OSF Healthcare is using predictive AI to streamline Care Navigator workload

Researchers at OSF Healthcare and its partners have developed an artificial intelligence algorithm that can predict each cancer patient navigator’s upcoming week of work for their current patients. They also built a second machine learning model that distributes new patients among sailors within specialties to better balance workload.

why it matters

Patient navigators can help health systems shorten the time from diagnosis to treatment, but programs are often underfunded, leading to a heavy workload.

OSF Healthcare, operated by the Sisters of the Third Order of St. Francis, has 10 acute care hospitals and five critical access hospitals with 2,084 licensed beds in Illinois and Michigan. The Peoria, Illinois-based health system’s Cancer Patient Navigators provide education, advocacy and support to cancer patients and help facilitate their care journey.

Innovative approaches help retain employees – “our greatest asset,” said Dr. Jonathan Handler, OSF senior fellow for healthcare innovation, in the health system announcement.

To achieve greater workload fairness for its pool of CPNs, OSF collaborated with the nearby University of Illinois College of Medicine, the University of Illinois at Urbana-Champaign, and Northwestern University on a retrospective study to develop and test machine-learned algorithms. In partnership with the Feinberg School of Medicine. The ability of the algorithm to outperform random assignment and create more equitable CPN workloads within a feature.

The researchers note that one sailor in a particular field may have a lot more work to do at a given time than others in the same specialization, he said.

In their report published in May in JCO Clinical Cancer Informatics, the researchers wrote, “CPNs do not transfer their existing patients to other CPNs when the workload becomes excessive.”

“They prefer to retain their patients for the laudable purpose of maintaining an ongoing patient-CPN relationship. Therefore, the algorithm uses the only lever available to equalize the workload: the distribution of new patients.

They used a three-year data set compiled from electronic health records, including demographics, cancer type and prior healthcare use, to assess the past workload of 13 specialty CPNs employed at the health system’s largest hospital. As per the report, the data set contained 273,057 records, which included 13,033 unique patients.

The researchers then built three supervised regression models, each built from one of the most common and successful open-source machine learning libraries. The third step was to develop a distribution model that could minimize the difference between those sailors in their upcoming week’s workload.

“Dozens of input features were used to make predictions each week for each patient,” he said.

“Our program strives to maintain the patient-CPN relationship, so the only consistency constraint imposed was on allocation to ensure that patients remained with their initially assigned CPNs throughout their time on the panel.”

In addition to their retrospective simulation analysis, the researchers compared the predictor-informed distribution with the random distribution and assessed the resulting workload differences between navigators in the same cancer specialty. They note that OSF’s current CPN workload decisions do not consider anticipated patient needs, navigator experience, and current workload.

They noted that the predictor-informed model achieved significantly higher workload fairness than the random distribution.

“To our knowledge, this work may represent the first description of an automated, algorithm-driven approach to equalizing CPN workloads,” the researchers said.

“Optimization has been applied to health care staffing and patient allocation in other health care domains, but it is usually applied to shifts rather than individuals.”

According to OSF’s announcement last week, the plan is to integrate the OSF Community Connect tool, a platform that automates workflows, and test its efficacy ahead of the planned opening of the OSF Cancer Institute in 2024.

big trend

Around the world, AI is being used or developed to address the unprecedented level of burnout experienced in the health care workforce.

Software companies and healthcare IT developers use machine learning to address a range of healthcare tasks, from transcribing audio or video, to addressing administrative inefficiencies and providing insights about patients and patient populations – all competencies and uplift the healthcare workforce to improve patient outcomes and prevent overburdening professionals.

For example, the UCLA health system is using algorithms to make its nursing workloads more equitable.

Nurse informaticists Meg Furukawa and Stesha Selsky, of UCLA Health, have developed a machine learning model that generates personalized real-time workload acuity scores for all nursing staff.

Charge nurses use the generated score for decision support, which all nursing staff can view, and they can adjust workload or request additional staff as needed.

Before HIMSS23, he told Healthcare IT News The ML model relies on existing patient chart information and other nursing documentation from electronic health records and other systems. Furukawa said that his guiding principle was to create a tool that would not add administrative burden and would rely on data generated from existing workflows.

Furukawa and Selsky said that by working collaboratively, the use of AI by UCLA Health’s nursing staff has helped achieve more equitable nursing resources and patient assignments.

“We included bedside nurses, of course, from the very beginning, we had nursing leadership, we had our workload acuity champions as part of the project who really gave us input and feedback and helped us develop tools and Used to help validate the equipment along the way, Furukawa explained.

On the record

“Our cancer patient nurse navigators are highly dedicated, and their workload can be overwhelming at times,” OSF’s Handler said in a statement about the new AI findings. “They never want to fall short of a patient, so they cut themselves short, work extra hours, and sacrifice their own well-being to help patients. We hope that our system can equalize those workloads and improve their work-life balance,” he said.

Andrea Fox is a senior editor for Healthcare IT News.
Email: afox@himss.org
Healthcare IT News is a HIMSS Media publication.