Table 1 Accuracy problems in studies using physiological measurements by smartphone apps
Measure | App | Smartphone | Participants | Finding | Study |
|---|---|---|---|---|---|
Blood pressure | Instant Blood Pressure | iPhone 5 and 6. | 85 patients and staff; 53% with hypertension. | Measures “were highly inaccurate” | Plante et al. (2016) |
Heart rate | Instant Heart Rate Heart Fitness Whats My Heart Rate Cardio Version | iPhone 4 and 5. | 108 patients, exluding those in critical condition. | “substantial performance differences” between the four apps | Coppetti et al. (2017) |
Heart rate Blood pressure Oxygen saturation | Instant Blood Pressure Instant Blood Pressure Pro Pulse Oximeter Pulse oximeter Pro | iPhone 5S | 100 healthy participants | “applications evaluated do not provide clinically meaningful data” “inaccurate data.. can potentially contribute to patient harm” | Alexander et al. (2017) |
Heart rate | Runtastic Heart Rate Monitor Instant heart rate+ | iPhone | 15 regularly active college students | “Poor correlation to ECG” during moderate to high intensity exercise | Bouts et al. (2018) |
Step counting | Argus: calorie counter and step | Android phones: Samsung, OnePlus, Moto, Oppo, Galazy, Huawei, LG, Google, Sony and Agora running Android 4.4 to 8.1 Apple: iPhone 6, 6S, 7, 8, running iOS10.3–11.4. | 48 healthy participants | “extraordinarily large error ranges for both..phones” “appear unsuitable to detect steps in short, slow, or non-stereotypical gait patterns” | Brodie et al. (2018) |
Sleep | Sleep time | iPhone 4 s and 5 | 20 participants with no sleep disorder | “absolute parameters and sleep staging…. correlate poorly with polysomnography” | Bhat et al. (2015) |
Sleep | MotionX 24/7 | iPhone 4 | 78 children and adolescents with suspected sleep disordered breathing | “did not accurately reflect sleep or wake and should be used with caution” | Toon et al. (2016) |