What you need to know: The Plante et al. Study
Updated: April 17th, 2016 4:54 PM PT
At AuraLife, we welcome third party researchers taking an active interest in assessing our technologies. Further, we believe open discussion with researchers is important in advancing health technology.
On March 1, 2016 our team became aware of a research letter titled "How Well Did an Instant Blood Pressure App Work?" which was published by JAMA Internal Medicine on March 2, 2016. A thorough review has been conducted of the study led by Timothy B. Plante, MD. We feel that critical flaws and deficiencies were present in how this study was conducted and the way its data analysis was performed. Here we present our concerns.
Study Deficiencies & Concerns
- Supported Range: The authors failed to disclose that approximately 20% of their data were associated with reference blood pressures outside IBP’s supported measurement range. As published on our our website and App Store description prior to the start of the authors’ data collection, IBP will not produce estimates outside of its supported range (Systolic 102-158 mmHg, Diastolic 65-99 mmHg). We believe the authors’ failure to disclose or mention IBP’s supported range misled readers.
- Improper Protocol Implementation: The authors did not properly implement key national guidelines which help ensure the reliability of comparative measures. The same-arm sequential measurement technique, defined by those guidelines in section 220.127.116.11, requires BPs to be collected as determination pairs with a reference device measurement taken before and after each test device measurement. This was not done. The ‘bounding’ of test device measurements ensures that any drift in a subject’s true BP during the protocol is properly accounted for and not incorrectly attributed to test device inaccuracies.
- Lack of Proper Study Controls: Proper study controls are required in conducting any sound scientific research. The authors failed to control for the most fundamental variable in their experiment – the version of the IBP algorithm itself. IBP is updated periodically, remotely on our servers, independent of any updates to a user’s phone. While each IBP algorithm update reflects an overall system improvement, each version is subject to unique data variability. The IBP algorithm was updated at least 5 times during the authors’ subject enrollment period.
- Intended Use: The authors investigated the effectiveness of IBP in diagnosing hypertension. However, as published within the app, on its website, and on its App Store description prior to the start of the authors’ study, IBP is not intended for such use. IBP is intended for use in estimating blood pressure before, during, or after recreational activity. For example, IBP may be used to better understand one’s natural, daily blood pressure rhythms or to estimate the impact of a meditation session or breathing exercise on blood pressure.
We share the authors’ concern at the idea of individuals, against the will of their physician, titrating their own therapies. We believe such a public health concern can be tackled through education. Many consumers understand that only a doctor is qualified to diagnose and treat disease such as hypertension and that they should not be self-treating or self-diagnosing. Nonetheless, we take an active and prominent role in ensuring our users understand Instant Blood Pressure’s specific intended use as demonstrated by our extensive in app disclosures and online materials.
It is our company’s mission to help people better access health insight. We envision a future where mobile technology plays a central role in helping people pursue their health and wellness goals. Instant Blood Pressure is a novel and constantly improving system which leverages recent advances across industry and academia in mobile sensor technology, machine learning techniques, and physiological understanding.
HOW THE INSTANT BP ALGORITHM WORKS
Through years of research and machine learning development we discovered unique pulse activity, mechanical cardiac activity, and vascular indices each of which correlate with blood pressure. Additionally, we leverage other biometrics which have been well documented in peer-reviewed clinical research to correlate with blood pressure. Instant Blood Pressure's algorithm leverages our unique discoveries, well understood biometrics, and each user's unique anthropometric data to produce every blood pressure estimate.