Smartphone Application-Based Vital Sign Monitors

Overview

Health-related applications for smartphones represent a new and popular market for consumer oriented products. These applications are generally not considered medical devices and, as such, are not under the purview of the Federal Drug Administration. Some of these applications purport to measure vital signs, but do not have any published data to confirm accuracy of such measurements and their developers have not been willing to divulge the methods by which vital sign readings are produced. Use of unvalidated application derived data may contribute to patient harm as patients may make healthcare decisions based on the results these applications provide. This study aims to assess the accuracy of several such applications by comparing their findings to those of clinical vital sign monitors commonly used in the perioperative setting.

Recruitment of volunteers take place via announcement at the Anesthesiology Departmental Grand Rounds as well as by flyers posted in the UT Southwestern Department of Anesthesiology \& Pain Management. After verbal consent is obtained, each volunteer will be placed in the seated position. Upper-arm non-invasive blood pressure will be measured using WelchAllyn Propaq CS via a standard size cuff placed in the standard fashion between the elbow and shoulder. Pulse oximetry measurement will be obtained using the WelchAllyn Propaq CS via the 4th digit of the hand. The smart phone applications are Instant Blood Pressure, Instant Blood Pressure Pro, Pulse Oximeter and Pulse Oximeter Pro and purchased by the principal investigator and downloaded via the iTunes App Store onto an iPhone 5S with iOS 7.1.2 owned by the principal investigator. The application-based measurements will be performed by the volunteer according to the instructions from the application to the user without explanation from the investigators. Systolic blood pressure (SBP), diastolic blood pressure (SBP), pulse (P) and peripheral capillary oxygen saturation (SpO2) will be collected from 100 volunteers. Serial measurements will be taken in the following fashion: right upper extremity, smartphone application measurements, left upper extremity and repeat smartphone application measurements. Thus, for each volunteer each data point referenced above will be recorded six times: two times using standard equipment readings (one per extremity) and four times with application-based readings (two times per application). The two applications to measure blood pressure (Instant Blood Pressure and Instant Blood Pressure Pro) both utilize the camera and flashlight on the iPhone. In addition, the Instant Blood Pressure application also utilizes the microphone and instructs the user to hold the iPhone against the chest in a specific fashion to complete its measurement. Both applications also request the user to input certain demographic data: Gender, Age, Height and Weight. In addition, one application (Instant Blood Pressure Pro) also instructs the user to input data on "Body Condition" and provides the following options: Poor, Below Average, Average, Above Average, Good, Excellent and Athlete. The application does not provide any direction on how one is to distinguish between these options. If these applications measure data based on some yet unexplained function using the iPhone's onboard technology, then changing the demographic data should not alter the result of the measurement. In the same way, a sphygmomanometer or pulse oximeter does not need to know one's gender, height, weight or age to accurately measure blood pressure or pulse. The application based vital sign readings may change in concert with changes to demographic variables even though clinical vital sign monitors remain relatively consistent. To test this hypothesis, in a subgroup (n=20), serial measurements using both clinical vital sign and application based vital sign readings will be evaluated by changing demographic variables in the application based vital sign monitors. For this subgroup the volunteer will perform serial measurements with the following changes in demographic data for each application: Correct data, Age + 10, Age + 20, Age - 10, Age - 20, Weight - 10%, Weight - 25%, Weight + 10% and Weight +25%. Additionally, for the Instant Blood Pressure Pro application additional readings will be taken changing the "Body Condition" demographic as follows: Poor, Average, Good and Athlete. For each measurement taken using the applications another will be taken using the clinical monitor, or thirteen total readings per volunteer.