For years, healthcare organizations have relied on retrospective chart reviews to identify documentation gaps and missed diagnoses. The challenge is that these reviews typically occur days after care has been delivered, when clinicians can no longer act on insights that may have affected treatment decisions, quality reporting or reimbursement.
As hospitals seek new ways to improve operational and financial performance, many are turning to AI-enabled workflows that bring clinical documentation intelligence directly into the care process.
During a Becker’s Healthcare webinar sponsored by Qventus, chief medical officer Jason Cohen, MD, and senior director of product management Katie Saindon explored how real-time AI interventions are helping health systems close documentation gaps before discharge, capture clinical complexity more accurately and improve outcomes across the continuum of care.
Below are four key takeaways from the conversation.
1. The gap is bigger than most organizations realize
Research shows 20 to 30 percent of inpatients are malnourished, yet most health systems code the condition in only five to 15 percent of cases. That discrepancy has real financial consequences — roughly $8,000 in missed reimbursement per uncoded malnutrition patient, and as much as $30,000 for surgical patients.
Across a health system, Dr. Cohen said, those gaps compound fast.
“We’re already providing most of this care for these patients,” Dr. Cohen said. “But we don’t capture their clinical complexity and it is literally like money on the table that we could be using to serve our mission for other patient populations.”
Beyond reimbursement, undetected malnutrition correlates with readmission rates more than twice as high and longer lengths of stay, factors that affect quality scores and expected mortality metrics.
2. The missing piece
Webinar attendees identified physician documentation that accurately reflects clinical complexity as their single biggest barrier to closing care gaps, a finding Dr. Cohen said tracks with his experience practicing as a hospitalist. The issue isn’t awareness, he argued. It’s that existing CDI tools weren’t built around the full care workflow.
“You need to always be biased towards action,” Dr. Cohen said. “You have to put it where people are already working so that they can see it, take advantage of it and it doesn’t slow them down.”
Ms. Saindon added that inconsistency between what dieticians diagnose and what physicians document is a persistent source of leakage — in one organization she analyzed, physicians’ notes failed to reflect a dietician’s malnutrition diagnosis 47 percent of the time.
3. When AI works best
The Qventus Care Gap and Coding Automation Suite runs AI models daily on all adult inpatients, drawing on both structured data (weight history, BMI, labs) and unstructured sources (physician progress notes, nursing documentation, physical therapy notes) to flag patients at risk. When a patient is identified, the solution surfaces a prompt directly in the physician’s chart and pre-populates a nutrition consult order, then all the clinician has to do is sign it.
Ms. Saindon shared a case from a go-live she attended the day before the webinar: a patient who scored zero on the standard malnutrition screening tool but was ultimately diagnosed with severe malnutrition after the AI model flagged clinical evidence in the chart.
“If I had not gotten this prompt,” the dietician told her, “this patient would not have been diagnosed and they would’ve left the hospital with nothing.”
4. The outcomes are strong
One large Midwestern academic medical center had already been capturing malnutrition at above-average rates — 15 to 18 percent of patients — before working with Qventus. Post go-live, the system saw a two-plus percentage point increase in capture rates and a relative lift of eight to 24 percent across three hospitals, translating to $256,000 in incremental reimbursement in the first six weeks alone.
Annualized, Dr. Cohen projected that outcome to around $2.2 million, consistent with what Ms. Saindon said Qventus typically sees for a 500-bed system. Combined reimbursement, mortality O/E improvement and dietician productivity gains add up to roughly $40 per discharge, she noted, with malnutrition serving as the first condition in a suite that will expand to include pressure injuries, AKI, delirium and sepsis.
“I want AI to actually help me get the right care for my patients,” Dr. Cohen said. “I want it to decrease the cost of care that we’re providing to patients and make sure that our health systems are able to meet our mission to our communities.”
At Becker's 4th Annual CEO + CFO Roundtable, taking place November 2–5 in Chicago, more than 1,500 hospital and health system executives tackle decisions that determine whether organizations thrive or merely survive: protecting margins under cost pressure, choosing where to grow, renegotiating payer relationships, stabilizing the workforce and proving real ROI on technology. This is where leaders work through them together, face-to-face. Apply for complimentary registration now.
