FDA grants EUA to COVID-19 screening device using machine learning

Dive Brief:

  • FDA has granted emergency use authorization to a machine learning system designed to spot signs of COVID-19 infection in otherwise asymptomatic individuals.
  • The product, which is not classed as a diagnostic device, consists of an armband that uses light sensors and machine learning to identify whether a person is showing signs of COVID-19 such as hypercoagulation. FDA cleared the product for use in people without fever or other COVID-19 symptoms.
  • Tiger Tech Solutions secured the EUA on the strength of data from studies in school and hospital settings that found it correctly predicted COVID-19 status the vast majority of the time.

Dive Insight:

Estimates of the proportion of COVID-19 cases caused by asymptomatic individuals vary but make up a significant minority of transmission events even at the low end of the range. The role of people with no significant symptoms in the transmission of the coronavirus has led to the proliferation of efforts to identify hidden carriers of the virus such as screening programs that use antigen tests.

Tiger Tech has added another product to the portfolio of screening tools. The device is intended for use in people with a temperature below 100.4ºF, indicating no evidence of a fever, and no other symptoms of COVID-19.

The person being screened wears the device around their arm for three to five minutes. At the end of that time, the device displays a light of one of three colors to indicate whether it found biomarkers of conditions such as hypercoagulation or generated a negative or inconclusive result.

Tiger Tech’s biomarkers are signs the person may have COVID-19, which affects blood clotting. In the event of a positive result, the person should be treated as if they were found to have a high temperature. That could mean denying the person entry to the site performing the screening and requiring them to get tested for COVID-19 and quarantine until the result.

FDA granted EUA based on results of two studies. Used in hospital settings, the product found 69 positive cases among 467 asymptomatic individuals. Positive percent agreement, also known as sensitivity, was 98.6% and the negative percent agreement was 94.5%. Tiger Tech saw similar results in the confirmatory school study.

The product provides institutions with another way of determining whether someone may have the coronavirus before granting them entry. However, with antigen testing scaling up at sites that need to screen people, most notably schools, and offering results in around 15 minutes, Tiger Tech faces competition for the asymptomatic screening market.