6.2 Malnutrition Criteria, Severity, and Audit Risk

6.2 Malnutrition Criteria, Severity, and Audit Risk

Quick Answer: Treat registered dietitian (RD) assessments as clinical indicators, not diagnoses. Cite Academy of Nutrition and Dietetics / American Society for Parenteral and Enteral Nutrition (Academy/ASPEN) or Global Leadership Initiative on Malnutrition (GLIM) criteria neutrally in a query. Severe protein-calorie malnutrition often groups as an MCC when the provider documents it and the record supports it. Do not use a yes/no query to introduce malnutrition from an RD note alone—use multiple-choice with Other, please specify.

This independent OpenExamPrep section is inpatient CDI study material. It cites Academy/ASPEN and GLIM as publicly used clinical frameworks. Citing a framework in a query is allowed under the 2026 ACDIS/AHIMA brief. Converting a dietitian impression into a coded MCC without a provider diagnosis is not.

Why malnutrition draws auditors

Malnutrition is common on medical and surgical units and is one of the diagnoses payers and reviewers most often challenge. Severe protein-calorie malnutrition (also documented as severe malnutrition or severe protein-energy malnutrition) often acts as an MCC in Medicare Severity Diagnosis Related Group (MS-DRG) grouping when it is reported and not excluded. That severity effect is exactly why Recovery Audit Contractors (RACs), Medicare Administrative Contractors (MACs), and other reviewers test clinical validity. CDI success is not “capturing the MCC.” CDI success is a defensible provider diagnosis, supported by indicators, stated without leading language, and consistent with UHDDS secondary-diagnosis rules.

Exact CC/MCC lists live in the current IPPS tables and can change with the fiscal year. This guide does not reprint those tables. Know the directional fact—severe protein-calorie malnutrition often functions as an MCC—then confirm the current table in practice.

RD notes versus provider diagnoses

Nutrition assessments, calorie counts, subjective global assessment tools, and nutrition diagnoses in the dietitian note are clinical indicators. For ICD-10-CM reporting, the malnutrition diagnosis comes from a provider (physician or other qualified practitioner whose documentation is used for coding). A wound-care nurse’s “malnutrition” and an RD’s “severe malnutrition related to inadequate intake” work the same way: they can trigger a query; they cannot stand alone as the coded diagnosis.

The 2026 ACDIS/AHIMA Guidelines for Achieving a Compliant Query Practice is explicit on format. A yes/no query may not introduce a new diagnosis. Yes/no is appropriate to clarify POA (and must include unable to determine), to substantiate a diagnosis already documented (for example a pathology confirmation), or to ask cause-and-effect between conditions that are already named. Using yes/no solely from an RD or wound-nurse diagnosis to “confirm malnutrition” introduces a new diagnosis and is the wrong format. Use multiple-choice, include at least one clinically valid option, and include a required open-ended Other, please specify (or similar). Do not mention reimbursement, MCC status, or quality scores in the query text.

Verbal queries follow the same clinical standard. Record date, time, people, indicators and source, a nonleading statement, and options if any. The provider’s response must reach the permanent health record before coding.

Academy/ASPEN and GLIM as neutral criteria

You may cite recognized clinical definitions neutrally at the end of the query template. Neutral means you display the criteria as a reference. You do not tell the provider that the patient “meets criteria” and therefore must diagnose a specific severity.

Academy/ASPEN (2012 adult consensus) identifies malnutrition when two or more of six characteristics are present:

Non-severe (moderate) versus severe depends on the degree of those findings in the published tables. CDI specialists should not “upgrade” moderate malnutrition to severe because an MCC is more desirable. That is both a leading-query problem and a False Claims Act pattern risk if unsupported MCCs repeat.

GLIM uses a different structure: one phenotypic criterion (non-volitional weight loss, low BMI, or reduced muscle mass) plus one etiologic criterion (reduced food intake or assimilation, or inflammation / disease burden). Either framework may appear on a query as a cited reference. Do not mix thresholds from both systems in a way that forces a severity the provider did not diagnose.

