DRG grouper
Group and price a claim, with CDI review.
Shows the grouping path step by step, the code that set the severity, the relative weight and the other DRGs in the family.
Claim
Codes
Patient and stay
Hospital and charges
Import a claim JSON
Four shapes convert in your browser: native (diagnoses, procedures, age, sex, dischargeStatus), a vendor PatientClaim/Dx claim, a FHIR R4 Claim resource, and an Epic Tapestry-style claim.
Coded diagnoses
Clinical note
Ask AI is opt-in: it sends this note through our server to our AI model provider (Anthropic) for one scan (leave out names, dates of birth and other identifiers), and every code it proposes is checked against the ICD-10-CM and CC/MCC tables. The rule-based scan runs in this page and matches named conditions only. Note text is never put in links or saved in this browser.
Result
Grouping path
DRG family
Codes that drive the DRG
| Code | POA | Level | Effect on the DRG |
|---|
What holds this DRG
Grouper response (raw)
Grouping path
Possible documentation gaps for provider clarification
Impact if documented and clinically supported
How the grouping would change if the provider documented the condition, it is clinically supported, and it were coded. Shown so coders understand the grouping, not to decide what to query: queries follow the clinical evidence in the record.
Checks
Validation edits
Grouper notes
Version and engine
Price in Inpatient claimHow suggestions are made
- Rule-based
- Matches named conditions (for example "acute kidney injury") against a fixed list of common documentation gaps. Runs in this page. Not AI.
- AI-assisted
- Only when you press Ask AI. A model reads the note, including findings such as lab values, and proposes conditions. Every code is checked against the ICD-10-CM and CC/MCC tables, and codes already coded are dropped.
- Saved example
- A model scan of the sample note captured earlier and shipped with the page, so you can see the output without running anything.
Why the rule-based scan found nothing
This sample names no diagnoses, only findings. The rule-based scan matches named conditions, so it finds nothing here. Reading findings and inferring the documented condition is what a model scan adds, and any inferred condition still needs the provider to document it.
Documentation integrity
Each item is a possible documentation gap for the provider to clarify, with the note text that prompted it. Code only what the provider documents and the record clinically supports. Provider queries copied from this page are open-ended, offer other answers, and never mention payment or the DRG.