AI Documentation and Your EHR: The Pros and Cons of Embedded, Integrated, and Built-In Solutions

In the ever-growing dandelion field of healthcare AI vendors, orgs like yours have a lot to think about once you decide to implement an AI solution. One of the major points on your checklist is what type of AI solution makes sense for you. Terms like “embedded tech,” “EHR overlay,” “in-depth EHR integration,” and “built-in workflow” are just words. What matters is what those words mean for your organization, your providers, and your tech stack.

At Eleos, we believe clinician-centered, behavioral health-specific AI can profoundly impact provider lives, so we put on our investigative hats and asked: when it comes to AI documentation solutions, what is out there—and how do orgs like yours know what kind of solution makes the most sense (and cents) for their team?

How AI Tech Works With Your EHR

Where does documentation live? In the EHR, of course. So, how AI-generated documentation is informed by—and gets into—the EHR is a key variable in the tech equation. And there are a lot of technological definitions rolling around that pertain to behavioral health AI. So for the purposes of this post, here’s how we define and categorize AI solution types.

Keep these definitions handy as you read through this article (here’s a downloadable version you can have open in another window).

AI and Your EHR: Integrated, Embedded, or Built-in?

How can you know the difference between these three methods of connecting an AI tool to your EHR? We’ve done our best to explain below. (Don’t worry, we’re not going to get too technical. This is not a piece for your CIO.) 

Third Party Integrated Solutions

An integrated solution is one that theoretically exists within one workflow. You shouldn’t have to click into or out of separate screens, though the screens may automatically shift for you (like when you open an app on your phone and click a link in the menu, and the app pushes you to a new app or to your web browser). Integrated doesn’t mean built-in, but it should mean seamless. So if your EHR is promising a truly integrated solution, don’t let them off the hook with clunky workflows.

Third Party Embedded and Overlaid Tech

An embedded, or overlaid, solution lives on top, or inside, of another solution—like the bacon in a great BLT. Or the lettuce. Or even the tomato. Where the bread is your EHR, and the ingredients inside are all where they’re supposed to be—but nothing is directly mixed together. This is not a melted, gooey BLT; it’s a crisp and refreshing one.

Built-in or Attached to the EHR 

A built-in solution is one that is actually part of your EHR—a true all-in-one bundle. None of today’s built-in AI solutions have the same level of sophistication and functionality as the top-tier integrated or embedded solutions that are currently on the market—though that may change as this technology evolves.

Customer Review: “Embedded > Integrated”
“Eleos is the goddess of caring/compassion and it has been very refreshing to work with a company who puts ‘care’ at the center of who they are. Pros: I like that it is an embedded product, not a separate application. It enhances our EHR instead of competing with it and keeps our workflows intact.”

Carin S.

Now let’s take a closer look at what makes each AI solution type the right (or not-so-right) fit for your org. And remember, we’re talking about this in the context of the tech itself—not necessarily the individual companies building the tech. While company-specific factors should certainly play a role in your decision, this particular section focuses more on how a solution works (or doesn’t work)—and less on things like culture or company-specific traits. We know every situation is different—and you have to look at every aspect of an AI tool before bringing it into your org.

What are the Pros and Cons of Each Type of AI Solution?

Now that we’ve defined each AI solution type, let’s run through the biggest pros and cons of each. (Note that for the purposes of this blog, a built-in solution is a product created and owned by the EHR—which you can add onto or bundle with your EHR contract.)

Third-Party Deep Integrations: The Pros

• Deep integration: At the risk of sounding redundant, a deep integration is, well, deep. It’s likely fully mapped and customized to your EHR fields—which is awesome!
• Customized note templates: Your EHR, your integration, your template. Nuff said.
• In-workflow (maybe?): An integrated solution should live in-workflow, but some vendors may still have some clunky kinks to work out. So, even with a deep integration, you may still need to copy/paste.

Third-Party Deep Integrations: The Cons

• Heavy coding/API work: With a deep integration comes deep coding—and potentially lengthy technical timelines. Your IT team will be digging in with the vendor’s IT/implementation teams, and between all the back-and-forth, you could be looking at a delay of several months before you can actually go live with the new tool.
• Limited adaptability: Once you do go live, any tweaks, changes, or updates to note templates can trigger domino effects—which means more email chains with IT teams. Furthermore, your contract may limit anything that can be changed/updated once implementation is complete.
• Heavy change management: Heavy code brings heavy change (and project) management, which could mean lots of extra admin hours. 
• EHR-specific: When you push your integration to a level where fields and templates are customized to your specific EHR, it likely means you are bound to that EHR—for better or for worse.

EHR-built: The Pros

• Years of documentation experience: EHRs exist to store healthcare data, so they have years of experience building structures, systems, fields, and other mechanisms tailored to data storage—they’re very good at it.
• Payer and regulatory familiarity: Like every healthcare vendor, EHRs are on the hook for regulatory compliance, but most are also deeply familiar with payer rules and requirements.
• Large companies: Most EHRs are large enterprises that are relatively stable from a business perspective.
• Built for the business: EHRs are essentially a standard of doing business in the healthcare industry, and that’s how they’re engineered—to serve the business.

