Each company overcame dozens of other entrants in the innovation competition's largest participant pool to date.

Online prehabilitation service, NLP case data summarization tool named as RWJF pitch contest champions

By Dave Muoio
Share

The team from Ooney makes their pitch to the judges.

Following a session of live pitch presentations and evaluations, judges have awarded Ooney and Social Impact AI Lab — New York with the top prizes in two healthcare innovation competitions focused on, respectively, home/community care and social determinants of health.

The innovation contests are part of an annual event held here at the Health 2.0 Conference in Santa Clara, and are conducted in partnership with the Robert Wood Johnson Foundation (RWJF) and Catalyst @ Health 2.0. This year’s victors were each presented with a $40,000 prize, with $25,000 and $10,000 awarded to second and third place.

“In all the years we’ve been doing this, we had the greatest number of applicants [this year],” Indu Subaiya, senior advisor and cofounder of Health 2.0, said yesterday while introducing the event. “The Home and Community-based Care Challenge drew 97 applicants, the Social Determinants of Health [Challenge] there were 114. And we had over 50 judges from a number of different disciplines of expertise — academia, non-profit, venture capital and healthcare — all weighing in.”

Ooney’s individualized surgery preparation tool was followed up by AI home caregiver assistant Wizeview and physician house call service Heal, while Social Impact AI Lab — New York’s automated case note consolidation tool won out over Community Resource Network’s SDOH profile builder and Open City Labs’ referral and enrollment aid.

Online prehabilitation to reduce surgical complications

For its pitch, Ooney demoed its online app for patients preparing to undergo surgery. By providing a selection of targeted exercises or routines, the company looks to reduce the chance of poor outcomes, explained founder Dr. Emily Finlayson.

“Every year, over 4 million operations are performed on frail older adults. That’s nearly 50% of all major high-risk operations in the US, and we know from epidemiology that older adults suffer complications at exponentially higher rates than their younger counterparts,” she said. “Research has shown that focused prehabilitation can improve outcomes after surgery, accelerate recovery, accelerate function recovery, prevent debilitating and costly complications and decrease hospital length of stay. … The problem is that this is an activity at that takes place at home, and has to have engagement at home.”

To address this challenge, Ooney’s Prehab Pal tasks patients with completing a 47-question survey about their condition and upcoming procedure. From this, the tool devises a personalized program of exercises complete with video instructions on how to complete each motion.

Of particular note are the adherence tools included within the web app, such as regular reminder emails, overview summaries of completed and upcoming tasks, online Q&A with a trainer, weekly check-in calls and other escalation protocols of the patient doesn’t check in for multiple days.

Natural language processing to intelligently compress case notes

The event’s second winner was Social Impact AI Lab — New York, a collaboration between non-profit provider MercyFirst, software services firm Augmented Intelligence and other local human service agencies. Together, the groups built a tool that scans a large volume of social work case notes and records to generate an easy-to-read case summary for service providers. Along with a top-level timeline of major events, the platform user is also shown a timeline graph displaying a quantified risk value, which they can use to hone in on past events that may have caused a family’s risk to spike and quickly understand the most pertinent aspects of their case history.

“What’s going on beneath the surface is the application of natural language processing AI. AI systems are trained to understand context from raw data,” Marty Elisco, CEO of Augmented Intelligence, said. “We take millions of case notes from the agencies we work with to train them on child welfare content, so when a new case … come[s] in we’ll know what to work for — particularly, SDOH factors that are presented to the case worker in a way so they can quickly get up speed on the case.”

Elisco and MercyFirst Chief Data Officer Besa Bauta said that the current plan is to ensure that the tool is performing well within the New York region, and then build a replicable model for broader application.

Responding to questions from the judges, the pair highlighted their tool’s potential for time savings among social case workers as its major value add. However, they made it clear that this capability shouldn’t result in greater caseloads, but rather a chance for these workers to spend more time with their charges and improve the services being provided.

“I don’t want them to do have 30 families, I want them to do the best job they can with five to 10 families,” Bauta said. “Our goal is to provide better care to help so they’re not in foster care … so I want them to be able to pick up that information so that we’re providing greater care, greater community collaboration, so they can return to their homes.”

“It’s much less about quantity and value, than it is about quality of care by picking up the information they need to make decisions,” Elisco added.