AI Playbook for Sepsis Documentation for Quality Reporting
A hospital's coding and quality team has identified a documentation gap: physicians are treating sepsis but documenting 'infection' or 'SIRS' without capturing sepsis diagnosis. As a result, the hospital's reported sepsis mortality rate appears artificially high and its case volume is undercounted. CMS quality scores are affected.
When to use this playbook
- Use this playbook when the decision looks like the situation above: A hospital's coding and quality team has identified a documentation gap: physicians are treating sepsis but documenting 'infection' or 'SIRS' without capturing sepsis diagnosis.
- It is a fit when you have source files in hand and need a structured, reviewable analysis — not a generic chat answer about "Sepsis Documentation for Quality Reporting".
- Do not use it as a substitute for licensed, legal, clinical, or authorized official judgment in the domain.
What you'll need
- Coding abstractions for 200 cases flagged as potential undercoded sepsis
- Clinical documentation from those cases (physician notes, nursing flowsheets, lab data)
- Sepsis-3 definition and documentation criteria
- CMS SEP-1 abstraction specifications
- Current physician education materials on sepsis documentation
Attachments: Documents (Documents)
The Prompt
You are a clinical documentation specialist addressing a sepsis documentation gap for CMS quality reporting. I am attaching: Work only from the attached source files. If a conclusion is not supported, say so. Produce: 1. For the 200 flagged cases, assess how many meet Sepsis-3 clinical criteria based on the available documentation—even if the physician did not use the word 'sepsis.' 2. Identify the specific documentation elements missing that would allow coders to capture the sepsis diagnosis: organ dysfunction documentation, clinical decision-making reflecting sepsis treatment. 3. Assess the financial impact of the undercoding: what is the DRG value difference between sepsis and the current coded diagnoses? 4. Design the physician education program: what specific documentation language resolves the coding gap without overcoding. 5. Tell me the CMS quality score impact if documentation is corrected and whether a query program carries any compliance risk. Call out where independent models are likely to disagree, and list follow-up documents a reviewer should request.
What to expect
- Sepsis-3 criteria meeting case count
- Missing documentation element analysis
- DRG and financial impact of undercoding
- Physician education program design
- CMS quality score impact and query program compliance assessment
Review before you act
- Validate this output against source files before relying on it: For the 200 flagged cases, assess how many meet Sepsis-3 clinical criteria based on the available documentation—even if the physician did not use the word 'sepsis.'.
- Validate this output against source files before relying on it: Identify the specific documentation elements missing that would allow coders to capture the sepsis diagnosis: organ dysfunction documentation, clinical decision-making reflecting sepsis treatment.
- Validate this output against source files before relying on it: Assess the financial impact of the undercoding: what is the DRG value difference between sepsis and the current coded diagnoses?.
- Validate this output against source files before relying on it: Design the physician education program: what specific documentation language resolves the coding gap without overcoding.
- Confirm every cited figure, date, counterparty, or requirement against the attached originals — models compress and can drop a qualifier.
- Treat disagreement between models as a review item, especially on classification, materiality, and recommended next action.
- Do not authorize an operational, clinical, legal, credit, or enforcement action solely because the models agree.
Why compare models on this
For Sepsis Documentation for Quality Reporting, running the same attachments across independent models is useful because the hard part is classification and completeness, not fluency. The workflow is already designed to surface sepsis-3 criteria meeting case count; missing documentation element analysis; drg and financial impact of undercoding; physician education program design. Those are comparison artifacts — they only exist if more than one model runs. Models disagree on whether an alert is noise, whether a death was sepsis-attributable, and whether a risk model is calibrated. Those disagreements belong in a morbidity-and-mortality style review, not an auto-implemented rule.
See governed multi-model AI on your own prompt
Compare GPT-5, Claude, and Gemini side by side, with human review and a decision record built in.

