Clinical documentation integrity teams have long relied on retrospective queries to close gaps in physician documentation. But as payers accelerate AI-driven claims review and increasingly deny diagnoses added to the chart after the clinical moment, that approach is becoming a liability — and the infrastructure built around it increasingly hard to justify.
During a May 7 webinar hosted by Becker’s Hospital Review and sponsored by Regard, three leaders examined what it takes to move documentation upstream and how CDI and the technology supporting it have to evolve to make that shift work. The panelists were:
● David Kirk, MD, Chief Medical Officer of Regard
● Carolyn Bauer, MD, Medical Director of Clinical Documentation at Montefiore Medical Center in New York City
● Fran Jurcak, MSN, RN, independent consultant in CDI and revenue cycle
Below are four takeaways from their conversation.
Note: Quotes have been edited for length and clarity.
- Queries themselves are becoming a denial signal
Dr. Kirk opened the discussion with a theme that surfaced throughout the conversation: payer scrutiny of documentation changes made after the clinical moment is intensifying, and diagnoses added through retrospective queries can be more difficult to defend than those captured during care.
That has direct implications for work CDI teams have done for years. Ms. Jurcak returned to the theme later in the session: “We are getting into a world where any query, especially a retrospective query, is going to be a reason for an automatic denial.”
That puts new pressure on the retrospective infrastructure CDI teams have built around closing those gaps.
- The retrospective infrastructure is a hidden tax on cost to collect
Dr. Bauer detailed the layered cost of catching documentation problems downstream: a CDI query to the physician, the physician’s time responding, post-discharge chart review, a second round of queries, safety-net vendors, AI tools combing the record and additional headcount to fight denials.
“If we can scale that back by getting it right the first time, you can get cleaner claims out the door quicker,” Dr. Bauer said. Hospital leaders, she added, are starting to ask how many touches a single record requires before it goes out the door and how many come back as denials anyway.
- Ambient AI captures more words, but not always the full picture
The panel was careful to separate ambient documentation tools from the broader point-of-care AI category. Ambient scribes capture what is said during the encounter to produce a more complete record of the conversation. However, they don’t have the ability to reason across labs, vitals, prior history or imaging.
Ms. Jurcak argued that the value of an AI assistant is in surfacing supporting evidence and statistical likelihood, not in declaring diagnoses outright.
“If you gave the statistical likelihood that it’s 70% likely to be sepsis, a physician is going to look at that a different way and better consider whether or not they should be documenting it,” she said. That framing also addresses a common physician complaint about tools that surface stale or irrelevant diagnoses pulled from older notes.
Dr. Kirk added color to the conversation, sharing that as an ICU physician what kept him at night was the fear of missing something hidden in the chart that would have changed the trajectory of care. Ambient doesn’t solve the problem of cutting through the noise of documentation, and in some cases, exacerbates it.
- AI doesn’t replace CDI — it refocuses it on the complex 20%
Removing routine queries from CDI’s workload doesn’t shrink the function; it redirects it. Documentation, quality, and revenue problems are ultimately diagnosis problems, and that’s where CDI’s expertise is most needed. Ms. Jurcak framed the split as roughly 80/20 — the simpler queries automation can handle, and the complex clinical validation, denial-prevention work, payer-trend monitoring and ICD-11 preparation that still require experienced human judgment.
Dr. Bauer noted CDI teams already run chronically understaffed, which makes the redirect more urgent than threatening. The harder pitch internally, she said, is convincing leadership that physicians will accept a documentation tool at all — and the answer is that they do when it saves time, surfaces evidence and reduces query burden rather than adding to it.
What this means for CDI and revenue cycle leaders
The panel’s throughline was that physician satisfaction is the leading indicator of whether a point-of-care documentation strategy is working. When physicians use the tool, the rest of the metrics — case mix index (CMI), severity scoring, denial rates, time to bill — start moving in the same direction. When they don’t, no amount of retrospective rework will close the gap that payer scrutiny is widening each quarter. The underlying argument is simple: get the diagnosis right at the point of care, and the documentation, the quality scores, and the revenue follow. Everything else is recovery.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.