Digital health can help screen test for COVID-19

A project funded by the National Institutes of Health will soon roll out a digital health tool to provide screening and testing for COVID-19.

Vibrent Health, which focuses on furthering precision medicine and research through digital tools, received its initial grant from NIH in September as part of the institute’s effort to support digital health solutions to address the pandemic. The National Cancer Institute and National Institute of Biomedical Imaging and Bioengineering selected the projects out of nearly 200 applications.

Fairfax, Virginia-based Vibrent Health received a total of $4 million and is aiming for commercial launch in the fall, Vibrent CEO Praduman Jain said.

Vibrent’s COVID-CARE program, once rolled out, will comprise multiple parts. Consisting of both a web and mobile application, users will be asked a series of questions related to symptoms, location, and demographics.

The answers are then inputted into an algorithm that determines the likelihood of the user having COVID-19. If a user is flagged as likely having the virus, a Quickvue At-Home test distributed from either Quidel or Everlywell is shipped that the patient can conduct at home. Jain said test kits can be delivered “within a few hours” in a surrounding geographic area and results are delivered within 10 minutes.

Results will be entered into the app, and the data can be used by health departments to pinpoint outbreaks or track infections.

“It starts to close the loop between screening and the test kit,” Jain said.

The program resulted after a year of strategy conversations with NIH, looking at the market and the need for infectious disease detection, he said. It is intended to help reduce unnecessary testing for patients who likely don’t have the virus.

As software has become more sophisticated, a software-enabled approach to screening, testing, and tracing became a viable option, Jain said.

The company traditionally has worked with partners rather than distributing its products directly to consumers and is talking to a variety of potential distributors for the commercial launch, he said.

Both NIH and Vibrent have been focused on using the tool for COVID-19, but Jain said it’s “going to evolve” to help the nation “get ready for the next pandemic.” The proprietary algorithm Vibrent uses was kept easily programmable to be used for other diseases, including influenza, he said. The app also collects data each day from different zip codes to monitor hotspots and infection trends, Jain added.

The machine learning algorithm relies on prevalence data from the NIH, Centers for Disease Control and Prevention, combined with symptom data from other data sets, including electronic health record data from EPIC Systems and Cerner’s networks and insurance claim data, to determine the likelihood of COVID-19, Jain said.

This data is de-identified and shared by public and government sources, he added.

The company used CDC and NIH data to develop 45 models based on different scenarios for COVID-19, influenza, and influenza-like illness spread in a patient’s region. The company randomly sampled 10,000 cases from NIH and the CDC so that the prevalence of diseases reflected one of the 45 scenarios, which were then organized from prevalence of COVID-19, influenza, and influenza-like illnesses.

In December, most regions in the US were in the scenario with a 2 percent prevalence of COVID-19, .01 percent prevalence of influenza, and 3.5 percent prevalence of influenza-like illnesses, Jain said.

Social isolation and restrictions on access to healthcare providers may have reduced the number of influenza and similar illnesses that were reported to CDC, he added.

NIH also asked Vibrent to demonstrate that its product could improve access to screening and testing in underrepresented populations.

The issues Vibrent is looking to address in these populations include trust in the system, convenience, the digital divide, and digital inequities of mobile technologies, healthcare information, and access to care, Jain said. The platform works both in self-supported and staff-supported settings, with options for virtual healthcare visits for users who request them.

To address issues with access and familiarity with technology, Vibrent purposefully designed the program to work across devices and software – it has so far been used by more than 400,000 people, 80 percent of whom “are from backgrounds underrepresented in biomedical research,” Jain said.

One of those groups involved several hundred Black participants in Richmond, Virginia, and the company is integrating feedback from that cohort to improve the program, Jain said.

Multiple languages are also included in the product, with an emphasis on straightforward language and quizzes provided to gauge understanding, he said.

A spokesperson from NIH’s National Cancer Institute added that the agency would facilitate the availability of the data collected, as well as associated analysis, from the studies that were funded under the contract.

This story first appeared in our sister publication 360Dx, which provides in-depth coverage of in vitro diagnostics and the clinical lab market.