'Probable SBP' Meets a Consultant Who Disagrees: The Inpatient Uncertain-Diagnosis Rule and the Attending of Record
A 55-year-old man with alcohol-related cirrhosis is admitted with fever, abdominal pain, and tense ascites. Diagnostic paracentesis yields 480 PMN cells/mm3; ascitic culture is negative at five days. The hospitalist documents 'spontaneous bacterial peritonitis, probable' and completes a five-day course of ceftriaxone. The gastroenterology consultant writes: 'Ascitic fluid neutrophilic; SBP unlikely given negative culture - consider secondary peritonitis or pancreatitis; recommend further imaging.' The patient is discharged on day 7, and the hospitalist's discharge summary restates 'probable SBP.' A CDI specialist who joined the program two weeks ago wants to query the gastroenterologist to 'clarify the diagnosis' before the claim goes out.
1. Is a query warranted here, and to whom should it be directed? Explain the role of the inpatient uncertain-diagnosis rule.
Read the worked answer
A query is warranted, but not the one the new specialist proposed, and not to that physician.
- Recipient: queries go to the attending of record responsible for the final diagnosis. A consultant may support or question a diagnosis, but the attending is accountable for what is coded. Querying the gastroenterologist invites a second, conflicting 'final' answer.
- The uncertain-diagnosis rule: in the inpatient setting, a condition documented as probable, suspected, or likely is coded as though it were confirmed. That means 'probable SBP' is already codeable as SBP - so a query simply to 'confirm SBP' is unnecessary, and asking it signals that the CDI is steering toward or away from the diagnosis.
- What actually needs resolving is the conflict between two providers' notes. That is a documentation-integrity issue affecting the principal diagnosis, not a specificity issue.
Compliant query to the hospitalist: 'The record documents probable spontaneous bacterial peritonitis with 480 PMN cells/mm3 and treatment with ceftriaxone for five days. The gastroenterology consult note states SBP is unlikely given a negative culture and suggests considering secondary peritonitis. Please clarify your final diagnosis for this admission: [ ] Spontaneous bacterial peritonitis [ ] Secondary peritonitis [ ] Ascites with neutrophilic fluid, SBP excluded [ ] Other: ____ [ ] Unable to determine.'
2. The hospitalist responds: 'SBP was treated empirically; consultant's opinion noted.' Is that codable, and what are the CDI's options?
Read the worked answer
No. That response restates treatment and acknowledges a disagreement without resolving it. Treatment is not a diagnosis - giving ceftriaxone does not establish SBP, and the query response must produce an answer that can be coded.
Options, in order:
- Re-query once, narrowly, asking for a yes/no on the final diagnosis, keeping the clinical indicators and the 'unable to determine' option. Say explicitly why the earlier answer was insufficient: 'I need to know which diagnosis you want reflected as the final diagnosis so the record is accurate.'
- If no further clarification is obtainable, the attending's own documentation governs. 'Probable SBP' is coded as SBP under the inpatient uncertain-diagnosis rule; a consultant's note does not override the attending. Document the query attempt and route the case to the physician advisor or coding manager for a final determination rather than deciding unilaterally.
- If the attending later withdraws the diagnosis ('SBP excluded, this was ascites with neutrophilic fluid'), SBP comes out of the record and the DRG changes accordingly. That outcome is acceptable and should be reported as-is.
Note the asymmetry the examiner is testing: the CDI may not code a diagnosis the attending has ruled out, and may not keep a diagnosis alive simply because it is already documented and financially favorable.
3. What would a noncompliant version of this query look like, and why does the difference matter beyond etiquette?
Read the worked answer
Noncompliant versions share one feature: they narrow the physician's choice to the answer the hospital wants.
- 'SBP has a higher-weighted DRG than ascites - please confirm SBP.' (Reimbursement motive disclosed in writing.)
- 'The patient was treated with ceftriaxone for five days; please document SBP.' (Treatment presented as proof; no alternative, no clinical indicators from the lab data.)
- 'Please clarify whether this is SBP or secondary peritonitis.' (Two options only; forecloses 'neither,' which the consultant's note makes a real possibility.)
- 'Per the consultant, SBP is unlikely - please amend the discharge summary.' (Leads the physician toward a specific answer and invites an amendment based on another provider's opinion.)
Why it matters beyond etiquette: a query is part of the medical record and is discoverable. A pattern of closed-ended, financially motivated queries is the evidentiary core of a reverse false claims case, and it also corrupts the clinical data the hospital uses for quality, mortality review, and risk adjustment. The compliant query above cites objective indicators, offers a genuine 'unable to determine,' avoids etiology suggestions, and leaves the clinical judgment where it belongs.