CDI Insights #2
Updated: 2 days ago
Every year the claims side of the house gets a new survey, and this year's numbers should stop CDI leaders cold. Experian Health released its State of Claims 2026 report in late October, fielded among 204 healthcare staff in finance, billing, registration, reimbursement, claims and collections. Forty percent of them report that 10 percent or more of their claims are denied. That number has barely moved in two years. Eighty-nine percent now call denial reduction an organizational priority, the highest level since the survey began.
Read the survey's top denial reasons and you will see coverage issues leading the list. Then two entries tied for third place at 25 percent: code inaccuracy and prior authorization. Code inaccuracy is ours. It sits in the middle revenue cycle, in the space between the chart and the claim, and it is the one denial reason CDI can actually influence.
The question I keep asking is whether our current metrics would even tell us if we were part of the problem.
What the survey says, and what it does not
The State of Claims 2026 report describes a system under sustained pressure. Seventy-nine percent of respondents say payer reimbursement is more challenging than a year ago, and 78 percent say payer policy changes are more frequent. Fifty-three percent say errors in submitted claims are on the rise.
The respondents were claims and finance staff, so the survey speaks the language of the claim. But here is the gap I see, and this is my read rather than the survey's. A claim that carries an unsupported MCC or a debatable principal diagnosis passes every claim scrubber on its way out the door. Scrubbers check validity and format, not clinical truth. The claim looks clean right up until a payer's clinician reads the note and denies it for clinical validation. By then it is not a clean claim problem. It is a documentation and CDI problem that the clean claim rate never saw coming.
The metric most CDI departments still do not have
A widely read ICD10monitor column published November 6 made an argument I have been making in my own teaching for years: most CDI departments are measured on query rates, response rates and agreement rates, and none of those metrics hold us accountable when a record we reviewed is later denied for a documentation issue we could have addressed. The column proposes tracking query accuracy, meaning the share of CDI-reviewed records that survive payer review versus the share later denied, downgraded or taken back for an addressable documentation or clinical validation issue.
I agree with the direction, and I would add the clinical piece the claims side always misses. Accuracy is not just whether the query was sent on time and worded compliantly. It is whether the clinical indicators in the note actually supported the diagnosis being queried, and whether the resulting documentation would hold up when a payer clinician reads it cold. A query that removes an unsupported diagnosis and lowers the DRG is an accurate query even though it hurts your case mix index. A query sent only to copy a diagnosis into the discharge summary adds delay without adding anything a payer will accept.
This is why I care more about what the record can withstand than what it adds to the claim.
Why this matters more in 2027, not less
Acuity keeps rising and payers keep refining their clinical validation programs, so the number of diagnoses available to challenge grows every year. The same survey found 64 percent of providers plan to invest in AI for claims or denial management. When payers and providers both automate, the records that fail are the ones where the clinical story was never fully told in the first place.
Technology flags faster. It does not decide whether the clinical picture supports the diagnosis. That judgment is the work, and it is the part that cannot be outsourced.
What to do this week
Ask your revenue cycle partner for your first submission pass rate, not just your clean claim rate. The first tells you whether claims were paid correctly the first time. The second only tells you they were accepted for processing.
Build a feedback loop from denials back to CDI by encounter, diagnosis and reviewer, including takebacks that arrive by letter or portal and never appear in the remittance file.
Pull the last 10 denials with a clinical validation or DRG downgrade component. For each one, ask the honest question: could a better, earlier or differently worded query have prevented this?
Make that review educational, never punitive. The goal is a team that can tell the difference between a record that is fine and a record that merely looks fine.
Bring one finding to your next finance meeting. Denial prevention has executive attention right now, and CDI should be at that table with evidence, not activity counts.
The claims survey measures what payers do to us. Query accuracy measures what we do to ourselves. The second one is the only one we control.
If your CDI team is reviewing hard records but denials tied to documentation and clinical validation keep coming, the CLINICAL Logic Framework is the program I built for exactly that team. You can find it at https://www.thequerycorner.com/clinical-logic-framework
Sources:
Experian Health, "State of Claims 2026: Why denials remain an uphill climb," November 6, 2026: https://www.experian.com/blogs/healthcare/state-of-claims-2026-why-denials-remain-an-uphill-climb/ (survey fielded August 5 to 23, 2026 among 204 healthcare staff; 40 percent report 10 percent or more of claims denied, compared with 41 percent in 2025 and 31 percent in 2022; 89 percent call denial reduction an organizational priority, up from 82 percent; 79 percent say reimbursement is more challenging; 78 percent say payer policy changes more frequent; 53 percent say claim errors rising; "services not covered" cited in 53 percent of top three reasons, payer policy changes 31 percent, code inaccuracy and prior authorization tied at 25 percent; 64 percent plan AI investment, 9 percent currently use AI)
ICD10monitor, "The Query Accuracy Gap Contributing to Administrative Burden," Cheryl Ericson, RN, MS, CCDS, CDIP, November 6, 2026: https://medlearn.com/icd10monitor/the-query-accuracy-gap-contributing-to-administrative-burden/ (argument for tracking query accuracy; clean claim rate versus first submission pass rate)




Comments