Amidst a worker shortage, the healthcare industry has been exploring generative AI and its potential to eliminate mundane, burdensome tasks. On Wednesday, two established organizations — Google Cloud and Mayo Clinic — announced a partnership focused on the technology.
Mayo Clinic will be piloting a new HIPAA-compliant Google Cloud service called Enterprise Search on Generative AI App Builder — Gen App Builder for short. The aim of the partnership is to make it easier for clinicians to search seemingly boundless repositories of patient data.
With the Rise of AI, What IP Disputes in Healthcare Are Likely to Emerge?
Munck Wilson Mandala Partner Greg Howison shared his perspective on some of the legal ramifications around AI, IP, connected devices and the data they generate, in response to emailed questions.
For clinicians, gathering the information they need to take care of patients can be quite time-consuming and onerous. These pieces of information — such as clinical guidelines, research papers and patient data in the electronic health record — are often all stored in different places using various formats.
Gen App Builder enables providers to create a search system for their data, equipped with conversational features powered by Google’s large language models. The service is designed to harmonize data, as well as make it easier to interpret, said Vish Anantraman, Mayo Clinic’s chief technology officer, in an interview.
Mayo Clinic expects that the tool will unlock sources of information that clinicians typically aren’t able to easily search — such as imaging, complex medical histories, labs and genomics. Getting these insights more readily could lead to faster treatments and higher patient satisfaction, Anantraman pointed out.
For example, say a physician wants to see information about Black maternity patients ages 30 through 40, including their insurance status and medical history. Using the tool’s conversational AI abilities, the physician can ask Gen App builder to yield the information. The tool can also let physicians know which clinical trials a patient could be a good match for, Anantraman added.
A Personalized Approach to Medication Nonadherence
At the Abarca Forward conference earlier this year, George Van Antwerp, managing director at Deloitte, discussed how social determinants of health and a personalized member experience can improve medication adherence and health outcomes.
Anantraman said Mayo Clinic will test various use cases for Gen App Builder in the coming months, beginning with use cases that aid clinicians with simple administrative tasks that cause burnout. For instance, the tool can help with clinical documentation and gathering patients’ billing information to submit to payers.
In addition to exploring administrative uses cases, Mayo Clinic will see whether the service can be helpful for clinical use cases, but Anantraman noted it will have to be “very, very thoughtful” about this. Though it is rare, generative AI services like ChatGPT can occasionally provide inaccurate or misleading information. Because of this, the health system doesn’t want to dive into clinical use until it knows the AI model is ready.
“We have to create the right set of evaluation metrics before we even talk about that to ensure that the accuracy and the appropriateness are measured more than anything else,” he explained.
This partnership is part of a 10-year collaboration between Mayo Clinic and Google Cloud, which they announced in 2019. Google Cloud will be working closely with the health system to hear its clinicians’ feedback as they test Gen App Builder, but Mayo Clinic isn’t the only provider that can use the tool — it’s available for other health systems to adopt as well and is being sold on a SaaS model, a Google Cloud spokesperson said.
Google Cloud’s creation of this healthcare data search service, as well as its deployment at Mayo Clinic, is a response to escalating competition with Microsoft and other major tech companies to attract health system customers in search of the most advanced AI solutions. Only time will tell whether this tool will deliver sufficient time savings to justify its adoption.
Photo: metamorworks, Getty Images
