
...

By Alexis Kayser
Healthcare Editor
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
There is no shortage of health care AI applications on the market, including solutions from major brands like Epic and Microsoft. But despite the ever-expanding array of options, there wasn't a commercial AI solution that fit City of Hope's needs.
That's why the Los Angeles-based system—which is one of the nation's largest cancer research and treatment organizations—decided to launch its own generative AI model in late June.
HopeLLM is powered by City of Hope's proprietary data and technology, and works to improve both provider efficiency and patient experience. The tool can assist with patient onboarding, summarizing vast medical records in seconds. It also works to match patients with clinical trials and pull relevant data for research.
Read More
How Cognizant Is Powering Health Care’s Data Revolution3 min read

What AI Really Can Do Now: 6 Lessons for Harnessing Artificial Intelligence18 min read

New Microsoft AI Research Edges Towards ‘Medical Superintelligence’5 min read

The tool has been rolled out across City of Hope's enterprise, including locations in Los Angeles, Chicago, Phoenix and Atlanta. It's been a success so far, Simon Nazarian, the system's chief digital and technology officer, told Newsweek in an exclusive interview on July 1.

...
"I got hugged the other day," Nazarian said, speaking about health care professionals' response to HopeLLM. "We get a lot of accolades [from providers] about the fact that they're not spending as much 'pajama time,' and they have this available at the point of care. It's to the point that they want to get it out there in the hands of everyone."
HopeLLM had already saved clinicians "thousands" of hours during its first week of deployment, and that number continues to rise, according to Nazarian.
While more health systems are leaning on AI-powered tools like ambient scribes and scheduling agents, few have the resources and infrastructure to launch their own platforms.
There was no better option for City of Hope, per Nazarian. The health system was an early adopter of AI, and has been creating and deploying its own AI-driven predictive models since 2017. Those solutions focused on risk identification—rating patients' odds of becoming septic or having surgical complications—while the new app helps make sense of expansive medical records and strengthens the link between different arms of the business.
"Whether in care delivery or research, we want to use AI in order to accelerate discovery to continuously improve our quality of care, and deliver more precise therapies," Nazarian said.
City of Hope onboarded more than 150,000 new patients in 2024. These patients typically have complex medical histories and see a number of providers before reaching the specialty cancer network. It can be daunting for both patients and physicians to parse through their records, which often include 10 to 20 years of test results and visit notes.
A study from the University of Wisconsin found that one in five patients has an electronic health record (EHR) the size of Herman Melville's Moby Dick, more than 206,000 words. Cancer patients tend to have longer, more layered stories than the average patient, hence the need for a tailored oncology AI solution.
The health system's dedicated applied AI and advanced analytics team worked alongside doctors and researchers to leverage all that data and build HopeLLM. It uses multiple commercial and open-source models, and has been integrated into City of Hope's EHR, Epic.

...
While individual providers can use the app to receive a personalized summary of their patient's records, the health system can tap into its analytical features for a longitudinal view of their care journey. According to Nazarian, City of Hope can marry these insights with other data points on social determinants of health and genomics—helping them create precision therapies that account for an individual's entire background.
AI governance protocols were essential throughout the model's construction and are an ongoing focus for the health system rather than a "one and done," he added. Stakeholders from across the enterprise, including physicians, legal and compliance experts, weighed in on the platform's development and continue to monitor it for quality and safety.
"It really is important for us to make sure that [our AI models] have perfect accuracy, with more information coverage compared to the human-generated summaries that have been done in the past," Nazarian said.
HopeLLM has garnered interest from the broader oncology market, particularly from pharmaceutical companies that would like to take advantage of its trial-matching capabilities. City of Hope is exploring avenues to make the tool available to others, and partner with external stakeholders that could help improve it.
"There is no way for us to predict where a revolutionary technology like this is going to go," Nazarian said, "but we are staying close to it and will be at the forefront."
Nazarian will be speaking at Newsweek's Digital Health Care Forum in New York City on September 16. For more information or to register for the event, visit events.newsweek.com/digitalhealthcare.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines