From chart reviews to deep reasoning: Why I believe the future of CDI is human + AI

By Nita Bruce, Director, Strategic Client Solutions 

For twenty years, I sat inside the clinical record. Not as a clinician, but as a CDI consultant working side by side with Clinical Documentation Specialists and coding professionals, combing through charts one by one, hunting for what the documentation missed. We looked for the clinical evidence that was right there in the labs, vitals, medications, and treatment plans—evidence that a condition was being monitored, evaluated, or treated, and therefore could and should be captured as a diagnosis. When we found it, we built queries. We educated physicians. We helped health systems tell the complete and accurate story of how sick their patients truly were. 

It was meaningful work, and I loved it. But I will be the first to admit: it was painstakingly slow. 

A human reviewer can only get through so many charts in a day. No matter how experienced, no matter how sharp, there is a ceiling. We triaged. We prioritized the cases most likely to yield results and accepted that some charts would never get a second look. The math was simple and unforgiving—there were always more records than hours. 

This is not a knock on the incredible professionals I have worked alongside for two decades. It is an honest reckoning with what technology can now do. HDAI has built deep reasoning large language models and predictive analytics that can do what I used to do—trace the thread from a lab value to a medication to a clinical indication and recognize a documentable diagnosis—but at a scale and speed that no team of human reviewers could match. The AI does not get tired at 3 p.m. It does not skip a chart because the queue is too long. It does not have a bad Monday. 

Watching it work felt like seeing twenty years of my career compressed into seconds. The same logic I saw painstakingly applied chart by chart, the AI was applying across an entire patient population, simultaneously. 

Now, I know what you might be thinking. Is she saying AI should replace CDI specialists and coders? 

Absolutely not. 

Here is what two decades in the trenches taught me: clinical documentation is not a mechanical exercise. It requires judgment. It requires understanding of the patient story, the physician’s intent, the nuance of language, and the regulatory landscape that shifts beneath our feet every year. It requires a human being who can pick up the phone, have a conversation with a provider, and build the kind of trust that leads to lasting behavior change. 

AI does not do that. AI does not sit in a physician advisory council meeting and read the room. It does not mentor a new CDS through her first query on sepsis. It does not navigate the politics of a documentation improvement program across service lines. 

What AI does extraordinarily well is handle the volume. It can review every single encounter—not just the ones that make it to the top of the worklist. It can flag patterns that a human reviewer might not notice until months of data have accumulated. It can surface the low-hanging fruit so that CDI professionals can focus their expertise where it matters most: on the complex cases, the education, the relationship-building, and the strategic work that actually moves the needle. 

Think of it this way. For years, we have asked brilliant, experienced professionals to spend a significant portion of their day on tasks that, while necessary, do not require the full depth of their training. AI can take on that foundational layer of review—the initial screen, the pattern recognition, the data synthesis—and free the human experts to do what only they can do. 

I have spent my career watching CDI evolve. From retrospective-only programs to concurrent review. From paper-based queries to electronic workflows. From standalone departments to enterprise-wide integrity initiatives. Each evolution made the work more efficient. AI is the next evolution, and like everyone before it, it will elevate the value of CDI professionals. 

At HDAI, we are augmenting CDI teams with superpowers that will help them make sure that every patient's story is told completely and accurately —not just the ones that a human happened to have time to review. 

After twenty years of doing this the hard way, I can tell you: the future is not AI or humans. It is AI and humans, working together, each covering the other’s blind spots. And from where I sit now, that future looks incredibly bright. 

HDAI 

Transforming clinical documentation integrity through deep reasoning AI and predictive analytics — built by people who have lived the work. 

Want to learn how? Let’s talk.