A neutral closing line sounds like a reference, not an order: “For your reference, adult malnutrition frameworks such as Academy/ASPEN (identification by two or more characteristics) and GLIM (one phenotypic plus one etiologic criterion) are sometimes used in clinical validation. Please document your independent diagnostic impression, including severity if malnutrition is present, or document an alternative finding.” That is the 2026 idea. Do not paste a copyrighted criteria table into the query body as if it were a scoring mandate.

Severity, MCC effect, and clinical validation

When the provider documents severe protein-calorie malnutrition, that diagnosis often groups as an MCC. Moderate malnutrition and unspecified malnutrition do not carry the same typical severity effect. If the attending documents severe malnutrition but the patient has a stable weight, is eating a full diet, has no nutrition-focused physical findings, and receives no nutrition intervention, send a clinical validation query. The 2026 brief treats multiple-choice as the optimal format for validation and expects options such as other, ruled out, no longer valid, confirmed with support, or an alternative diagnosis. Technology-generated queries follow the same standard as human-authored ones.

UHDDS still applies. A malnutrition diagnosis should be evaluated, treated (nutrition support, monitoring, medications), or otherwise meet secondary-diagnosis criteria. A copy-forward problem-list entry from a prior admission, never addressed this stay, is a validation target—not an automatic MCC.

POA for malnutrition is usually Y when the patient is admitted already wasted or with long-standing intake failure. New, hospital-acquired starvation over a long stay is uncommon but possible; do not assume N without a timeline. U is for insufficient documentation. W is clinically undetermined and is paid like Y under the HAC payment provision. Malnutrition is not itself one of the fourteen CMS HAC payment categories, but POA discipline still matters for other conditions on the same case.

Labs and treatments that support a query without becoming the diagnosis

Prealbumin is an acute-phase reactant and falls with inflammation; it is a weak standalone nutrition proof. Albumin tracks illness and dilution more than meal history. Phosphorus and potassium shifts after refeeding, orders for enteral or parenteral nutrition, and documented calorie counts are stronger process indicators. None of them is a coded malnutrition diagnosis. Use them in the indicator paragraph of a multiple-choice query, sourced to the note and date.

Exam-style scenarios

Scenario: RD-only severe malnutrition. The dietitian documents “severe malnutrition, Academy/ASPEN.” The attending never names malnutrition. Do not code from the RD note. Do not send “Does the patient have severe malnutrition? Yes/No.” Send multiple-choice options that include clinically valid nutrition diagnoses, an alternative such as well nourished or unspecified weight loss, and Other, please specify. Cite Academy/ASPEN or GLIM only as a neutral reference at the end.

Scenario: MCC-seeking query language. A template reads, “Please document severe protein-calorie malnutrition so this case groups with MCC.” That sentence is reimbursement language and is noncompliant under the 2026 brief regardless of how accurate the BMI looks. Rewrite to sourced indicators and independent judgment.

Scenario: conflicting severity. The RD says severe. The attending says “well nourished, obesity.” Leaving both in the record without reconciliation is an audit exhibit. Clinical validation multiple-choice can ask whether malnutrition is confirmed with support, ruled out, no longer valid, or replaced by an alternative diagnosis.

Scenario: copy-forward MCC. Last year’s discharge summary listed severe malnutrition. This stay shows a BMI of 31, a regular diet, and no nutrition consult. Query for this-stay validity. UHDDS is about this encounter, not a fossilized problem list.

Audit patterns to anticipate

Reviewers look for severe malnutrition documented only in an RD note; severe malnutrition with no weight trajectory, intake data, or physical findings; queries that list only “severe malnutrition” and “undetermined”; queries that mention CC/MCC, DRG, or quality bonus; POA errors on malnutrition that was clearly chronic at the inpatient order; and unreconciled severity conflicts. Build the habit of pairing every malnutrition query with sourced indicators and a format that cannot introduce the diagnosis from ancillary staff alone.