|
AMIE Prime Time, Heidi Hardware, and FC Most Innovative Companies
March 26, 2026
|
|
|
|
|
Together with
|
|
|
|
“If 2025 was about simulated benchmarks for medical AI, 2026 is about real-world prospective studies with actual patients.”
|
|
Google Research Scientist Mike Schäkermann
|
|
|
|
The gap between benchmark scores and real-world performance has been the theme of the year in AI research, so Google was right on cue with its first prospective clinical trial for AMIE using actual patients.
Meet the Articulate Medical Intelligence Explorer. AMIE is Google’s flagship “medical AI researcher,” and it teamed up with Beth Israel Deaconess Medical Center to gauge performance in real clinical workflows.
- 100 patients completed an AMIE interaction before their primary care visit, with AMIE taking medical histories and equipping patients with potential diagnoses to discuss with their PCP.
- PCPs received the transcript, summary, and AMIE’s management plan prior to the visit. All interactions were monitored live by physicians trained to intervene if safety criteria weren’t met.
AMIE got a gold star. Not only were there zero safety stops across all 100 interactions, patients reported that their attitudes toward AI significantly improved after chatting with AMIE.
- AMIE’s differential included the correct final diagnosis in 90% of cases (per chart review 8 weeks post-encounter), with 75% top-3 accuracy.
- PCPs using AMIE reported increased visit preparedness in 75% of cases, as well as potential behavior change in nearly 60%.
- The quality of AMIE’s differential diagnosis and management plan appropriateness was similar to PCPs, although PCPs won on management plan practicality and cost-effectiveness.
Other findings were less obvious. PCPs had the chart, the physical exam, and the pre-visit transcript, yet AMIE still matched them on differential quality and management safety without taking a single peak at the EHR.
- That speaks to the ceiling (or lack there-of) for structured AI history-taking, and shows that AI is gearing up to improve patient care in more ways than just making predictions.
- The fact that PCPs reported better visit preparedness and potential behavior change in over half of cases also highlights how AI can augment – not just replace – clinical reasoning.
The Takeaway
The distance between the bench and bedside is getting shorter, and Google’s AMIE results suggest that conversational AI in primary care is closer to reality than most people might think.
|
|
VBC 202: Intro to VBC and Risk Adjustment
Navina and Out-of-Pocket teamed up to create VBC 202, a three-part online course for anyone looking to build or scale value-based care programs. Watch the first session on-demand to learn why VBC emerged, how policy shapes provider behavior, and the fundamentals of risk, quality, and financial performance. Gary Pilling II joins as the guest instructor to dive into risk adjustment basics, including RAF scoring, HCC coding, and the CMS-HCC V28 model.
|
|
Scale RPM With BPM Pro 2
BPM Pro 2 is the next generation of cellular blood pressure monitors, empowering care teams to scale remote patient monitoring and streamline operations. Discover why leading providers are choosing BPM Pro 2 to collect highly precise measurements and enrich data with Patient Insights from their daily lives.
|
|
- Heidi Hardware: Heidi just went where few (if any) clinical AI software companies have gone before: clinical AI hardware. The launch of Heidi Remote gives clinicians reliable audio capture without relying on a phone or a laptop, which should be a welcome capability in settings where hands are in high demand like the ED. Heidi Remote transcribes patient sessions offline, stores them securely on the small clip-on device, then syncs to Heidi whenever the user is back in range. Not a bad way to follow up a flurry of conference announcements.
- Death, Taxes, Consolidation: Few things are certain in healthcare, but hospital mega-mergers are one of them. California-based Sutter Health is acquiring Allina Health, and the California-based system pledged to invest $2B in Allina’s Minnesota and Wisconsin operations. Sutter said that combining its AI platform development with Allina’s expertise in med-tech and engineering will position it as a national leader in improving patient and provider experiences through new tech. The combined organization will account for $26B in revenue across 39 hospitals, 88k employees, and 400 care sites.
- FC Most Innovative Companies: Fast Company released its annual Most Innovative Companies List, highlighting businesses across all industries “that are doing things differently.” Healthcare surprisingly only had two companies rank in the top 50 (grats to Abridge and Cresilon), and if you zoom in on just the AI companies Abridge ranked third overall behind a couple of tiny startups called Google and Anthropic. The full list is definitely worth checking out even if just for its great startup overviews hiding behind the article icon next to their names.
- Overly Agreeable Models: New research out of Stanford was raising eyebrows after showing that clinicians can easily steer LLMs toward harmful decisions. The study evaluated 21 top reasoning models (including GPT-5, Claude Opus 4.5, Gemini 3) on differential diagnosis and next-step recommendations using 61 NEJM cases and 92 real-world clinician-AI interactions. Adversarial clinician input (a plausible-but-wrong differential, no correct final diagnosis) caused a significant drop in diagnostic accuracy for 14 of the 21 models, with an average impact of -5.4 percentage points.
