<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/1804a276bb464527b6f08c467280eeb1&quot; frameborder=&quot;0&quot; width=&quot;1760&quot; height=&quot;1320&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>1320</height><width>1760</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>1320</thumbnail_height><thumbnail_width>1760</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/1804a276bb464527b6f08c467280eeb1-f81362db94d26bb5.gif</thumbnail_url><duration>542.072</duration><title>How I Use UpToDate Expert AI in Real Clinical Workflows</title><description>I&apos;ve been using UpToDate since medical school — it&apos;s one of the most trusted clinical decision-support tools we have. Now they&apos;ve added an AI feature, and I wanted to show you exactly how I use it at the bedside, in real time.

In this video, I walk through a real clinical scenario: a 65-year-old woman with newly diagnosed atrial fibrillation, hypertension, and diabetes — admitted as an inpatient. Should we anticoagulate? When? What happens when you layer in bleeding risk, CKD, and fall risk?

UpToDate&apos;s Expert AI walks through the reasoning, shows its assumptions, calculates CHA2DS2-VASc, flags HAS-BLED considerations, and — importantly — tells you when it doesn&apos;t have enough data to answer. That last part is something most consumer AI tools get wrong.

This is the workflow I covered in my Healthcare Huddle newsletter. If you haven&apos;t read it yet, the link is below.

[http://healthcarehuddle.com/p/how-uptodate-expert-ai-builds-trust-at-the-point-of-care]</description></oembed>