A novel high specificity COVID-19 screening method based on simple blood exams and artificial intelligence

Felipe Soares; Aline Villavicencio; Michel Jose Anzanello; Flavio Sanson Fogliatto; Marco Idiart; Mark Stevenson

By the simple introduction of the AI model, we can potentially improve COVID-19 screening in emergency rooms. Given its NPV of 94.92%, we would be able to identify patients with very low risk for COVID-19. , The use of this tool along with clinical evaluation could reduce the number of traditional tests already adopted at the hospital's ER. Therefore, the first (and possibly most important) benefit is the possible reduction in the number of tests required to be performed on patients that are negative for COVID-19. Another benefit is the potential to develop prioritization queues for patients that our model classifies as likely positive, thus speeding up results for potentially infected patients. That is, with a specificity of 85.98%, suspected COVID-19 cases would be prioritized for further testing. In a global scenario of an extreme shortage of testing kits and laboratory supplies for PCR related exams, we foresee that our proposed method may have a great impact on patient diagnosis in ERs, especially in developing and low-income countries.