Google Cloud launches AI search tool to help medical staff quickly access clinical information

Google Cloud announced on Monday the launch of a new artificial intelligence search feature that will help healthcare professionals extract accurate clinical information from different types of medical records.

The healthcare industry has a wealth of valuable information and data, but doctors often struggle to find it because it is typically stored in multiple systems and formats. Google Cloud's new search tool will allow doctors to extract information from clinical notes, scanned documents, and electronic health records for easy access in one place.

The company said the new feature will ultimately save healthcare professionals a significant amount of time and effort.

"Such a search tool can save time, prevent frustration for doctors, and ensure that they can find answers more easily," said Lisa O'Malley, Director of Advanced Product Management for Google Cloud AI, in an interview.

For example, if a doctor wants to know a patient's medical history, they no longer need to read notes, faxes, and electronic health records separately. Instead, they can search a question like "What medications has this patient used in the past 12 months?" and see relevant information in one place.

Google's new search feature can also be used for other critical applications, such as applying the correct billing codes and determining if a patient meets the criteria for a clinical trial, according to O'Malley.

She also added that the technology can cite and link to the original sources of information, which come directly from an organization's own internal data. This will help alleviate doctors' concerns about AI potentially generating illusions or providing incorrect answers.

These search features are particularly valuable for healthcare professionals who are already facing manpower shortages and heavy paperwork.

A study funded by the American Medical Association in 2016 found that for every hour doctors spend with patients, they need to spend an additional two hours on administrative work. The study also revealed that doctors spend an extra one to two hours on paperwork outside of working hours, which many professionals refer to as "pajama time."

According to a survey by Medscape in January 2022, 53% of doctors reported feeling burned out, an increase from 42% in 2018.

Google hopes that its new search feature will reduce the time doctors need to spend reviewing additional records and databases.

"Google can apply our search technology, healthcare expertise, and research capabilities to make the journey of clinical doctors, healthcare providers, and payers faster, more efficient, and cost-saving. I believe this ultimately benefits us as patients as well," said O'Malley.

These new features will be available to health and life science organizations through Google's Vertex AI Search platform. Companies from other industries have already been able to use the platform for searching public websites, documents, and other databases. The specific healthcare functionalities are built on top of Google's existing Healthcare API and Healthcare Data Engine products.

Aashima Gupta, Global Head of Healthcare Strategy and Solutions at Google Cloud, said that the new Vertex AI Search feature can be seamlessly integrated into the workflows of clinical professionals, which is crucial for customers in the industry.

The healthcare industry has been skeptical about adopting new technologies, and adoption can be even more challenging if healthcare professionals find new solutions distracting or difficult to use. Gupta said that Google has been closely paying attention to this.

"These are the workflows that doctors and nurses live day in and day out. We can't add friction to it," she said. "We are very mindful of making sure that we respect the platforms they use, ensuring that the workflows remain unchanged but can benefit from the power of this technology."

Starting from Monday, customers can apply for early access to Vertex AI Search for healthcare and life science fields. However, Google Cloud has already conducted functional testing with healthcare institutions such as Mayo Clinic, Hackensack Meridian Health, and Highmark Health.

Cris Ross, Chief Information Officer at Mayo Clinic, said that the clinic has not yet used the new Vertex AI search tool in clinical care but has started with administrative use cases.

"We are curious, and we are cautious," he said. "We won't put anything into patient care until we are truly ready."

Ross said that in the future, Mayo Clinic plans to explore how to use the Vertex AI search tool to help nurses summarize lengthy surgical notes, organize complex patient histories, and easily answer questions like "What is the patient's smoking status?" However, for now, the institution will start with small-scale experiments to study where Google's AI solutions are most useful.

Richard Clarke, Chief Analytics Officer at Highmark Health, said that the organization's initial response to the search tool has been "very positive," and the company already has a backlog of over 200 use cases. However, similar to Mayo Clinic, he said the challenge lies in determining where the technology is most useful, building trust among employees, and conducting large-scale deployments.

"We are still in a very early stage, just working with small teams and getting a lot of support, and we are seriously considering this," Clarke said. "We haven't done large-scale deployments yet, but from early indications, this will be very useful and potentially transformative in many cases."

Google Cloud does not access customer data or use it to train models, and the company stated that the new service complies with the Health Insurance Portability and Accountability Act (HIPAA).

Gupta said that as a patient, interacting with the healthcare system can feel fragmented and challenging, so she is excited about how clinical doctors can use Google's new tools to create a more complete picture.

"From my perspective as a patient, it's a process of going from one point to another in healthcare, but it's difficult," she said. "Now, we are at a stage where AI is starting to play a helpful role in these very practical use cases."