Newsbytes

Editors: Raymond D. Aller, MD & Dennis Winsten

AI for clinical decision support: from progress to predictions

November 2025—Following the announcement of Wolters Kluwer’s new generative artificial intelligence-powered clinical decision support tool, UpToDate Expert AI, Robert Wachter, MD, professor and chair of the Department of Medicine, University of California, San Francisco, assessed the field of AI clinical decision support in the article “Medicine’s AI knowledge war heats up,” published on Substack (substack.com/home/post/p-174968874). Here is an edited excerpt of what he had to say.

UpToDate remains an extraordinary resource, created by over 7,500 human experts charged with culling and interpreting the medical literature and guidelines to produce chapters on every conceivable clinical topic—and keeping the chapters, as the name says, up to date. From the time it emerged in the late 1990s until about 2023, it deservedly had what felt like an unshakeable position as the dominant point-of-care tool for health systems and clinicians.

And then, OpenEvidence displaced UpToDate—particularly among doctors in training—because it could take an entire clinical case, in all its staggering complexity, and produce an AI-generated “curbside consult” that was impressively accurate, context specific, and, yes, up to date.

Two very different approaches
Let me demonstrate what I mean. Here’s a relatively complex case I saw a couple of years ago when I was a visiting professor at Yale. I’ll present it briefly, as I might if I had run into my favorite hepatologist in the hospital cafeteria and asked her for a curbside consult:

“I’m caring for a 75-year-old man with Waldenstrom’s macroglobulinemia who came in with fever, hypoxia, and pulmonary infiltrates. We started him on Zosyn, and he rapidly developed liver dysfunction, with transaminases in the thousands but normal alk phos and bili. What do you think is going on?”

When I entered my long prompt in UpToDate, it produced links to four of its chapters:

• Approach to the immunocompromised patient with fever and pulmonary infiltrates
• Epidemiology, pathogenesis, clinical manifestations, and diagnosis of Waldenstrom macroglobulinemia
• Sepsis in children: definition, clinical manifestations, and diagnosis
• Evaluation and management of fever in children and adults with sickle cell disease

The first two are off point. The last two links, to pediatric chapters, are bizarre [choices] since my query clearly concerned a 75-year-old man.

Recognizing that I was asking UpToDate to do something it isn’t built to do (grapple with a complex case rather than a narrowly framed problem), I tried deconstructing my query into a single focused question: “Can Zosyn cause liver injury?” This time UpToDate took me to its chapter on Zosyn (an intravenous antibiotic), which was fine but didn’t answer my question. Several more links followed, most of them clinically irrelevant. The most on-point link was seventh on the list: a chapter on acute liver injury.

In contrast, when I put my original prompt into OpenEvidence, it “thought” for about 10 seconds and then gave me an impressive answer that considered the entire case presentation. First, it reviewed the most likely diagnoses, which it considered to be ischemic hepatitis (“shock liver”), drug-induced hepatocellular injury, and sepsis-associated liver injury. It then offered several other “most important not to miss diagnoses,” including Tylenol overdose, Budd-Chiari syndrome (hepatic vein thrombosis), and fulminant viral hepatitis. These two lists mirror the way experienced clinicians approach clinical cases—we often create lists of both the most likely diagnoses and those that might be fatal if missed.

[At CAP TODAY press time,] UpToDate said it would soon roll out its own AI-based tool, called UpToDate Expert AI. Note the careful branding, designed to highlight the fact that the UpToDate tool won’t scour the entire medical literature or the unfiltered Internet for insights. Instead, UpToDate’s AI will draw its wisdom exclusively from its thousands of continuously updated chapters, written by experts.

What is the optimal source of truth?
This raises a question: Now that UpToDate is adding genAI capabilities, which tool will provide better results? Will it be OpenEvidence, whose AI reviews the world’s medical literature via searches of journals and society guidelines and then applies genAI to create answers and references? Or will UpToDate’s more human-crafted approach, in which the AI limits its source of truth to the chapters in the UpToDate data set, produce better results?

