AI & RoboticsNews

How AI-assisted app from Pfizer and Ada Health aims to help COVID sufferers

Check out all the on-demand sessions from the Intelligent Security Summit here.


Getting COVID-19 isn’t fun, but the key to getting better faster has a lot to do with getting treatment faster.

Digital health vendor Ada Health today announced that it has partnered with drugmaker Pfizer on an artificial intelligence (AI)-assisted application to help individuals that are experiencing the onset of COVID symptoms to get the right treatment as quickly as possible. The new COVID-19 Care Journey application, operated by Ada Health, in partnership with Pfizer, can assist individuals in determining what course of action they can take to help remediate symptoms. With the right conditions it can also be used to help direct users to get the right antiviral medication to reduce the impact of symptoms and accelerate recovery.

The COVID-19 Care Journey will be available as a standalone service, as well as being part of the Ada app, which is an AI-powered mobile application that the company claims has been downloaded more than 12 million times. The Ada app uses AI to help assist in the initial triage and diagnosis for potential health issues.

“We pretty much help our users find out what health issues they might have and then how they can address it as a next step,” Daniel Nathrath, CEO and cofounder of Ada, told VentureBeat. “Now we are taking the usefulness of Ada a step further and, rather than telling users what they might have and what they can do,  we endeavor to actually deliver the solution to them in a seamless end-to-end care journey.”

Event

Intelligent Security Summit On-Demand

Learn the critical role of AI & ML in cybersecurity and industry specific case studies. Watch on-demand sessions today.


Watch Here

How the Care Journey service works

Nathrath explained that what the service does is it helps to determine if an individual is eligible for antiviral treatment for COVID.

“COVID antiviral drugs are now available and they are quite effective, but they’re not necessarily reaching the people who actually need them most,” he said. “Part of that is because people don’t actually know whether they’re eligible or not and sometimes even the doctors don’t really know.”

The COVID-19 Care Journey provides a specific eligibility questionnaire to help determine if a user meets the criteria as outlined by the U.S. Centers for Disease Control and Prevention (CDC) for usage of antiviral COVID medications. For those who are eligible, the service can connect the user with a telehealth physician who can then prescribe the actual medication.

While the service is being developed in partnership with Pfizer, which manufactures the Paxlovid antiviral, Nathrath emphasized that physicians are free to prescribe whatever antiviral drug is appropriate.  

After a physician issues a prescription, the Care Journey service can also help to make sure that the drug reaches the patient as quickly as possible, which is crucial when it comes to antiviral treatments as they need to be taken in the first few days to be effective. 

“The idea here is really to make the whole journey a lot more painless for the patient,” Nathrath said. “If you have an infectious disease, the last thing you want — and the last thing everyone else wants — is that you sit in the waiting room of your family doctor’s office, and you pass along the infection to 10 more people.”

The AI doctor will see you now

The core of the Ada application relies on an AI system that the company has been developing over the last dozen years.

Nathrath explained that his company’s AI has been training on a very large curated medical knowledge base. It’s an effort that he said has involved hundreds of doctors to fine-tune and get the correct results.

As part of the AI development process, Ada Health developed its own domain-specific programming language. With that language, it is able to potentially identify several thousand different diseases and tens of thousands of symptoms. The information is all collected as structured data and behind it is a probabilistic reasoning engine.

A probabilistic engine works differently than the large language model (LLM) approach that has become popular in recent years with AI.

“We made a conscious decision to take the approach we’ve taken because it’s more explainable, it’s not a black box,” Nathrath said. “It’s a much more guided conversation, then, than the one you would have with something like ChatGPT and for this specific use case, we believe it’s actually the better solution.”

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.


Author: Sean Michael Kerner
Source: Venturebeat

Related posts
AI & RoboticsNews

DeepSeek’s first reasoning model R1-Lite-Preview turns heads, beating OpenAI o1 performance

AI & RoboticsNews

Snowflake beats Databricks to integrating Claude 3.5 directly

AI & RoboticsNews

OpenScholar: The open-source A.I. that’s outperforming GPT-4o in scientific research

DefenseNews

US Army fires Precision Strike Missile in salvo shot for first time

Sign up for our Newsletter and
stay informed!