ARTICLE: Critical Flaws in the Validation of the Instant Blood Pressure Smartphone App—A Letter from the App Developers—Reply
JOURNAL: JAMA Intern Med. 2016 Sep 1;176(9):1408. doi: 10.1001/jamainternmed.2016.4510.
In response to: http://archinte.jamanetwork.com/article.aspx?articleid=2545990
In Reply In a letter to the editor in this issue of JAMA Internal Medicine, Archdeacon et al raise 4 concerns about our assessment of the Instant Blood Pressure (IBP) app. First, regarding IBP’s limited functional range, we included all participants as blood pressure (BP) cannot be known without measurement, and assessing BP is the app’s purpose. We expected IBP to present out-of-range errors if BP fell outside its range. However, no such error was presented, and IBP commonly reported a falsely reassuring BP when the actual BP was beyond its upper limit. Among the 6 participants with reference device systolic BP greater than 158 mm Hg, IBP reported a mean (SD) systolic BP of 129.4 (6.8) mm Hg; only 1 user received a measurement in the hypertensive range (systolic BP≥140 mm Hg).
Second, with respect to validation procedures, we used an accepted alternative approach of random-order paired measurements (A-A-B-B or B-B-A-A).1 There were no statistically significant effects of order on accuracy measures. Traditional guidelines are appropriate when an observer takes measurements with the standard and investigational techniques while the participant remains still. For IBP readings, the participant is the observer, who manipulates their smartphone and therefore is not still.
For a link to the full article, click here: http://archinte.jamanetwork.com/article.aspx?articleid=2545999
Link to abstract online: https://www.ncbi.nlm.nih.gov/pubmed/27598762