Intermountain Health, the largest nonprofit healthcare provider in the Intermountain West region, struck a multi-year collaboration with a young generative AI startup this month.

The health system is teaming up with Layer Health, a Boston-based healthcare AI company that uses large language models (LLMs) to improve clinical data abstraction without increasing administrative burden. Intermountain’s venture capital arm is making a strategic investment in the startup, and the organization is deploying the technology across several of its clinical registries.

Clinical data abstraction is the process of manually reviewing and extracting key information from patients’ medical records — this typically means taking unstructured data and transforming it into standardized, usable data. Layer uses LLMs and a range of other machine learning techniques to automate this process, explained CEO David Sontag.

Sponsored Post The 5 Best Medical Billing Outsourcing Companies Medical billing is a tedious administrative task, but outsourcing it may be the key to increased revenue. Explore the top medical billing outsourcing companies and the benefits of partnering with one.

“For the clinical registry abstraction use case, we have fine tuned a range of AI models that can reason across a patient’s record and answer the questions found in the clinical registry, such as ‘Was Thrombolysis given for this patient’s stroke?’ and ‘If not, what was the contraindication?’ This AI-enabled chart review engine is used to power software that is used by Intermountain’s clinical abstraction team — the software provides suggested answers to the clinical registry questions, supported by snippets of evidence from the patient’s chart,” he stated.

The scope of this partnership includes registries in the cardiovascular, stroke and bariatric surgery spaces.

For each registry, the full longitudinal patient chart often needs to be parsed, including structured data elements, like labs, vitals and diagnosis codes, as well as unstructured data, like clinical notes and radiology reports, Sontag noted.

He pointed out that Intermountain participates in a large number of clinical registries, which requires a large manual human chart abstraction effort.

“Using Layer’s software, Intermountain has the opportunity to enable its quality staff to work ‘top of license’ — making operational improvements based on insights from the data, rather than spending the majority of their time collecting the data,” Sontag declared.

Phillip Wood, Intermountain Ventures’ executive director for partnerships, said that his team had been watching the market for solutions to improve clinical data abstraction — and that when they met with Layer’s team, they felt the technology was the right fit for their organization.

“When we were introduced to Layer Health, we were impressed by their team and approach and were excited by the early success they had so far in their initial market rollout. After technology demos with our clinical excellence team, we determined that Layer would be a good partner to work with in developing and validating this use-case of AI technology,” Wood stated.

To determine the success of the partnership, Intermountain will be measuring the platform’s efficiency and accuracy in comparison to its previous benchmarks, he added.

“If we can improve efficiency and accuracy via the tools, it will enable our caregivers to do more with other quality organizations and improvement initiatives,” Wood remarked.

Photo: ra2studio, Getty Images