigative report by iMedicalApps and then in a subsequent validation study at Johns Hopkins.
But the thoroughly pilloried app, the go-to example of digital snake oil, is back in the news today as researchers, including some of the same ones who worked on the 2016 validation study, have published a study in the Nature Partner Journal Digital Medicine attempting to ascertain why Instant Blood Pressure proved so popular while being so inaccurate. And the research suggests that the app's inaccuracy might actually have been part of its appeal.
"For mHealth apps, user experience may be at odds with accuracy of medical functionality," study authors, led by Dr. Timothy Plante, wrote. "For example, a theoretical app that estimates weight using a smartphone’s camera could improve enjoyment and promote a better user experience if it systematically underestimated a user’s weight and presented reassuring, lower-than-expected weight measurements to users. Such an app could foster greater app use and receive positive reviews despite its inaccuracy, thereby driving uptake of this app over other, more accurate apps. However, if users perceived this app as inaccurate, they may be less likely to use it."
To test the hypothesis, the team had users estimate their blood pressure, measure it with both the app and a standard of care sphygmometer, in a randomized order, and then rate the app. They were told their results from both devices afterward.
Researchers found that generally, when the app told someone their blood pressure was higher than they expected it to be, they were more likely to give it a lower rating. When it told them their blood pressure was lower than expected, they would give it a higher rating.
"For BP-measuring apps, the primary driver of use from a medical perspective should be accuracy," researchers wrote. "However, our exploration of user experience suggests that falsely reassuring BP results may play a role in promoting app use. In other words, inaccurate apps like IBP may be preferentially used because of a systematic underreporting of elevated BP results. Absent of formal validation, the app store ecosystem may drive use of inaccurate apps."
This could be an example of the self-serving bias, researchers theorized, a cognitive phenomenon that causes people to reject sources of inconvenient information. No matter the explanation, the study does fuel the widespread belief in the mobile health space that more vetting, validation, and curation of apps is needed.