Hospitals and health systems are using artificial intelligence to improve patient selection for clinical trials, prevent healthcare supply chain disruptions, identify diseases at an early stage, and shorten patient wait times – just to name a few. of the myriad AI use cases already in action, according to a virtual conference briefing hosted by Premier Inc. last week.
Diversify and standardize data
Premier, Inc. is a digital transformation company that works with 4,350 U.S. hospitals and healthcare systems and more than 300,000 other providers and organizations, according to its website. The company is finding that hospital systems are currently looking to AI to support patient data analytics and workflow optimization, according to Mason Ingram, director of payer, policy and government affairs.
In his Congressional advocacy roadmap Announced in August, the company said it believes that while “AI can and should play a critical role in advancing healthcare and driving innovation,” it also believes that “AI cannot and should not disrupt the practice of medicine to replace.”
However, connecting underserved patients to clinical trials is an area that AI can support and refine, the company said.
“The healthcare industry in general is constantly seeking high-quality data, but the data must be standardized and connected, representative of large, diverse communities, and reflect the state of clinical practice in the general population,” Ingram said.
Such data can lead to finding the most suitable patients for real-world research and clinical trials, said Denise Juliano, group vice president of life sciences.
“This structured and unstructured data and the use of AI, machine learning and natural language processing can complement our research work,” she said.
Working with unstructured data
With both structured and unstructured data, Premier has accelerated its efforts to get “the right patient to the right studies,” Juliano said.
With structured data, the company uses predictive analytics and AI to accelerate the work of clinical trials. But she said working with unstructured data – stories – is an area where AI can have a real impact. “Because this is very, very powerful information.”
Discharge summaries, conversations with doctors and patients are often rich in information that can help advance care for those patients, Juliano says. Use cases in radiology and histology show what AI can do for healthcare.
“It will never replace clinical practice, but it can help to really innovate and complement practice,” she said before diving into the first use case.
Analysis of lung scans from the COVID era
Based on the belief that a large number of patients would receive lung scans during the COVID-19 pandemic, the company worked with the nation’s largest health information exchange and partners in the life science industry, on AI, ML and NLP techniques.
She said researchers initially assumed they would find between 10,000 and 20,000 patients with indications of early-stage lung disease.
They found that 152,000 patients had pulmonary nodules in their lungs, Juliano said, noting that the analysis for 6.3 million patients took three weeks.
“Without this technology, the human eye or person would never be able to browse through 60 million documents at that high speed,” she said. “So this was really revolutionary. It was the largest observational trial of its kind.”
Nodules are often referred to as an incidental finding, an “incidental lung nodule,” but the literature suggests that one-third may have lung cancer, Juliano said.
“We could actually help find patients early on who might have lung cancer,” she explained. “It offers the opportunity to identify patients early in their potential disease state and allow them to develop a care plan.
For phase two, Juliano said Premier is working with a number of New York health care systems to bring in patients, identify them and develop a treatment plan.
“In this particular study, Black Americans made up 12% of the study population, which again is a lot higher than you would otherwise have found with the human eye for this work,” she said.
Predicting Alzheimer’s disease, improving dementia care
In the following example, Premier used both structured data from more than a million patients across 275 locations and unstructured data in early patient identification for Alzheimer’s disease and dementia.
What she said Premier calls data ontology helps predict whether a patient will reach a final diagnosis.
The use of predictive analytics and unstructured data can support early intervention in early mild cognitive impairment, Juliano explains. The AI ??helped “find the needle in the haystack” and at a later stage established the terms that could be indicators of diseases.
They also use these techniques to accelerate ongoing clinical trials “because 80% of US clinical trials fail to meet their recruitment deadlines,” she said, noting that it takes 90 months to complete a phase three or complete phase four study.
Cost also plays a role here: Bringing a new drug to market costs more than $2.6 billion, she said. Delays in the process could cost “$600,000 to $8 million per day.”
AI is a powerful way to “get the right people using the products that are right for their care,” says Juliano.
Diversification of clinical trials
Black Americans, Latinos or Latinx tend to be underrepresented in clinical trials, she said, so they developed a model to improve clinical trial diversity.
If you think about the traditional model, “they usually try to identify a physician-investigator on site first,” she said. ‘Then they hope that those patients are in healthcare.
“What we’ve developed at Premier is a flip-funnel model,” she explained. “We actually apply the inclusion and exclusion criteria to both our structured and unstructured data. We find out where those patients are, then we recruit the system, and then we recruit the physician researcher,” Juliano said.
Building resilience in the supply chain
Angela Lanning, Chief Operating Officer within Premier’s healthcare informatics division, shared how the company is deploying AI, NLP and ML in healthcare administration and the supply chain.
In 2019, Premier partnered with UPMC to use machine learning to guide purchasing and recommend formula management decisions in real time.
But after the pandemic exposed the stark reality of “how catastrophic supply disruptions are to our ability to provide safe, timely health care to our patients,” the company tapped into its big data and used AI “to track’ for leading indicators – line delays and product shortages.
Once health care systems can predict when a product will be in short supply within three to six weeks, “we can do something about it,” she said.
“This is important because we can do this for more than 80% of the medical and surgical supplies in today’s healthcare system.”
She said the company is working with healthcare systems and suppliers to mitigate shortages, either through a supplier increasing production or by helping healthcare systems identify alternatives they can use to minimize disruptions to the patient care supply chain .
End-to-end prior authorization automation
A Council for Healthcare Quality survey last year found that only 28% of prior authorization activities were entirely electronic, Lanning said.
When a health plan rejects a physician-recommended procedure, “you end up in a circular loop of the physician and the payer trying to get this procedure approved.”
Not only is it difficult for patients, “it’s also very manual and inefficient,” she said.
Doctors have to intervene with faxes and forms, or their administrators do. According to Lanning’s presentation, the Centers for Medicare and Medicaid Services estimates that physicians and physician group practices spend an average of 13 hours per week on manual pre-auth activities.
“We don’t have the people in healthcare who would have to take on these non-value-added processes when we have the ability to automate them,” she said.
There is a lot of tension in the prior authorization process, and AI can help address this and strengthen payer-provider relationships, she said.
“It was important for us to figure out how we could work with our healthcare systems and health plans to identify eligible patients, along with the appropriate clinical indications, to approve the procedure in real time,” Lanning said.
In 2014, there would be a mandate that health care systems would be required to use clinical decision support to achieve appropriate use compliance. Although that mandate has been postponed, Premier has “integrated decision support into the electronic health record so that it falls within the physician workflow.”
The evidence is coded in the EPD and is 100% automated.
“Although Medicare did not mandate this, this has helped us with commercial payers in obtaining approval for advanced imaging,” Lanning said.
Premier has also been working with healthcare systems for more than a decade to integrate CDS into EHRs to enable reimbursement for radiology tests.
While the agency requires providers to have significant clinical judgment for payment approval, Lanning said AI has helped automate known patient cases in real time.
“What we’ve done to help providers manage this process is give them only the most complex patients to assess based on human interaction,” she said. “We have also linked that to radiologists on our team so that they can provide the ordering physician with the best test to achieve the best outcome for the patient.”
It completes 30-minute processes in less than a minute, she said, and patients get a much better experience.
Premier, together with HIMSS (parent company of Healthcare IT news) and others, is part of the Patient ID Now coalition. In 2021, the organization introduced a national patient identity framework, which it said could streamline the flow of patient information, but there has been little legislative movement.
“It will also remove barriers to care coordination and nationwide interoperability, and save millions in related costs to the health care system,” said Blair Childs, Premier Inc.’s senior vice president of public affairs.