19.1 DRG Compliance, DRG Creep, and the OIG Work Plan
Key Takeaways
- DRG creep is a pattern of higher-weighted MS-DRG assignment driven by coding or documentation practice rather than verified clinical severity, not CMS's annual relative-weight recalibration.
- The HHS OIG Work Plan is a dynamic priority signal for audits and evaluations; it is not an ICD-10 code list, a hospital-specific do-not-bill list, or immunity for unlisted MS-DRGs.
- Use historically typical OIG issue types—sepsis billing, severe-malnutrition MCC capture, kwashiorkor-type implausible codes, and ventilation-hour DRGs—as method examples, not as a frozen 2026 target-DRG inventory.
- A working DRG compliance program combines targeted MS-DRG audits, query monitoring for leading or payment-driven language, and clinical validation of documented but unsupported diagnoses.
- A signed attending diagnosis does not replace UHDDS and Official Guidelines support; silent retention of unsupported CCs/MCCs recreates the payment-integrity pattern OIG samples.
19.1 DRG Compliance, DRG Creep, and the OIG Work Plan
Quick Answer: DRG creep is a pattern of assigning higher-weighted Medicare Severity Diagnosis-Related Groups (MS-DRGs) than the record supports—usually through unsupported principal-diagnosis selection, complication or comorbidity (CC) / major complication or comorbidity (MCC) capture, or procedure-duration coding. The U.S. Department of Health and Human Services Office of Inspector General (OIG) Work Plan is a dynamic priority signal for audits and evaluations, not a complete code list and not immunity for unlisted MS-DRGs. Inpatient clinical documentation integrity (CDI) programs answer that risk with targeted DRG audits, query monitoring, and clinical validation, not with query language that steers severity for payment.
This independent OpenExamPrep section helps learners study inpatient DRG compliance as it appears in Domain VII of the Association of Clinical Documentation Integrity Specialists (ACDIS) Certified Clinical Documentation Specialist (CCDS) examination. It is not an ACDIS or OIG product, and OpenExamPrep does not claim partnership, official review, or approval by ACDIS, OIG, or the Centers for Medicare & Medicaid Services (CMS).
Why Domain VII treats DRG assignment as an ethics problem
The Inpatient Prospective Payment System (IPPS) pays a predetermined amount based on the MS-DRG the grouper assigns from coded diagnoses, procedures, and a handful of other claim elements. A legitimate CDI program improves the accuracy of that assignment so the claim reflects the condition established after study as chiefly responsible for occasioning the admission—the Uniform Hospital Discharge Data Set (UHDDS) principal-diagnosis definition—and the secondary conditions that met UHDDS reporting criteria. The same tools—queries, education, working-DRG review—can be misused to inflate relative weight. When that inflation is systematic, oversight writing calls the pattern DRG creep (enforcement documents often say upcoding).
A single unsupported MCC is a documentation or coding error. Creep is the trend: case-mix index (CMI) or high-weighted DRG volume rising because of how the hospital documents and codes, not because the patients got sicker. Rising CMI after a well-governed CDI program with clinical validation is expected. Rising CMI after a template that always offers the highest-weighted diagnosis first, omits a genuine alternative, or cites reimbursement as the reason to query, is a compliance event.
What DRG creep looks like on an inpatient unit
Creep rarely announces itself as “code the higher DRG.” It shows up as process:
- Principal diagnosis chosen because it maps to a richer family—for example, treating a localized infection as sepsis without a provider diagnosis of sepsis and without supporting organ-dysfunction documentation when severe sepsis is reported.
- Secondary diagnoses reported as CC/MCC when they are integral to the principal diagnosis, ruled out, historical only, or not clinically significant for this stay.
- Procedure duration reported from a clock the record cannot support. The classic inpatient example is invasive mechanical ventilation greater than 96 consecutive hours, which can split certain respiratory and sepsis MS-DRGs.
