Wildly Popular, Dangerously Inaccurate: Blood Pressure Monitoring App Misses Mark

The idea of using a popular smartphone app to measure blood pressure has huge appeal, but the readings they give are erroneous and possibly even dangerous for patients who are relying on them, a small study suggests. 

“It’s important to check blood pressure occasionally, but it’s also important that the numbers are accurate,” Timothy B. Plante, MD (Johns Hopkins University School of Medicine; Baltimore, MD), told TCTMD. 

Plante and colleagues decided to test out a best-selling blood pressure app (Instant Blood Pressure; AuraLife), which works by holding the top edge of a smartphone to the left side of your chest while simultaneously placing your right index finger over the camera of the phone.

Highly Inaccurate for Systolic and Diastolic

They compared readings taken with the app and those taken with an automated sphygmomanometer in 85 participants. More than half had self-reported hypertension, with 91% of those reporting use of antihypertensives.

Compared with blood pressure taken with the automated cuff, the app’s readings were off by a mean of 12.4 mm Hg for systolic pressure and 10.1 mm Hg for diastolic. Overall, the app underestimated higher pressures and overestimated lower pressures. Systolic readings were within 5, 10, and 15 mm Hg of the cuff reading 24%, 44%, and 59% of the time, respectively. Diastolic readings were within that range 26%, 48%, and 70% of the time, respectively. This correlated with the lowest possible accuracy grade in all categories by British Hypertensive Society scoring, and with sensitivity and specificity for hypertension of 0.22 and 0.92, respectively.

Plante presented the results of the study March 2, 2016 at the American Heart Association Epi/Lifestyle 2016 Scientific Sessions in Phoenix, AZ, and it was simultaneously published online in JAMA Internal Medicine.

“I was optimistic that [the app] would work, but I was guarded with my optimism,” Plante said, adding that when he saw the results, they were “about as inaccurate as I thought they could be.”

Uptake of the app has been explosive. According to Plante, Instant Blood Pressure outsold the popular Angry Birds game on some days, and for more than 150 days in 2014 and 2015 it was one of the top 50 best-sellers among all apps for sale in the Apple Itunes store, selling at least 950 copies per day on each of those days. Plante and colleagues choose it when their study began, but by the time they started enrolling participants, the app, which was selling for about $4.99, had been pulled from the market despite not only its popularity, but also many favorable reviews from users. Removing the app from the store, however, has no bearing on whether regular users are still using it to this day.

No Proof That It Works

One of the first physicians to take notice of the app—and become alarmed by it—was Iltifat Husain, MD (Wake Forest School of Medicine, Winston-Salem, NC), who wrote a blog post in July 2014 sharing his concerns about it, ultimately motivating Plante and colleagues to test it out.

“It was quite obvious to me that using just an iPhone microphone and a light wasn’t necessarily sufficient to measure someone’s blood pressure,” Husain told TCMTD. “Even in the app itself there was a small little disclaimer that said … ‘this app is for entertainment purposes,’ which no one really listened to if you look at the comments.” Screen captures of those comments taken by Husain, who is editor and founder of iMedicalApps.com, show that users were convinced the app was accurate.

The developers claimed that it works via undisclosed, proprietary technology. Be that as it may, Husain said before marketing health products like this to consumers, app developers should conduct their own simple study similar to the one by Plante and colleagues, and publish those results.

“This is one of the first studies of its kind that shows how an application that was extremely popular has been tested … and failed miserably,” Husain observed. “There are tens of thousands of people who paid money to use it and Apple received a portion of those proceeds as well. It’s kind of disconcerting to me that Apple would not have asked the developers to provide more data on how the app worked.”

Similarly, Plante said he is convinced, based on the results of his study, that the app is “spitting out random numbers” based on the data that users input, namely their sex, weight, height and age.

“I think you can certainly say that there is no legitimate algorithm behind the data that the application produces,” Husain concurred. “How many times were people falsely reassured that their blood pressure was okay?” Some users, based on their comments, were using the app daily and for titration of medications, which Husain said, “is absolutely terrifying.”

Although this particular blood pressure app is no longer available, at least 8 or 9 similar apps are still in the Itunes store for customers to purchase. Plante recommends steering clear of them and tells patients to rely instead on the cuff in their physician’s office or at their local pharmacy. For now, he added, all similar apps should be removed from the market. Still, he remains hopeful that the field of health apps will continue to grow and be useful for clinicians and patients alike.

“I’m disheartened by the findings of [this paticular app] but I actually am very excited about this kind of technology—that we can use these portable devices like smart phones to help people engage more in their health,” Plante said.

---

Source:  
Plante TB, Urrea B, MacFarlane ZT, et al. Validation of the instant blood pressure smartphone app. JAMA Intern Med. 2016;Epub ahead of print. 

Disclosures:

• The study was supported by a PJ Schafer Cardiovascular Research Grant. 
• Plante and Husain report no relevant conflicts of interest. 

Related Stories:

• Health Apps, Bar Codes, and Big Data: The “Brave New World” of Medical Device Innovation
• iPhone CTA Readings Accurate Enough to Rule Out CAD