CS-Blog
Cedars-Sinai Blog

How AI Is Born: Patient Needs Drive Development

Doctor and nurse looking at computer

The ebb and flow of a medical center’s admissions and discharges, sick calls, empty staff positions, open beds, surgery schedule and more creates a complex equation driven by very real human lives and needs.

Jeremy Miller, RN, associate director of Capacity Management at Cedars-Sinai, sees firsthand how these factors create challenges for efficiently placing patients in a unit that meets their specific needs. Recruitment and hiring can be especially tricky—and in an optimistic scenario, it takes about 90 days to recruit and onboard a nurse, Miller says. Some days he wishes managers had a way to optimize the start date of the hiring process to stay ahead of the need—so, he suggested Cedars-Sinai invent one.

"We already know the seasonal trends for the hospital census, the trends for staff needing sick time and the times of year people are likely to leave their jobs," Miller said. “All that information and more is there, but it would take a much stronger mind than mine to put all those pieces together in a way that could make recommendations for when to start hiring staff."


“Our physicians, nurses and other care providers have direct insights into the roadblocks and opportunities in their daily jobs. By bringing them into the discussion about generative AI technologies, they can help develop meaningful solutions to these hurdles."


Miller participated in a Cedars-Sinai ideathon to pitch ideas for how artificial intelligence could improve hospital operations. Rather than relying on outside technology companies to invent AI to meet healthcare needs, Cedars-Sinai is turning to its own experts to help shape these tools. Ideathons are forums which invite clinicians to share their best suggestions on how AI can reshape the hospital experience. The events are not weighed down in programming specifications, hardware or software, machine learning or any of the technical trappings of generative AI.

Instead, they’re impassioned dialogues about the real needs of patients and people who directly care for them.

“Our physicians, nurses and other care providers have direct insights into the roadblocks and opportunities in their daily jobs," said Mike Thompson, Cedars-Sinai vice president of Enterprise Data Intelligence. “By bringing them into the discussion about generative AI technologies, they can help develop meaningful solutions to these hurdles. The result is better patient care, organizational efficiency and tools that help push the frontiers of medical research.”

Clinical and administrative leaders collaborate with AI experts in these sessions. The clinical team members are broken into groups, each with a moderator who has technical expertise. They then brainstorm ideas and decide together which ones to present to the larger group, which includes a panel of senior leaders from throughout the organization. That panel decides which ideas to advance, support and eventually put into practice.



Building AI Strategy

Cedars-Sinai’s Artificial Intelligence Council oversees the ideathons and guides the health system’s AI strategy. That strategy is shaped by three pillars:

  • Planning and investing
  • Transitioning innovation into adoption
  • Supporting the ethical, responsible and scientifically sound use of AI

“I really prefer the term augmented intelligence,” Thompson said. “AI is not a replacement for the doctors and nurses who provide care to people every day. It is a tool that can help them do their jobs more efficiently, assist in research and perform tasks that will give them more time to spend with their patients.”

AI is already a tool used daily at the medical center. It helps predict patient flow—analyzing what kind of beds will be most needed and helping to plan staffing around those needs. It’s successfully been used in numerous research studies, including helping patients and doctors avoid opioid misuse, reading imaging scans and predicting heart disease.



Ideas Into Action

Cedars-Sinai’s first ideathon gathered physicians from specialties such as cardiology, emergency medicine, surgery, OB-GYN and primary care. The next invited nurses, pharmacists, therapists and other staff from Cedars-Sinai; Huntington Health, a Cedars-Sinai affiliate; and Cedars-Sinai Marina del Rey Hospital to pose solutions to daily challenges.

“This is how artificial intelligence tools are born at Cedars-Sinai,” said David Marshall, JD, DNP, RN, senior vice president and chief nursing executive of Cedars-Sinai Health System. Marshall served on the senior leadership panel for the second ideathon.

Staff suggested multiple advances that could be catalyzed using AI:

  • Better predict the length of surgeries to optimize staffing and communication
  • Develop tools to communicate with patients in real time throughout their healthcare journeys
  • Help patients identify and manage health issues before they develop into life-threatening problems
  • Assist with language translation
  • Improve care coordination
  • Manage triage
  • Help with medication management
  • Boost clinical trial recruitment

“We talked about the obstacles between us and providing care that exceeds our own high expectations,” Marshall said. “We voiced concerns about patients who need our help the most. These are problems we’ve passionately devoted our careers to addressing. There’s nothing artificial about it.”