- Query design that makes the MCC the path of least resistance: the first option is always the highest-weighted condition, clinical indicators are thin, or the prompt mentions payment or quality scores.
The CCDS exam tests whether you can separate compliant severity capture from creep. The record must support the code under the ICD-10-CM/PCS Official Guidelines for Coding and Reporting and UHDDS definitions. Clinical criteria (Sepsis-3, Academy of Nutrition and Dietetics/American Society for Parenteral and Enteral Nutrition or GLIM malnutrition criteria, and similar) may be cited neutrally at the end of a query template. They do not replace a provider diagnosis, and they do not authorize the CDI specialist to “make the patient sicker on paper.”
OIG oversight is not the same as a MAC denial
OIG is HHS’s independent oversight arm. It conducts audits, evaluations, and investigations; it issues compliance program guidance; and it can pursue administrative Civil Monetary Penalties Law remedies and exclusions, often in parallel with Department of Justice False Claims Act cases taught in the next section. Medicare Administrative Contractors (MACs), Recovery Auditors, and the Comprehensive Error Rate Testing (CERT) program are payment-integrity contractors with different sampling and recoupment rules (Domain I). For Domain VII, remember the role: OIG sets public enforcement priorities and publishes findings that hospitals, contractors, and plaintiffs’ counsel all read.
OIG hospital-compliance reviews have long sampled inpatient claims that look at risk for overpayment—selected MS-DRGs, short stays, transfer billing, and coding that changes payment when a diagnosis is removed. Those reviews routinely recommend that the hospital identify, report, and return related overpayments under the statutory overpayment timetable. CDI is in that loop because the “related” claims are often the same documentation pattern you see every day.
The Work Plan is a method, not a codebook
OIG describes work planning as dynamic. Staff identify risk in HHS programs, an engagement committee approves projects, approved work is posted, the audit or evaluation is performed, a report issues, and OIG tracks whether CMS or the provider implements recommendations. Projects are added, revised, completed, or removed during the year. Some titles stay redacted until the report publishes.
Use it this way:
- Read current Work Plan items that mention inpatient hospital billing, MS-DRG, sepsis, malnutrition, ventilation, or coding.
- Translate each item into a local audit question: does our documentation support this code on a statistically valid sample?
- Pair the item with internal data—CMI by MS-DRG family, query yield, denial reasons, and Program for Evaluating Payment Patterns Electronic Report (PEPPER) percentiles (the compliance overlay for PEPPER is the next section).
- Do not treat silence as safety. A DRG that is not named this quarter can still be false if the record does not support it.
Do not treat the Work Plan as the ICD-10 code set, a hospital-specific “do not bill” list, ACDIS exam content, or a promise that unlisted MS-DRGs will not be audited by a MAC, Recovery Auditor, CERT, or a future OIG project.
This guide does not print a 2026 “target DRG list.” OIG’s posted inventory changes. Memorizing last year’s titles as if they were this year’s codes is the mistake the exam is built to catch. If a question stem asks what the Work Plan is, the defensible answer is a priority signal. If it asks what the Work Plan is not, the defensible answers are a code list, a cut score, a CMI target, and an accreditation process.
Historical examples of the type of issue OIG has scrutinized
The following rows are examples of issue types that have appeared in OIG audits, evaluations, or Work Plan postings. They illustrate method. They are not a 2026 Work Plan inventory unless you have verified the live posting yourself.
