Predicting Patient Outcomes: ClosedLoop.ai Secures $11M Series A For Health Models

Even before a patient walks into the doctor’s office, ClosedLoop.ai aims to predict whether that patient may be admitted to the hospital soon or is likely to cancel an appointment.

Subscribe to the Crunchbase Daily

The Austin-based company is using a new $11 million Series A cash infusion to continue developing its health care data science platform that uses artificial intelligence, machine learning and automation to enable data scientists to build and deploy predictive models.

“We consider ‘who is most likely to X,’ for example, admitted to the hospital, develop a chronic disease or be a no-show for an appointment,” ClosedLoop co-founder and CEO Andrew Eye told Crunchbase News. Eye and Dave DeCaprio founded the company in 2017.

The Series A was co-led by Greycroft and .406 Ventures with participation from Silicon Valley Bank and Meridian Street Capital. Including the new round, ClosedLoop has raised a total of $15 million, which includes a $3 million seed round in 2018, according to Crunchbase data.

The company intends to invest the funds in product and engineering, as well as sales and marketing. Eye said the company has attracted recent attention after being rated a “top performer in health care-focused AI” on the KLAS Healthcare AI 2020 report, as well as being named one of seven finalists in the Centers for Medicare & Medicaid Services’ Artificial Intelligence Health Outcomes Challenge.

“With all of this recognition, the timing for the new investment is good,” Eye added.

Predicting behavior

It’s well known that 5 percent of Americans account for 50 percent of the country’s health care costs. If doctors are able to identify the 5 percent of at-risk patients within their practices, they can intervene with them directly, Eye said.

“If you can find those people, and do something like a home visit, you have a bigger opportunity to impact negative health outcomes and cost impact,” he added.

ClosedLoop’s content library provides more than 2,000 health care features, including medication adherence, frailty and co-morbidities so that predictive and prescriptive models for common health care use cases can be built.

The company is disrupting the traditional model of health care, which involves fee-based care. Instead, value-based care is becoming the new way, and providers have a financial incentive to take care of a patient prior to them coming in, Eye said.

Using insurance claims, doctors can see health history to spot potential chronic diseases and other challenges, but social determinant information helps doctors see if someone has access to fresh food, live in a walkable area or may need transportation or referrals.

“You could identify people overutilizing the emergency department and give them a free Uber ride to follow-up visits,” Eye said. “It’s unlikely you can do that for everyone, but you could do it for people most likely to have these problems. You could also predict things like falls, and help the patient put a step stool next to the bed. It’s a $50 action versus repairing a broken hip.”

What’s next

Although Eye would not disclose revenue figures, he said ClosedLoop’s growth is a result of working within four market buckets: providers, payers, pharma and health IT. Health care spending was 17.7 percent of the nation’s Gross Domestic Product in 2018, according to the CMS.

Eye said that is more like 25 percent today, with $1 out of every $4 spent on health care.

“The industry is behind in using artificial intelligence to look at electronic health records and claims to see what is wrong with you and predict future health events,” he added. “By providing an end-to-end platform purpose-built for health care and the industry’s largest catalog of prebuilt models and features, we allow customers to go from raw, messy health care data to production-ready predictive models in as little as 24 hours.”

Illustration: Dom Guzman