Finally, there’ll be another competitor, one that will approach decision support from a very different angle. Epic, the largest electronic health record vendor, also announced a new set of AI tools. One of them, named Art for Clinicians, is designed to review individual patient records as well as Epic’s database (Cosmos) of deidentified records on millions of patients cared for in health systems that use the company’s EHR.

Over time, one assumes Epic will be able to mine two sources of data unavailable to UpToDate, OpenEvidence, or any of the general large language models—the diagnostic and treatment strategies of tens of thousands of clinicians that use Epic’s EHR and the clinical courses of millions of patients in Cosmos.

The promise of EHR integration
The AI tools being adopted in health care today—such as digital scribes, AI chart review, and prior authorization assistance—are helpful, but the real impact and savings will be in clinical decision support, particularly once we figure out how to tightly integrate it into the EHR.

This last point is crucial. Currently, although UpToDate is a click away on most clinicians’ Epic, Athena, or Oracle desktops (increasingly, OpenEvidence is as well), these tools aren’t integrated into EHRs in any meaningful way. Clinicians interested in the best treatment for high blood pressure or the proper tests to rule out lymphoma need to leave their EHR workflow and conduct a search of UpToDate or OpenEvidence for the answer. And that search won’t know anything about the patient beyond what the clinician types into the search box.

Imagine a world in which the clinician doesn’t need to enter, “This is a 75-year-old man with Waldenstrom’s who presents with fever and pulmonary infiltrates and develops elevated liver tests,” but rather one in which AI is reading the patient’s chart automatically so that it “knows” all that contextual information. The AI tool might suggest—and perhaps even tee up—recommended diagnostic tests or treatments based on its review of the patient’s data. It might even allow the clinician to ask very specific questions (“Given when the patient received the first dose of Zosyn, what are the chances this is drug-induced liver injury?”) based on the patient’s data.

Assuming the AI produces trustworthy insights and recommendations, this kind of EHR integration has the capacity to transform the practice of medicine. Of course, the source of the data it uses for the search will ultimately determine whether the results are, in fact, trustworthy.

The Substack article, “Medicine’s AI knowledge war heats up,” is adapted from the upcoming book A Giant Leap: How AI is Transforming Healthcare and What that Means for Our Future, by Robert Wachter, MD, to be published by Portfolio, an imprint of Penguin Publishing Group, a division of Penguin Random House LLC. Copyright © Robert M. Wachter, 2026.

Voicebrook partnerships create speech-enabled digital path ecosystem

With Voicebrook’s integration of its VoiceOver Pro pathology reporting solution into digital pathology workflow systems from PathAI, PathPresenter, and Gestalt Diagnostics, the company has created a fully speech-enabled digital pathology ecosystem.

Users of these leading digital pathology platforms can navigate slides, dictate findings, complete CAP protocols, and finalize reports using the speech-recognition technology, according to a press release from Voicebrook.

VoiceOver Pro is integrated with Gestalt Diagnostics’ PathFlow, enabling voice control of image navigation and reporting; PathAI’s AISight, allowing pathologists to dictate and capture structured data from digital slides; and PathPresenter’s PathPresenter system, providing voice-based slide navigation and report generation.

Voicebrook, 516-326-9400

Epic and Quest collaborate

Epic has announced plans to deliver a suite of technologies and services to Quest Diagnostics that will serve as the foundation for Quest’s Project Nova, a multi-year initiative to simplify and enhance several of the company’s customer-facing processes, regardless of the electronic health record platform used.

The implementation will provide Quest’s national laboratory operations with access to the breadth of Epic’s diagnostic enterprise system for laboratories, which includes the Aura data-exchange network, Beaker Laboratory lab information system, Care Everywhere health information exchange platform, MyChart patient portal, and Resolute Professional Billing and Claims.

“From creation of a single comprehensive test list to a better patient experience with MyChart and more, this is an opportunity to transform the health care experience at a national scale,” said Epic president Sumit Rana, in a company press statement.

