MedCityNews: Atropos Health Launches Evidence Network to Back Its Physician Consult Service

Insightful words from Katie Adams on Atropos Evidence Network and the impact of better evidence for equitable healthcare. Read the original article on MedCityNews.


Excerpted below:

Atropos Health — a company that delivers clinical data to physicians at the point of care to help them treat complex patients — recently launched a clinical advisory board and new evidence network. Both announcements are meant to enhance the credibility and transparency of its physician consult platform, CEO Brigham Hyde said at ViVE in Nashville.

When I talked to to Atropos CEO Brigham Hyde in August following the company’s $14 million Series A funding round, he described the company’s point-of-care report service as a “full boat, publication-grade observational research study on deidentified EMR patients, including the use of high dimensional propensity score matching for statistical error correction.” When we caught up this week at ViVE, a healthcare innovation conference in Nashville, he explained how the company has recently strengthened the credibility of its services through its newly formed clinical advisory board and evidence network.

“Our services are designed to feel like physicians are talking to a colleague, kind of like ChatGPT. They type their question to us in a few sentences, and we turn around a really clear and detailed response. And it takes them less than 15 minutes to consume the information, and it’s backed by millions of patients’ evidence, which gives them great confidence in decisions they’ll make,” Hyde explained.

“Let’s say I have a complicated oncology patient. Well, maybe we’ve only seen a few of these patients in my institution, so I need another option. Our partners like ASCO CancerLinQ have millions and millions of oncology patients. So we can take the question, and without any data having to be transferred or any HIPAA risk, we can send that question to ASCO and turn the answer quickly,” Hyde said.

When a physician receives their answer from the evidence network, it comes stamped with a “real world data score,” which Hyde described as a “credit score for data.” Atropos developed this data scoring system to address data bias and the fact that studies reproduced on different data often have different outcomes. The score evaluates the dataset’s quality based on elements such as size, completeness and longitudinality. “In doing that, we’re upping transparency by showing where there’s bias or weakness in results, and we’re also offering the opportunity to run your question by multiple datasets,” Hyde explained.

Read more of the article by Katie Adams on MedCityNews



Previous
Previous

Why Early Tests of ChatGPT in Medicine Miss the Mark

Next
Next

How Well Do Large Language Models Support Clinician Information Needs?