Feasibility of Commercially Marketed Health Devices for Potential Clinical Application

• Published in: Critical reviews in biomedical engineering Vol. 47; no. 2; p. 159
• Main Authors: Uchimura, Kevin D, Adamson, Teagan L, Karaniuk, Kara M, Spano, Mark L, La Belle, Jeffrey T
• Format: Journal Article
• Language: English
• Published: United States 2019
• ISSN: 1943-619X
• DOI: 10.1615/CritRevBiomedEng.2019026110

Concerns have been raised regarding the lack of validation on consumer-marketed health-monitoring devices. An investigation to characterize current health monitoring devices was carried out in the laboratory using widely accepted clinical and industry criteria. In total, 16 unique devices were examined. These devices were assessed according to their sensing modalities: step count, blood pressure, body temperature, electrocardiogram, blood oxygen saturation, and respiratory rate. Devices were tested at rest and immediately following exercise. Our results revealed that only four devices meet target requirements for accuracy. The AliveCor, a portable ECG monitor, accurately detected the heart rate for 87% of all recordings. To meet the target criterion for accuracy, the heart rate must be within ± 5 beats/minute or 10% of the standard measurement, whichever is lower. The Withings Pulse Ox, the Tinké, and the Santamedical SM-110 measured blood oxygen saturation with 2.1, 2.6, and 1.4 root-mean-square (rms) error, respectively. For blood oxygen saturation, the device should demonstrate rms error of < 3%. However, the Withings Pulse Ox and the Tinké failed to meet the accuracy criteria for their alternative biosensing capabilities: step count and respiratory rate, respectively. We conclude that the use of consumer-marketed health-monitoring devices for clinical or medical purposes should be undertaken with caution, especially in the absence of FDA or comparable clearance.