A cross-organization team experienced in large-scale implementations will lead the years-long phased project.

FujiFilm and Precidx team up

FujiFilm Healthcare Americas and Precidx have reported plans to integrate FujiFilm’s Synapse Pathology picture archiving and communications system with Precidx’ optimization platform to reduce digital pathology file storage.

“A key barrier in the wider spread adoption of digital pathology is the significant cost of storing multi-gigabyte whole slide digital images,” said John Cupp, MD, director of digital pathology, Hoag Hospital, Newport Beach, Calif., in a FujiFilm press release. “The POP [Precidx optimization platform] has achieved a great milestone in addressing this concern, all while maintaining diagnostic reproducibility.”

Hoag Health System, a large nonprofit regional health care organization in Orange County, Calif., conducted clinical research and validation on Synapse Pathology combined with Precidx’ optimization platform, which demonstrated “significant reductions in file sizes, while maintaining requisite image quality,” according to the press release.

LabWare revamps system

LabWare has released Clinical Health Solution 5.06, the latest version of its system for managing the operations of clinical diagnostics and public health laboratories.

Among the enhancements available via this update are the following:

• an out-of-the-box laboratory billing module as a standard feature
• an enhanced result review dashboard that provides streamlined result verification and improved user interfaces
• refined send-out functionality for more efficient tracking and management of external lab requests
• advanced contact management for documenting and managing customer communications and resolving issues
• a redesigned master data-management dashboard for more granular control over master data elements, such as test catalogs and submitters
• reorganized and ready-to-use role definitions for lab administrators managing access control and user permissions
• smarter search functionality, including improved template-driven data-visualization capabilities
• additional support for HL7 interface workflows that enables automatic flagging of tests that are added or canceled during order processing

“This version was released utilizing LabWare’s comprehensive automated testing tool, which allowed for a more rigorous and expanded validation testing footprint and will streamline testing for future versions,” said LabWare’s director of clinical solutions, Ed Krasovec, in a company press release.

LabWare, 302-661-2345

Proscia extends partnership with Amazon Web Services

Proscia has expanded its collaboration with Amazon Web Services by integrating its Concentriq platform for pathology with Amazon’s AWS HealthImaging to provide laboratories and life sciences organizations with a cloud-native foundation that supports AI-driven precision medicine initiatives. AWS HealthImaging is a service for storing, analyzing, and sharing whole slide images in the cloud at petabyte scale.

With Concentriq, AWS HealthImaging clients can access whole slide images in the cloud with subsecond time-to-first-image, enabling high-throughput workflows. Pathology data stored in AWS HealthImaging can also be used to develop artificial intelligence and by existing AI applications.

Proscia, 215-608-5411

LigoLab integrates PathPresenter clinical viewer into platform

LigoLab and PathPresenter have entered into a strategic partnership under which PathPresenter’s FDA 510(k)-cleared clinical viewer has been integrated into LigoLab’s all-in-one laboratory information system and revenue cycle management platform, providing LigoLab clients with an end-to-end digital pathology workflow.

“By integrating the PathPresenter viewer with our LIS platform, we’ve created a unified environment where pathologists can move effortlessly from case data to high-resolution whole slide images,” said LigoLab CEO Suren Avunjian, in a company press release.

LigoLab, 800-544-6522

Gestalt Diagnostics updates digital pathology system

Gestalt Diagnostics has released version six of its PathFlow digital pathology platform.

This latest version provides a robust visual interface that allows pathologists to navigate images with greater precision, speed, and clarity, thereby supporting complex case reviews and high-volume workflows, according to the company.

The solution also includes a tumor board module that supports real-time collaboration across pathology, oncology, and radiology.

Gestalt Diagnostics, 509-492-4912

Dr. Aller practices clinical informatics in Southern California. He can be reached at [email protected]. Dennis Winsten is founder of Dennis Winsten & Associates, Healthcare Systems Consultants. He can be reached at [email protected].