16.2 Integral Signs/Symptoms, the With Convention, and Uncertain Diagnoses

16.2 Integral Signs/Symptoms, the With Convention, and Uncertain Diagnoses

Quick Answer: Do not assign separate codes for integral signs and symptoms of a diagnosed disease. The words with and in in the ICD-10-CM Index or code title presume a causal relationship unless the record says the conditions are unrelated (or another guideline requires an explicit link). For inpatient discharges, possible / probable / likely (and similar) diagnoses documented at discharge are coded as if established. The 2026 ACDIS/AHIMA query guidance: use uncertain-diagnosis queries sparingly.

This independent OpenExamPrep teaching follows the ICD-10-CM Official Guidelines for Coding and Reporting and the August 2026 ACDIS/AHIMA Guidelines for Achieving a Compliant Query Practice. It is not an official ACDIS, AHIMA, or Cooperating Parties publication.

Integral signs and symptoms: when not to add a code

ICD-10-CM allows signs and symptoms when a related definitive diagnosis has not been established. Once a definitive diagnosis is established, signs and symptoms that are routinely associated with that disease process are not assigned as additional codes, unless the classification specifically says to add them.

Integral means the finding is part of how that disease presents, not a separately reportable problem. Typical inpatient examples:

• Cough, fever, and leukocytosis with pneumonia (unless a guideline or note tells you to code a specific associated condition).
• Chest pain that is the presenting symptom of acute coronary syndrome or unstable angina once that diagnosis is established.
• Dyspnea that is the expected manifestation of the acute respiratory failure or pulmonary edema already being coded as the disease process.
• Nausea with acute gastroenteritis when it is not evaluated or treated as a separate condition.

Not integral means the finding is not routinely part of the disease, or the classification/UHDDS treats it as its own reportable condition because it was evaluated, treated, monitored, or prolonged the stay. Examples CDI actually sees:

• Hyponatremia that is repleted, worked up, and named as a problem in a patient who also has pneumonia. Hyponatremia is not “just pneumonia.” If it meets UHDDS secondary criteria, it is eligible as an additional diagnosis (clinical validation still applies).
• Acute kidney injury in a patient admitted for decompensated heart failure, when the kidney injury is identified, monitored, and treated—not assumed to be “just heart failure.”
• Ileus after surgery when it is more than the expected brief postoperative bowel quietness and is documented as a condition requiring treatment.
• A symptom that points to a different process than the established diagnosis (for example, new focal neurologic findings in a patient whose PD is pneumonia).

The trap on CCDS items is coding every abnormal vital sign or lab as a secondary diagnosis “for severity.” Integral findings do not become CCs by being repeated in every note. Conversely, withholding a non-integral condition that was treated, just because a bigger diagnosis is present, under-reports the stay. Ask: is this finding routine to the disease, or is it a distinct problem the team managed?

If a code title or instructional note says to use an additional code for a manifestation or associated condition, follow the classification even if the finding feels clinically “part of” the disease. The Tabular List can override the general integral-symptom habit.

The with / in convention: presumed causal relationship

Section I.A.15 of the Official Guidelines (the “with” convention) is a high-yield CDI rule. The word with or in should be interpreted as associated with or due to when it appears in a code title, the Alphabetic Index, or an instructional note in the Tabular List. The classification presumes a causal relationship between the two conditions linked by those terms. Those conditions are coded as related even without provider documentation that explicitly says “due to” or “secondary to,” unless:

• the documentation clearly states the conditions are unrelated, or
• another guideline specifically requires a documented linkage between the two conditions.

In the Index, with is listed immediately after the main term (not in alphabetical order among subterms). CDI specialists who only skim progress notes for the phrase “due to” will undercode combination codes the classification already considers related.

Classic applications

Diabetes with stated manifestations in the Index. Type 2 diabetes with chronic kidney disease is a combination (or combination-plus-additional-code) situation because the Index links diabetes with CKD via with. You do not need “diabetic nephropathy” verbatim if both conditions are diagnosed and nothing says they are unrelated.

Hypertension with heart or kidney involvement. The classification links hypertension with heart disease and hypertension with chronic kidney disease through with in the Index. Hypertensive heart disease and hypertensive CKD concepts follow that presumption unless the provider documents that the heart or kidney disease is unrelated to hypertension.

Other Index-linked pairs. Many “X with Y” Index lines (for example, certain infection-with-organism constructions, or disease-with-complication constructions) work the same way: the presumed relationship lives in the classification, not in a required causal sentence.

