Health systems are increasingly turning to big data to drive innovation in their facilities. While innovation can take many forms, one Midwestern health system chose to focus on what it calls "radical access to care," successfully integrating social determinants of health data into clinical workflows to support disadvantaged populations.
In a webinar sponsored by Pieces Technologies and hosted by Becker's Hospital Review, Ruben Amarasingham, MD, founder and CEO of Pieces Technologies, and Matthew Warrens, vice president of innovation partnerships at Peoria, Ill.-based OSF HealthCare, explained how connecting community-based organizations to the health system is improving patient outcomes.
Pieces Technologies, a health IT firm and spinoff of Dallas-based Parkland Center for Clinical Innovation, applies artificial intelligence, natural language processing and predictive modeling capabilities to patients' clinical and social determinants of health data to follow their care journeys and provide decision support across the care — and community — continuum.
Like many health systems, OSF HealthCare was looking to hone its innovation strategy with concrete, measurable goals. To do so, officials focused their energies on a few select priorities, including determining how to best serve disadvantaged populations through social determinants of health and improving what Mr. Warrens called "radical access to care."
"When you think about a patient's healthcare journey, the health system is really just one short part of that," Mr. Warrens said, noting OSF HealthCare has created a community-based health and wellness steering team and deployed digital tools — like those from Pieces Technologies — to connect acute and community care teams.
As part of their mission to drive positive outcomes across their community, OSF HealthCare and nearby community-based organizations worked with two of Pieces Technologies' data software platforms: Pieces Decision Sciences (DS) and Pieces Iris.
1. Predicting adverse events in the hospital setting. To improve outcomes within the hospital's four walls, OSF HealthCare licensed Pieces DS, a cloud-based software platform that sits atop EHRs and clinical data warehouses, to predict and monitor the likelihood of events like readmissions or excess length of stay.
Pieces DS uses natural language processing to read structured and unstructured data in patient charts, such as free-text notes from physicians and case managers, to identify at-risk patients. It assists clinicians in creating a holistic and personalized discharge plan, noting its recommendations in the patient's record.
"It's really trying to understand context," Dr. Amarasingham said. The system analyzes clinical notes to identify social issues such as homelessness and flags patients based on pressing clinical concerns, like high risk of developing sepsis. "This is all to make sure that the internal resources of the health system are alerted to important information about the patient," he added.
Many of the software's care recommendations involve referring patients to outside organizations — for example, connecting a homeless patient with a local food pantry or shelter. Hospitals and community-based organizations deliver these referrals through a separate Pieces Technologies platform called Pieces Iris. The system monitors the patient across their care journey, alerting relevant stakeholders whether he or she presents at the referral organization for care or services.
"We're learning from our complex care managers that this is the real 'delighter,'" Mr. Warrens said. "They can go in and see that a referral they made was completed by the patient, and, more importantly, if [the patient] hasn't made it, they can reach out to the patient with an intervention to try to re-engage them."
2. Creating connections across community-based organizations. In OSF HealthCare's surrounding community, Pieces Technologies worked with OSF HealthCare's new outpatient wellness center, as well as social service providers to implement Pieces Iris, a cloud-based case management platform that standardizes patient data to support operations and referrals. The platform connects health systems and community-based organizations.
For Dr. Amarasingham, the software aims to enable community-based organizations to collaborate with one another. For example, a food pantry might use the system to refer an individual in need of a job to an employment assistance organization, while a social support organization might tap a nearby health system to evaluate medication adherence.
"Our vision for how we improve health systems' connection to the community really focuses on the community organizations themselves," Dr. Amarasingham said. "Often separate and apart from the health systems, you want to be able to build what we call the 'musculature of the community.'"
Hospitals and community-based organizations see value in using Pieces Iris to manage follow-up care, encouraging collaboration between hospitals and post-acute or home health providers. Across organizations using its software, Pieces Technologies has seen a 26 percent relative reduction in readmissions for Medicare patients.
"You can't divorce the decision sciences support we're giving health systems from the rest of the community," Dr. Amarasingham said. "The deeper the connection and the number of organizations using the software, the more power levers that a health system and community organizations have to achieve their missions."
Listen to the webinar recording here. View the webinar slides here.