Enhancing Clinical Decision Support (CDS) with AI and FHIR®

You can love or hate it, but you can’t ignore it.

That’s AI today. 

One thing I’ve learned from working in healthcare is that FOMO is not the right motivator. So, as the hype grew, instead of chasing headlines, I decided to explore what AI really means for healthcare. I started by talking to our existing customers, doing the research, asking the hard questions, and even going back to school (yes, after all these years) to learn the fundamentals. I wanted to know: is this real or just hype?

Spoiler: Turns out, it's math, not black-box magic.

During the initial research, it quickly became clear that most AI solutions in healthcare are swimming in the same red ocean - focused on documentation and billing. The irony? These are problems created by the system itself. Now, we’re throwing AI at them to make the mess more manageable. That's not the kind of impact we were looking to make. It might streamline the workflows, but it won’t improve healthcare, at least not directly. Rather than try to boil that already-red ocean, we looked for a blue one - an area with real clinical potential and insufficient attention: leveraging AI for Clinical Decision Support (CDS). 

 Over the past few months, our team at Darena Health has had countless conversations (in-person and virtual) with existing customers, EHRs, healthcare organizations, developers, model builders, and industry leaders. The more we talked, the more we realized how much was missing from the conversation: actual evidence (pun intended) of how AI can change clinical decision-making like evidence-based does today. What's worse, the discussions seemed to have gone from solving documentation problems to replacing doctors and completely skipped the idea of supporting clinicians.

 That’s when we felt the need for a structured approach to bringing the focus back to how clinicians practice medicine. The goal was to develop a framework that aligns with how it is done today and how AI can support it, NOT override it.

That’s how the CDS STAT™ framework was born that we shared in a recent post.

In short, it’s a structure for responsibly aligning AI with evidence-based CDS. It's not a product; it is a framework for anyone to implement. It may not be perfect. It’s evolving (feedback is welcome). But it’s a start. It was shaped by feedback from a wide cross-section of the healthcare community. 

Now, we’re taking it to the field.

In this newsletter, I’ll share that journey. I'll talk about how we’re applying the framework, learning from real-world implementations, and continuing to refine it. It’s our way of bringing structure to something that’s evolving quickly while keeping the focus where it belongs: on supporting clinicians and improving care.

What you can expect

🔥Lot of FHIR - yes, we’ll get geeky. Our core focus is CDS on FHIR. From CDS Hooks to SMART apps to CQL to backend services. It’s where we think the most meaningful impact is possible. But the CDS STAT™ framework goes beyond just FHIR, and we’re always open to ideas, feedback, and collaboration.

🧠 Conversations I’m having with customers, partners, and clinical leaders

🔍 Insights from the field - things we learn from our team, customers and the community

🔧 Implementation stories from what we’re building at Darena Health

🎥 Hands-on videos and walkthroughs (because I’m from the Show-Me State)

👨💻 Dev talks with engineers building for CDS + AI + FHIR

🗞️ Post and comments recaps - in case you missed something (or ignored it the first time)

📦 Flashbacks to past work - yes, there might be reruns, but I promise to keep it worth your scroll

About the frequency

LinkedIn made me pick weekly. At the pace AI is moving, weekly feels like dial-up internet in a fiber world. So, let’s call it “as it makes sense.”

No spam. No “rise and grind” energy. Just thoughtful, real-world updates when there's something worth sharing. Sounds good?

If you’re building, designing, or just thinking about how AI can support evidence-based CDS, welcome to this space. I’d love your feedback, questions, and, most importantly, thoughtful disagreements.

Let’s build something better together.