| Issue type (historical example) | What oversight was testing | CDI/compliance takeaway |
|---|---|---|
| Sepsis inpatient billing | Whether hospitals bill a broader sepsis construct than the clinical definition in use, and how billing varies across hospitals | Principal-diagnosis sepsis needs a provider diagnosis plus indicators; SIRS alone is not Sepsis-3; do not upgrade pneumonia or UTI to sepsis from labs alone |
| Severe protein-calorie malnutrition as MCC (OIG’s later audits sampled codes such as E41 and E43 on claims where removing the code changed the MS-DRG) | Whether the severe malnutrition code was supported, or whether a less severe code—or no malnutrition code—was appropriate | Severe malnutrition can be an MCC; auditors expect recognized criteria plus provider documentation, not albumin or a dietitian note standing alone |
| Kwashiorkor (classic ICD-9 code 260 hospital reviews) | Adult U.S. inpatients billed with a diagnosis that is rare outside famine settings | If the coded condition is epidemiologically implausible, clinical validation is mandatory before the bill |
| Invasive mechanical ventilation >96 hours (MS-DRGs that require the hour threshold, including respiratory and sepsis families) | Whether consecutive hours and the ICD-10-PCS ventilation code matched the record | Count hours from the record; a start date several days before discharge does not automatically prove more than 96 consecutive hours |
Those four rows share a structure: a payment-sensitive code or MS-DRG, a documentation gap that is easy to industrialize, and an oversight finding that the hospital should have known the record did not support the higher payment. That structure is what you study—not trivia about current project numbers.
DRG compliance initiatives that actually change risk
A CCDS-level program does not “monitor CMI” and stop. It runs controls.
Targeted DRG audits. Sample MS-DRG families where payment jumps when one code is added or removed—MCC splits, ventilation-hour splits, sepsis versus localized infection. Use a clinician or CDI reviewer plus a coder. Score principal diagnosis, secondary UHDDS criteria, present-on-admission when relevant, and PCS duration. Extrapolate only with a method your compliance office owns.
Query monitoring. Review a sample of queries for leading language, missing clinical indicators, missing open-ended Other, please specify (required on multiple-choice queries under the ACDIS/AHIMA 2026 Guidelines for Achieving a Compliant Query Practice), reimbursement or quality-outcome language, and technology-generated prompts that fail the same standard. Track agree rates that are implausibly high for MCC-producing queries. A query program that never receives “not clinically significant” is not a quality program; it is a capture engine.
Clinical validation. When a diagnosis is documented but the record lacks support, the compliant move is a clinical validation query—multiple-choice with other / ruled out / no longer valid / confirmed with support / alternative diagnosis—not silent deletion by the CDI specialist and not silent retention “because the attending wrote it.” Validation is how you prove CMI movement is real severity, not creep.
Governance. Written policies that CDI goals include integrity, not a CMI target. Escalation to compliance when a pattern appears. Education that uses de-identified cases. If an audit finds overpayments, the hospital’s identified-overpayment process—not a quiet recode of one chart—owns the rest.
Worked pattern (composite)
A hospital’s sepsis MS-DRG volume rises after CDI launches a “lactate plus SIRS” screen that auto-suggests sepsis as principal diagnosis. Query agree rate for that template is 96 percent. Denials for clinical validity of sepsis climb. This is the creep pattern even if every query looks polite. The compliance response is to stop the screen, rewrite the query to require a provider diagnosis and organ-dysfunction linkage where severe sepsis is at issue, sample paid claims, and involve compliance in any repayment. Checking whether sepsis currently appears on the OIG Work Plan is useful prioritization. It is not the legal test of whether the claims were supportable.
Exam traps
- Equating any CMI increase with fraud. Legitimate capture exists.
- Treating the Work Plan as a safe harbor for unnamed MS-DRGs.
- Inventing a memorized list of “this year’s OIG DRGs” instead of describing how OIG signals priority.
- Believing a signed attending note immunizes an unsupported MCC.
- Using leading queries as a “compliance initiative” because they were labeled educational.
A CDI manager sees CMI rise after a new query template that always lists severe protein-calorie malnutrition first whenever albumin is low, with no alternative malnutrition severity and no open-ended option. Which description best captures the compliance concern?
A CDI educator is building next quarter’s audit calendar. Which use of the HHS OIG Work Plan is appropriate?
A hospital copies last year’s OIG Work Plan into policy and treats any MS-DRG not named there as cleared for billing. What is the error in that policy?
Which DRG-compliance initiative most directly tests whether a documented MCC is clinically supported before the bill drops?