- Turquoise Series C: Turquoise Health hauled in $40M of Series C funding to become the operating system for healthcare contracts and payments. The platform has come a long way since we first covered Turquoise while the ink was still drying on CMS’ original transparency requirements, and it’s now taking a two-pronged approach to dissolving administrative waste. Clear Rates synthesizes files, claims data, Medicare benchmarks and other relevant pricing signals into a single rate for every payor-provider combination. Contracts then leverages AI to tag every rate and provision across a customers’ contract portfolio to help quickly find contract details, compare performance, and model negotiation scenarios in a few clicks.
- PCPs Are Key for Chronic Care: A Milbank Memorial Fund report underscored the importance of having a usual source of care, particularly when it comes to managing chronic diseases. Using data from the Medical Expenditure Panel Survey, the report found that nearly all adults (95.5%) with a regular PCP received preventive services for chronic conditions, compared to just 67.6% of those without a regular PCP. The starkest contrasts were seen with blood pressure checks (95.5% vs. 67.6%), cholesterol screening (89.2% vs. 56.7%), mammograms (83.8% versus 54.2%), and colon cancer screening (79% versus 59.5%).
- Reimagine Care Remi Upgrades: Virtual oncology provider Reimagine Care debuted “a major evolution” of its Remi assistant, including deeper AI-powered patient understanding and expanded safeguards. Remi is designed to support cancer patients between visits and help oncology care teams respond to treatment concerns earlier. The AI overhaul gives Remi the ability to analyze complex messages (like when patients report multiple symptoms at once) and translate them into structured insights that help route patients through appropriate clinical pathways.
- GLP-1s Post-MI: The GLP-1 revolution has officially moved into the acute care setting. An analysis of over 280k patients without diabetes showed that initiating tirzepatide (Zepbound/Mounjaro) within 14 days of a heart attack or stroke significantly improves two-year clinical outcomes. While the study found no difference in total major adverse cardiac events, patients on early tirzepatide therapy experienced a significantly lower risk of heart failure hospitalization, acute kidney injury, and recurrent ischemic stroke.
- Parallel Lands $20M: Parallel hauled in $20M of Series A funding to tackle administrative waste with AI agents (definitely the flavor of the year for young digital health startups). Instead of requiring hospitals to replace existing systems, Parallel runs as an AI layer on top of them, navigating software exactly like a human would. That enables it to quickly automate repetitive tasks across coding, billing, and admissions, while also allowing it to be up and running within a week.
- Deepfake Images Fool Radiologists: It’s no secret it’s getting harder to tell the difference between real and AI-generated content – and that extends to medical imaging. In an experiment in Radiology, a panel of 17 radiologists had trouble differentiating real radiographs from AI-generated deepfakes. When presented with a set of 154 images, only 41% recognized that some were synthetically created. When asked to differentiate real images from fakes, they had mean accuracy of 75%, illustrating the challenge that could occur in identifying malicious use of deepfakes.
|
|
State of Payor Enrollment and Credentialing
AI is changing the way that healthcare leaders tackle provider network management. Medallion’s latest report breaks down the biggest challenges, emerging trends, and how automation is transforming the landscape. Get the insights you need – read the full report today.
|
|
The Virtual-First Difference at MetroHealth
When the MetroHealth System needed a scalable solution to help its patients access care, it turned to Ovatient’s virtual-first care platform. Ovatient allows its partners to offer fully coordinated telehealth - including urgent care, primary care, and integrated behavioral health – all built on Epic. Discover how MetroHealth’s virtual-first approach is keeping its patients connected to high-quality care whether they’re at home, at work, or on the go.
|
|
- Medical Transcription That Gets Terminology Right: AssemblyAI’s clinical speech-to-text model was purpose-trained on medical terminology, pharmaceutical names, and industry acronyms. That means it performs the best where it counts the most. 66% fewer missed medical terms. BAA + SOC 2 Type II. Real clinical environments. See why AssemblyAI is the trusted Voice AI infrastructure for healthcare solutions.
- Abridge & Availity Redefine Payer-Provider Synergy: Abridge is teaming up with Availity to redefine payer-provider synergy at the point of conversation. The collaboration aligns Abridge’s evidence-aware intelligence with Availity’s real-time health information network to create a first-of-its-kind prior authorization experience, with a shared understanding between patients, providers, and payers. Find out how Abridge and Availity are extending conversational intelligence across the revenue cycle.
|
|
|
|
|
