Becker's Hospital Review

23 INNOVATION What hospital-at-home leaders can learn from hospitalists By Giles Bruce T o truly scale "hospital at home," its leaders should take a page from hospitalists, a similarly disruptive care model that grew exponentially, two physician leaders wrote in Health Affairs. Hospital-at-home physicians also need a catchy name, a la hospitalists, the authors said in the March 26 article. How about "home hospitalists"? "Like hospitalists, HaH providers must define the specific clinical and operational competencies of a home-based site-defined generalist specialist," wrote Robert Zimbroff, MD, associate medical director of San Francisco-based UCSF Health's Care at Home program, and Robert Wachter, MD, professor and chair of the department of medicine at the University of California, San Francisco. "ey would be helped by naming their physician workforce and buttressing the arguments for this being a new specialty." Hospital-at-home programs can also be expensive to set up, with command centers costing upward of $5 million and an extensive supply chain network required, while continued reimbursement from CMS is uncertain, the authors noted. So leaders also need to show the business value to their health systems' executive teams. "In making their business case, hospitalists convinced hospital and healthcare executives that it would be more expensive not to implement hospital medicine programs than to build and support them," Drs. Zimbroff and Wachter wrote. "For HaH to thrive, it must do the same." For example, they said, program leaders could explain how home- based care could free up hospital beds for "higher-acuity patients reimbursed at higher rates," with the increased revenue from the "bed arbitrage" being invested back into "hospital at home." n Why healthcare AI treatments cost more By Giles Bruce M ammograms read by artificial intelligence don't have a billing code so radiologists may pass the extra $40 to $100 in out-of-pocket costs on to patients, The New York Times reported. The financial ramifications are just one angle patients — and providers — must consider when deciding to use healthcare AI. Others include how well the technology works among a diverse patient population and whether it can improve disease survival rates, according to the April 8 story. "There's a need for more diverse training and testing of these AI tools and algorithms in order to develop them across different races and different ethnicities," Katerina Dodelzon, MD, a breast imaging radiologist at New York City-based NewYork-Presbyterian, told the newspaper. "AI is just a tool that learns based on what it sees." Some of the approximately two dozen AI mammography tools approved by the FDA are either being tested by hospitals or offered at a small number of clinics, The Times reported. Health systems that aren't charging patients for the added price tag of AI mammograms might be absorbing the costs themselves or using the tools strictly for research purposes at this point, according to the story. n How AdventHealth is advancing responsible AI By Naomi Diaz A ltamonte Springs, Fla.-based AdventHealth is par- ticipating in two initiatives that focus on implement- ing and developing artificial intelligence responsi- bly and safely. AdventHealth joined the Coalition for Health AI (CHAI), a nonprofit organization aiming to establish a network of laboratories across the country to test healthcare artificial intelligence tools. The health system is a founding member of the organization, according to an April 17 news release from AdventHealth. Additionally, AdventHealth is collaborating with the Trustworthy and Responsible AI Network (TRAIN). This network is dedicated to responsible development and use of artificial intelligence in healthcare. Sixteen health systems, including AdventHealth, have signed on to the network and will collaborate with Microsoft, the network's technology partner, and OCHIN, a national network of community health organizations. "By sharing our knowledge and collaborating with other health care leaders through CHAI and TRAIN, we can significantly accelerate the development of AI solutions that will improve patient outcomes and shape the future of healthcare," Rob Purinton, vice president of analytics and performance improvement for AdventHealth, said in the release. n

• Cover