When you must not presume

Unrelated is documented. Nephrology states chronic kidney disease is due to longstanding hypertension and is not diabetic kidney disease, despite coexisting type 2 diabetes. Do not assign the diabetes-with-CKD combination. Code the conditions as the record describes them.

Another guideline requires linkage. The sepsis guideline is the example CDI must not flatten into the with convention. Severe sepsis in ICD-10-CM requires organ dysfunction linked to sepsis (R65.2- category logic). If acute kidney injury or respiratory failure is attributed to another cause (for example, contrast nephropathy or flash pulmonary edema from systolic failure) and is not associated with the sepsis, you do not build severe sepsis from a silent “with.” That is a documented association rule, not a with-convention shortcut.

The Index never linked them. Two conditions appearing in the same note are not related just because both are serious. Atrial fibrillation and community-acquired pneumonia are not a with pair unless the classification or the provider creates that link. Do not invent etiology.

The documentation is unclear whether they are related. Then the correct move is often a nonleading query, not a silent presumption and not a silent refusal to use the combination. The with convention applies when the Index/title supplies the link and the record does not contradict it. If the record is ambiguous about relatedness, query rather than guessing.

Compliant queries on relatedness must remain nonleading, cite clinical indicators, and avoid reimbursement language. Multiple-choice relatedness queries still need a clinically valid option plus an open-ended Other, please specify path under the 2026 query guidance. Yes/no format is appropriate for cause-and-effect between conditions already documented, not for introducing a new diagnosis.

Uncertain diagnoses at inpatient discharge

If the diagnosis documented at the time of discharge is qualified as probable, suspected, likely, questionable, possible, still to be ruled out, compatible with, consistent with, or similar terms, code the condition as if it existed or was established. This uncertain-diagnosis guideline applies to inpatient admissions to short-term acute, long-term care, and psychiatric hospitals. It does not apply to outpatient encounters (including ED visits that do not become inpatient, clinic, and same-day surgery without inpatient admission).

Code it: Discharge summary: “probable aspiration pneumonia; treat with a full antibiotic course.” For an inpatient claim, assign aspiration pneumonia as if established (with correct POA, covered in 16.3).

Do not code it as established: The same “possible pneumonia” statement on an ED encounter that ends in discharge home. Outpatient coding reports confirmed diagnoses, not possible ones.

Do not code it: “Ruled out pulmonary embolism.” Ruled out is the opposite of uncertain-but-still-in-play. Do not assign the PE code.

Discharge timing matters. The guideline is anchored to documentation at discharge, not to every mid-stay hedge. A day-2 note that says “possible UTI” does not force a UTI code if the discharge summary and hospital course rule it out or never establish it. Conversely, a discharge diagnosis of probable CVA is coded as CVA even if some earlier notes were cautious. Read the discharge diagnostic statement in light of the whole record.

Conflict is different from uncertainty. One attending says possible colitis; the consultant says there is no colitis and imaging is normal. That is a conflict (and often a clinical-validation issue), not a clean probable diagnosis. Query rather than picking the code that helps the working DRG.

2026 brief: uncertain-diagnosis queries used sparingly

The 2026 ACDIS/AHIMA Guidelines for Achieving a Compliant Query Practice (August 2026; supersedes 2022) still require nonleading queries with sourced clinical indicators and no reimbursement or quality-outcome language. On uncertain diagnoses, the practical brief for CCDS candidates is: use uncertain-diagnosis queries sparingly.

That means:

• If the provider already documented possible/probable/likely pneumonia at discharge, the inpatient guideline already tells coding to report pneumonia. A query whose only goal is to upgrade “probable” to “confirmed” so the record “looks better” is usually unnecessary and can become leading if it pressures certainty the clinician does not have.
• Query when uncertainty is not the real problem: conflicting statements (possible versus ruled out), a diagnosis that lacks clinical support (clinical validation), missing POA, or an Index-relatedness question the with convention cannot settle.
• Do not use a yes/no query to introduce a new diagnosis that was never documented. Yes/no may not introduce a new diagnosis under the 2026 guidance.
• If you do query an uncertain diagnostic statement, keep the provider’s independent judgment. Multiple-choice should include Other, please specify. Clinical-validation queries should allow ruled-out / no longer valid / confirmed with support / alternative diagnosis pathways.

Integral symptoms, the with convention, and uncertain diagnoses travel together on the exam because all three are classification rules that CDI cannot override with a preference for extra codes, extra certainty, or extra payment. Apply the rule in front of you: omit the integral symptom, honor the Index with unless unrelated is stated, code the inpatient probable diagnosis, and rarely query solely to erase a compliant uncertainty qualifier.