EHR-built: The Cons

• Built for the business: What’s great for the org isn’t always great for the provider (i.e.,  the end user). EHRs are not necessarily engineered with clinicians in mind—and sometimes, it shows.
• EHR-bound: This one sounds obvious, but in the spirit of fairness, we want to mention it. EHR-built solutions are 100% tied to the EHR—so changing EHRs means losing your AI solution.
• Limited roadmap access/configurability: EHRs typically can’t customize or configure for specific use cases, so your AI solution is likely to be pretty out-of-the-box.
• Broad, general solution: EHRs are big tech, and they are broad by necessity. Without a laser focus on AI, it could be difficult for EHR-based AI solutions to keep up with the innovation curve.

Overlaid or Embedded Tech: The Pros

• Lightweight: As described above, an overlaid solution doesn’t require heavy lifting from your org’s IT teams, because no code integrations are needed—just an internet connection.
  • Bonus: Lightweight tech means lightweight change management—and that’s a win-win.
• EHR-agnostic: When a technology doesn’t require heavy coding or API connections to work with your EHR, that means it is generally compatible with any EHR. 🎉 Why is that important? Because if you ever change EHRs, your AI solution won’t need a full re-code. It just moves with you.
• In-workflow: The overlay sits on top or alongside of your EHR, which means you’re not clicking around trying to find it (or use it). It simply exists within the documentation workflow you’re used to.
• Works where you work: An overlaid solution goes where your job takes you—whether that’s your office, in the field, or a mix of both.
• Innovation-centric: When a third-party is on the leading edge of innovation, in this case using smart embed and overlay technology instead of deep coding, they’re likely to stay ahead of the curve and keep pace with—or even lead—technology advances in the field. 🏆

Overlaid or Embedded Tech: The Cons

• Possibility of limited scope: Any third-party solution can be limited in how well it works with your EHR, how flexible it is, and what functions it can actually perform. But in the name of being thorough (and unbiased) in our representations, we’re going to mention it. An overlaid or embedded experience may not be as capable as a fully-integrated or built-in solution, especially if the company is brand-new to the game—unlike us old-timers. 😉
• Potential instability/immaturity: With any external vendor, there is a risk that the company is too young, too small, or possibly even too unstable to handle the data load required for a large organization. Especially in AI—and especially in healthcare—it’s important to do your research and identify any red flags early. 👀

Speaking of doing research, this seems like a great time to mention how important security certifications and infrastructure—and a security-focused culture—are when it comes to any AI solution, no matter what type of technology blocks are used to build the product.

Beyond the Tech: Let’s Talk About Specialization

Third-party AI solutions (like Eleos) come in all shapes, sizes, and specialties. Some are built to serve individual providers, while others are made for entire organizations. Some work for multiple disciplines, while others focus on just one or two use cases. So, to wrap up this educational blog 🤓, we’re going to focus on some unique perks of an enterprise-grade AI specializer—like us!

As an anonymous attendee of our 2024 CIO Summit stated, “There’s something to be said for an org that solely focuses its resources on AI, versus an org that offers AI as a feature.”

What Makes an AI Specializer Special?

• Specializers know AI inside and out, because AI is all they (cough, we) do. From the nuances of artificial intelligence versus augmented intelligence to the unique AI use cases for their target market, a true AI specialist lives and breathes the latest information, research, and best practices for this technology in their particular niche—in this case, behavioral health.
  • Bonus: Specialists often come with super smart code wizards who know a thing or two, because they’ve developed a thing or two. Check out one of our super smart data wizard’s recent post on Medium: Baking the Future of AI with RAG and VectorDB.
• They’re enterprise-grade, but they’re engineered with a startup mentality. Staying fully focused on client needs, specializers can offer nimble roadmaps, quick fixes and updates, agile teams, and flexible configurations—all in the name of doing what’s best for everyone.
  • Bonus: When you work with an agile specializer like Eleos, you unlock roadmap influence. Case in point? Trilogy helped us build our case manager solution.
• Specializing in AI is one thing—specializing in behavioral health AI is another. With a solution like Eleos, you get both: sophisticated AI that is purpose-built for behavioral health. And with a strong clinical presence within our in-house team, Eleos engineers its products with clinicians in mind: every button, every field, and every feature is considered from a provider’s perspective.

“For any technology that leverages AI models, effectively training the models to perform with a high degree of accuracy is critical. For a model specific to behavioral health, that level of training is only possible with input from clinical experts—which is why our clinical team is a core differentiator for us at Eleos. Furthermore, the overall effectiveness of the model hinges on the quality of the data used to train it. By training our models with real-world session data, we enhance their applicability in the intended context: community-based behavioral health clinics.” – Eleos Chief Clinical Officer, Dr. Shiri Sharvit 

• Specializers have the bandwidth (and drive) to support a passionate commitment to science. It isn’t enough to have happy customers singing praises; the science also needs to back up the solution—or at least it does for this specializer.

When it Comes to AI and EHRs, There is no Right or Wrong—Only Best for Your Organization

If you’re thinking about implementing an AI solution, there is no easy button—and in healthcare, there really shouldn’t be. So, do your due diligence, ask the right questions, and be sure to look out for red flags. You’ve got this!

The Final Word: Who Do You Trust with Your Teeth?

When you have a toothache, do you go to your primary care doctor? Probably not. Specialists are specialists for a reason—and to be honest, the same is true when it comes to technology vendors. 

Eleos is a specialized, purpose-built solution that puts clinicians first—and believes in a science-driven approach. It’s just what we do. Ready to learn more? Talk to an expert.

Book your demo!