Testing the reliability of commercially available smart devices to remotely assess the intensity of physical exertion during respiratory rehabilitation

Smartwatches and similar devices, such as smart bands or smart rings, have became in-creasingly popular given their capacity to monitor an expanding range of physiological parameters such as heart rate, peripheral oxygen saturation, skin temperature, and to es-timate others such as sleep or physical activity patterns. The aim was to assess the accu-racy and reliability of data collected using four commercially available devices ( cum se cheama ) under a real life scenario in order to assess the usefulness of such devices in tel-erehabilitation programs. two real life n=1 studies were conducted recording simultaneously physiological data using two devices; main parameters were heart rate and peripheral oxygen saturation; other derived measures were also collected such as walk distance, walk cadence, sleep time or estimated energy expenditure. direct measurement physiological parameters such as heart rate and peripheral oxygen satura-tion were reliably and consistently recorded by the investigated models. There were signif-icant differences between devices for composite measurements such as distance walked and walk cadence; such differences seem related to the algorithms used to process the raw data and not to the physical performance of the sensors. commercially available smartwatches may be used in clinical rehabilitation scenarios, provided data are cautiously interpreted. Using such devices just to acquire raw biological signals and con-ducting centralized data analysis might mitigate some of the reliability problems expected to occur when using non medical grade equipment.