Validation of the Performance of the COBOX Solution for Measuring Heart Rate and Respiratory Rate

Overview

Quantiq.io "COBOX" is a software as a medical device. It relies on remote photoplethysmography, allowing for the remote measurement of a patient's physiological parameters from a healthcare professional's or patient's device which can be a smartphone, a computer or a health station with an optical camera. Quantiq.io software is a medical device which is : contactless for the measurement non-invasive for the patient In order to meet safety and quality standards, new health technologies require rigorous clinical validation with measurement performance at least equivalent to that of one CE marking measure device. In this context, we assume that the COBOX device will take measurements of heart and respiratory rates equivalent to those taken by medical devices and current practices.

The planned procedure is: 1. Oral information to the patient about the test by the nursing staff. 2. Patient signs the consent. 3. Collection of the patient's age, gender and Fitzpatrick phototype. 4. Collection of the patient identification number. 5. The nursing staff installs the oximeter on the patient's finger. 6. The nursing staff starts COBOX and begins to count the patient's breath in parallel 7. After 15, a first measurement is made by COBOX. The pulse oximeter reference result is acquired in parallel. 8. After 20s (5s after the previous measurement) a second measurement is made by COBOX. 9. After 25s (5 after the previous measurement) a third measurement is made by COBOX. 10. After 30s, the nursing staff enters the number of patient's breaths counted since the start. The performance of the tool will be estimated by the concordance between the measurements obtained from the application of the reference technical process via a CE marked device and with the measurements obtained by COBOX. An intraclass correlation coefficient (ICC) with its 95% confidence interval (95% CI) will be calculated. Measurement errors will be calculated and expressed as a percentage (root mean square error, etc.). The two study samples will be compared by Student's t test and Fischer's F, with a desired significance for P value = 0.05, for all the analyses. A DIXON test will be performed to assess the presence of statistically outliers within the samples. These data will enable the COBOX tool, a class IIA medical device, to be validated in terms of measurement performance compared to a CE-marked device used in a reference technical process for taking measurements and a reference method for manual counting.