How Context-Aware Decision Support is Reshaping Clinical Practice: Dr. Brian Dilcher of West Virginia University Medicine

In high-pressure care settings like the emergency department, time and information are everything, but in scarce supply. Clinicians must immediately assess problems and make decisions, all while new patients continue to arrive. There is rarely time to consult all of the latest, most-relevant literature and guidance. And new evidence is published at an impossible clip.

Abridge's clinical intelligence provides evidence at the point of care, contextualized to the patient in the room. It leverages the patient conversation alongside information from the electronic health record—including prior notes, labs, diagnoses, and other chart data—to shape the evidence it surfaces during care. No switching tabs. No digging for labs. Just evidence-based insights, tailored to each patient's clinical situation, now implemented enterprise-wide at more than 300 health systems.

We sat down with Dr. Brian Dilcher, an emergency medicine clinician and associate Chief Medical Information Officer (CMIO) for emergency medicine at WVU Medicine’s Ruby Memorial Hospital in West Virginia. As an informatics leader, he decides whether technologies like Abridge are rigorous enough to meet the needs of fellow clinicians and residents. As a practicing clinician, he's also the one using it, so he has a personal stake. In our conversation, he shared how he’s using Abridge’s clinical intelligence in his own practice and how it's earned his trust as both a clinician and an informatics leader.

In Conversation with Dr. Brian Dilcher

Tell us about your practice and role at West Virginia University Medicine.

I'm an emergency medicine physician at WVU Medicine’s Ruby Memorial Hospital—a Level I trauma center and tertiary/quaternary referral center. We see about 55,000 visits a year in the main ED, and there's an adjacent children's ED that sees about 16,000-17,000 visits a year. I work mostly in the adult ED, with a shift or two a quarter in the pediatric ED. My role is split, half clinical, half informatics and administration, as associate CMIO for emergency medicine. I work nights, so I'm predominantly a nocturnist.

Take us back to when you first used decision support within Abridge. What was your initial reaction?

What stood out right away was that the suggested questions were already informed by the patient context. Before this, I'd use a well-known clinical database. Take something like hyponatremia as an example: There's a big flow diagram that's hard to memorize, so I'd search the database for the right article, rely on muscle memory to find it, and maybe “Ctrl+F” once I got there to find the section I needed. Now I can just ask the question in Abridge directly, in natural language, and get an evidence-based answer specific to the patient in front of me. Since part of my role is validating Abridge as a resource before I endorse it to partners and residents, I sometimes use it just to double-check something I already know, since it adds the color of the patient context.

How does it hold up in an environment as fast-moving as the ED?

Between trauma pages, stroke alerts, transfer calls, and constant interruptions, you rarely have time for a full chart review in the ED. Most clinicians get skilled at making sound decisions without that complete picture, not because they want to, but because there simply isn't time to dig through every prior visit. Abridge changes that. It can surface that a patient saw their doctor two weeks ago for the same issue, or went to urgent care the day before, pulling together a trend I'd otherwise have to piece together manually. Catching that kind of pattern quickly, instead of missing it in the noise of a busy shift, can meaningfully change a patient's outcome.

As an informatics leader and clinician yourself, what gives you confidence in using and recommending decision support within Abridge?

For me it comes down to repeatability, consistency, and transparency. I want to know I'll get sound responses that pass the sniff test, and that I can see exactly where those responses come from. It's not enough to just see a citation. I want to be able to click through to verify that a source actually supports what's being suggested and trace responses back quickly, which Abridge allows me to do. Being involved in how these technologies get built and implemented also gives me more confidence than a general physician might have going in, since I'm not just using it, I'm vetting it, using it regularly, and giving feedback based on real experience. That's the standard I'd want from anything I put in front of a patient, and it's the same standard I hold this to before I'll recommend it to anyone else.

What’s the No. 1 way that Abridge’s context-aware clinical intelligence has changed your practice?

That habit of spot-checking my decisions against the evidence started early and just stuck, and it's probably saved me the most time day-to-day. It sounds like a small thing, but in an environment like the ED, where every extra minute matters, having that kind of quick, dependable check has made a real difference.

This interview was edited for length and clarity.

Want to hear from more clinicians like Brian? Check out the rest of the series on our newsfeed here. And if you're curious how Abridge's context-aware decision support could fit into your workflow, head to the CDS page on our website to learn more.