Home
/
Resource Center/
Case Study/
Smarter Selection, Sharper Results: How Precision AI Unlocked $1M+ in Inpatient DRG Savings
Smarter Selection, Sharper Results: How Precision AI Unlocked $1M+ in Inpatient DRG Savings
By combining intelligent claim selection with precision AI, a Medicare Advantage health plan uncovered more than $1M in incremental inpatient DRG savings.
Published: July 16, 2025
Topics:
Payment Integrity
The Challenge
Cleared claims that were never truly clean
Across health plan payment integrity (PI) programs, there is a consistent challenge: inpatient DRG claims are among the most clinically complex, making them the richest source of high-value clinical and coding opportunities–but also the most difficult to identify consistently. Unsupported diagnoses, documentation gaps, coding errors, and DRG assignment discrepancies are often buried within lengthy medical records, requiring deep clinical and coding expertise to uncover.
For this Medicare Advantage health plan, the question wasn't whether claims were being reviewed–it was whether the right technology was in place to find everything worth finding. Reviewing large volumes of medical records without the ability to intelligently prioritize those most likely to contain payment issues spreads resources thin and limits savings potential. Without visibility across the full audit population, the plan had no reliable way to know whether existing review programs were capturing every meaningful opportunity–or leaving value on the table.
The challenge wasn't replacing the health plan's existing PI program. It was enhancing it with technology capable of identifying opportunities hidden within claims that had already moved through traditional review workflows–without creating unnecessary provider abrasion through duplicate outreach or redundant dispute cycles.
Cohere's Opportunity
Improve identification of complex DRG issues
Focus review efforts on the highest-value claims
Reveal savings hidden within existing workflows
The Solution
A fresh look at the “no findings”–not a duplicate audit.
To strengthen its inpatient DRG review strategy, the plan implemented Cohere Complete for PI–leveraging AI-powered technology and intelligent claim selection to surface opportunities within previously reviewed claims.
Rather than casting a wide net across every previously reviewed inpatient audit, Cohere first applied intelligent claim selection to analyze the entire claims population and the corresponding medical records to prioritize only those with the highest probability of genuine clinical or coding issues. Every potential finding was supported with transparent clinical evidence and clear rationale, ensuring results were both actionable and highly defensible.
The process begins with Cohere Health’s smart selection technology–built on decades of clinical and coding expertise and continuously refined through real-world audit outcomes. AI-powered proprietary models identify claims most likely to contain savings opportunities, including unusual lengths of stay, coding inconsistencies, unsupported documentation, DRG mismatches, and other clinical indicators that warrant further review.
Once high-priority claims are identified, Cohere’s AI agents review medical records at scale–rapidly separating genuinely clean claims from those requiring deeper review. Clean claims are returned to the plan immediately, eliminating unnecessary provider outreach or wasted audit efforts. For claims needing review, the AI highlights exactly where clinical documentation fails to support the billing, allowing reviewers to focus immediately on the relevant evidence rather than manually searching through the record.
Clinical and coding experts then validate AI-identified findings, applying specialty expertise to ensure every determination is clinically sound, aligned with payer policy, and highly defensible.
Once findings are confirmed, Cohere issues transparent findings letters to providers grounded in clinical evidence–not opinions or opaque determinations. Each letter provides a root-cause explanation, giving the plan and its providers a clear picture of all issues. That level of transparency makes findings harder to dispute and gives providers the information they need to correct upstream billing practices.
The Cohere Difference
Three key capabilities powered results.
Intelligent Claims Selection
AI-powered proprietary models identify clinical and coding issues standard reviews overlook
AI/ML continuously analyzes claims to surface high-value savings opportunities
Models are continuously refined based on results–growing more precise over time
Precision AI
AI agents read medical records at scale, pre-screening for issues before auditor review
Unsupported billing is highlighted–so auditors can go straight to the issue
Audit rationale is auto-generated and parsed into a letter template
Transparent Findings
Findings grounded in clinical evidence and payer policy with clear rationale
Root cause explanations pinpoint the billing issue for improved provider education
Detailed findings documentation ensures audit results are highly defensible
The Results
Finding more by looking at the right things.
1 in 3
“no finding” audits converted to findings
$1.15 PMPM
in savings that would have gone uncaptured
(in 4 months)
(in 4 months)
The numbers tell a story that's difficult to ignore. Across the subset of "no finding" claims identified as most likely to have issues, 1 in 3 returned validated findings. This result reflects the precision of Cohere’s intelligent claim selection–Cohere didn’t find more by reviewing more claims. It found more by looking at the right claims.
Cohere didn’t find more by reviewing more claims. It found more by looking at the right claims.
In just four months, the program surfaced $1.08M in identified medical expense savings–and without a more intelligent approach to claim selection, that $1.15 PMPM in opportunity would have remained undiscovered entirely. With zero appeals filed to date, the program demonstrated the strength of its transparent, evidence-based approach.
Beyond the financial impact, Cohere delivered something equally valuable: visibility. The insights captured in each findings letter close the loop on each claim, giving the plan a clear picture of systemic billing patterns that can be addressed upstream and reducing overpayments before they happen.
By combining intelligent claim selection, precision AI, expert clinical validation, and transparent findings, Cohere Health demonstrated how modern payment integrity programs can move beyond broader review toward smarter review–helping health plans uncover meaningful savings.
