The cost-effectiveness of early COVID-19 detection using AI embedded wearable devices: Evidence from the COVID-RED Trial.

Abstract Trial Design : Randomized single-blinded two-period, two-sequence, crossover trial. Methods Objectives Participants All residents of the Netherlands above 18 years of age who reads, understands and writes dutch. All eligible study participants who owns a smartphone, were recruited during the first quarter of 2021 Interventions All study participants recieved a bracelet which based on the dutch COVID-19 guidelines (control group), or the embedded AI (intervention group), warns the wearer of a possible COVID-19 infection. Objective Estimate the time-difference in COVID-19 detection using AI, and the dutch COVID-19 guidelines. Outcomes The incremental cost effectiveness ratio (ICER) to assess the costeffectiveness of using AI embedded bracelets. Results Number randomised 17,825 eligible study participants were randomized to the intervention group (N = 8,915) and control group (N = 8,910) Recruitment The trial has ended, and is currently being studied. Number analysed 2,952 and 2884 study participants opted out from the study in the intervention and control group respectively. The number of analysed study participants is 5,963 and 6,026 respectively. Outcome COVID-19 can be detected 2 days earlier using AI embedded bracelets, than following the dutch COVID-19 guidelines. This comes at an additional cost of 0.85 e. The ICER is -0.42 (-0.47, -0.38) e. Conclusions COVID-19 can be detected 2 days earlier using AI embedded bracelets and comes at a low additional cost per study participant. Trial registration Dutch Trial Register, NL930. Registered February 18, 2021, https://onderzoekmetmensen.nl/en/trial/23180 Funding Innovative Medicines Initiative (https://www.imi.europa.eu) 2 Joint Undertaking under grant agreement No 101005177.