Google announces new generative AI search capabilities for doctors

Google Cloud on Monday announced new artificial intelligence-powered search capabilities that it said will help health-care workers quickly pull accurate clinical information from different types of medical records. 

The health-care industry is home to troves of valuable information and data, but it can be challenging for clinicians to find since it’s often stored across multiple systems and formats. Google Cloud’s new search tool will allow doctors to pull information from clinical notes, scanned documents and electronic health records so it can be accessed in one place. 

The company said the new capabilities will ultimately save health-care workers a significant amount of time and energy.

“While it should save time to be able to do that search, it should also prevent frustration on behalf of clinicians and [make] sure that they get to an answer easier,” Lisa O’Malley, senior director of product management for Cloud AI at Google Cloud told CNBC in an interview.

For instance, if doctors want to know about a patient’s history, they no longer need to read through their notes, faxes and electronic health records separately. Instead, they can search questions such as “What medications has this patient taken in the last 12 months?” and see the relevant information in one place. 

Google’s new search capabilities can also be used for other crucial applications such as applying the correct billing codes and determining whether patients meet the criteria to enroll in a clinical trial, O’Malley said.

She added that the technology can cite and link to the original source of the information, which will come directly from an organization’s own internal data. This should help alleviate clinicians’ concerns that the AI might be hallucinating, or generating inaccurate responses. 

The search features will be especially valuable to health-care workers who are already burdened with staffing shortages and daunting amounts of clerical paperwork. 

A study funded by the American Medical Association in 2016 found that for every hour a physician spent with a patient, they spent an additional two hours on administrative work. The study said physicians also tend to spend an additional one to two hours doing clerical work outside of working hours, which many in the industry refer to as “pajama time.”

In 2022, 53% of physicians reported that they were feeling burned out, up from 42% in 2018, according to a January survey from Medscape.

Google hopes its new search offerings will reduce the amount of time clinicians need to spend digging through additional records and databases.

“Anything that Google can do by applying our search technologies, our health-care technologies and research capabilities to make the journey of the clinicians and health-care providers and payers more quick, more efficient, saving them cost, I think ultimately benefits us as patients,” O’Malley said.  

The new features will be offered to health and life sciences organizations through Google’s Vertex AI Search platform, which companies in other industries can already use to conduct searches across public websites, documents and other databases. The specific offering for health care builds on Google’s existing Healthcare API and Healthcare Data Engine products. 

Aashima Gupta, global director of health care strategy and solutions at Google Cloud, said the new Vertex AI Search capabilities can integrate directly into a clinician’s workflow, which is of high importance for customers in the field. 

The health-care industry has historically been more hesitant to embrace new technology, and adoption can be even harder if health-care workers find new solutions distracting or hard to work with. It’s something Gupta said Google has been paying close attention to. 

“These are the workflows that the physicians and nurses live by day in and day out. You can’t be adding friction to it,” Gupta told CNBC in an interview. “We are very cautious of that — that we are respecting the surface they use, that the workflow doesn’t change, but yet they get the power of this technology.”

Customers can sign up for early access to Vertex AI Search for health care and life sciences starting Monday, but Google Cloud has already been testing the capabilities with health organizations such as Mayo Clinic, Hackensack Meridian Health and Highmark Health.

Mayo Clinic is not using the new Vertex AI Search tools in clinical care yet, said Cris Ross, Mayo’s chief information officer; it is starting with administrative use cases. 

“We are curious, we’re enthusiastic, we’re also careful,” he told CNBC in an interview. “And we’re not going to put anything into patient care until it’s really ready to be in patient care.”

Down the line, Ross said, Mayo Clinic is looking to explore how Vertex AI Search tools could be used to help nurses summarize long surgical notes, sort through patients’ complex medical histories, and easily answer questions such as “What is the smoking status of this patient?” But for now, the organization is starting slow and examining where AI solutions like Google’s will be the most useful.

Richard Clarke, chief analytics officer at Highmark Health, said the initial reaction to the search tools at the organization has been “tremendous” and the company already has a backlog of more than 200 use-case ideas. But similar to Mayo Clinic, he said the challenge will be prioritizing where the technology can be most useful, building employees’ trust in it and deploying it at scale.

“This is still very early days, deployed with small teams with lots of support, really thinking about this,” Clarke told CNBC in an interview.  “We haven’t gone big and wide yet, but all early signs say that this is going to be tremendously useful, and frankly, in many cases, transformational for us.”

Google Cloud does not access customer data or use it to train models, and the company said the new service is compliant with the Health Insurance Portability and Accountability Act, or HIPAA.

Gupta said that as a patient, interacting with the health-care system can feel like a very fragmented and challenging experience, so she is excited to see how clinicians can ultimately leverage Google’s new tools to create a fuller picture. 

“To me, connecting the dots from the patient perspective has long been health care’s journey, but it’s hard,” Gupta said. “Now, we are at a point where AI is being helpful in these very practical use cases.”