Audit Alert
OIG Nationwide Acute Stroke Audit
97 out of 97
failed validation.
OIG's nationwide acute stroke audit found every sampled high-risk acute stroke HCC was unsupported.
01
Why this matters
This was a visible coding pattern — not a vague documentation issue.
The finding did not hinge on subjective chart interpretation. Reviewers identified acute stroke HCCs on physician records and simply checked for matching hospital evidence in the same service year. When it wasn't there, the diagnosis failed. That makes the pattern detectable, repeatable, and exactly the kind of signal a RADV audit is built to catch.
02
The pattern OIG found
Three steps, every time.
STEP 01
Acute stroke code appears on a physician record
STEP 02
No matching acute stroke hospital evidence in the same service year
STEP 03
Diagnosis fails validation
That is a RADV red flag.
03
How this error starts
A code that outlives the event.
THE EVENT
Acute stroke code added during the event
DAYS LATER
Remains on the problem list
MONTHS OR YEARS LATER
Carried forward into follow-up documentation
TODAY
Current documentation may no longer support acute stroke
04
The clinical distinction
Not every prior stroke is an acute stroke.
Acute stroke
A current event with hospital-level evidence.
Stroke sequelae
Documented residual deficits from a prior stroke.
History of stroke
A prior stroke with no current deficits.
The bottom line: the code should follow the current clinical reality.
05
What to do now
Run the same check on yourself.
Pull physician records with selected acute stroke codes
Cross-reference hospital records from the same service year
Review physician-only cases
Validate what the documentation actually supports
Submit corrections where needed
06
Questions your team should ask now
5 questions to ask yourself before RADV does.
Q1
Which acute stroke codes are on our physician records, and how many are physician-only?
Q2
For each one, is there matching hospital evidence in the same service year?
Q3
How many acute stroke codes are being carried forward on problem lists past the event year?
Q4
Does our current documentation support acute stroke, or does it actually describe sequelae or history?
Q5
What internal review process catches this before submission — and who owns it?
High-risk HCC patterns are visible.
The question is whether you find them before CMS does.
Read the full breakdown
Learn how to catch acute stroke coding errors before they become RADV findings.
Read the full blog
Source: HHS-OIG nationwide acute stroke audit. Findings apply to the sampled high-risk acute stroke HCCs only.