7.3 Medications, Diagnostic Tests, and Abbreviations as Clinical Indicators
Key Takeaways
- Medications such as megestrol, vasopressors, insulin infusions, and lactulose are clinical indicators, not standalone diagnoses
- Laboratory values and imaging reports support queries; they do not replace a provider diagnostic statement
- Official Guidelines (Section III.B): abnormal findings are not coded and reported unless the provider indicates their clinical significance
- If a finding is outside the normal range and the provider has ordered further testing or treatment, it is appropriate to ask whether the finding should be added
- Standard abbreviations save time; ambiguous abbreviations (MS, PT, and similar) are dangerous and should be queried rather than guessed
7.3 Medications, Diagnostic Tests, and Abbreviations as Clinical Indicators
Quick Answer: Read drugs, labs, and imaging as clinical indicators. Megestrol (Megace) points toward poor appetite, not toward a coded malnutrition diagnosis by itself. Vasopressors point toward shock type that still needs a name. Insulin infusions point toward severe hyperglycemia, diabetic ketoacidosis, or a hyperosmolar state. Lactulose points toward hepatic encephalopathy. The ICD-10-CM Official Guidelines (Section III.B) say abnormal findings (laboratory, x-ray, pathologic, and other diagnostic results) are not coded unless the provider indicates their clinical significance. Ambiguous abbreviations must be clarified, not decoded in favor of an MCC.
This independent OpenExamPrep section helps inpatient CDI specialists study indicator literacy for the CCDS exam. It is not an ACDIS product and does not claim ACDIS approval or partnership. The CCDS sitting allows DRG Expert and one listed drug handbook; those books help you interpret what a drug is used for. They still do not let you assign a diagnosis from the medication administration record (MAR).
Drugs are clues, not codes
A drug on the MAR tells you what the team is treating or preventing. It does not tell the grouper what to report. The provider diagnostic statement, plus UHDDS relevance for secondary diagnoses, still governs the claim.
| Medication | What it commonly indicates | What it does not prove by itself |
|---|---|---|
| Megestrol acetate (Megace) | Appetite stimulation; possible malnutrition, cachexia, or cancer-related wasting under evaluation | Severe protein-calorie malnutrition or cachexia as a coded diagnosis |
| Norepinephrine, vasopressin, epinephrine, phenylephrine, dopamine | Shock physiology; look for type (septic, cardiogenic, hypovolemic, obstructive) | Septic shock or any shock type without a provider name |
| Insulin infusion | Severe hyperglycemia, diabetic ketoacidosis, hyperosmolar state, or perioperative glucose control | DKA or hyperosmolar crisis from the drip rate alone |
| Lactulose, rifaximin | Hepatic encephalopathy or constipation in cirrhosis | Hepatic encephalopathy without a provider diagnosis |
| Naloxone | Opioid toxicity or oversedation | Poisoning versus adverse effect versus intent |
| Acetylcysteine | Acetaminophen toxicity protocol | A finished poisoning diagnosis and intent |
| CIWA-triggered benzodiazepines | Alcohol withdrawal | Withdrawal delirium or seizures unless documented |
| Vancomycin plus piperacillin-tazobactam | Concern for serious bacterial infection | Sepsis or severe sepsis |
| Stress-dose hydrocortisone | Septic-shock adjunct or adrenal insufficiency | Adrenal crisis |
| Thiamine | Nutritional risk in alcohol-use disorder | Wernicke encephalopathy |
Original scenario. Ms. N. has a body mass index of 17.4, a registered dietitian nutrition diagnosis of severe protein-calorie malnutrition, and a new megestrol order. No provider has diagnosed malnutrition. Megestrol is an indicator. The dietitian note is an indicator. The August 2026 ACDIS/AHIMA query brief allows Academy/ASPEN or GLIM criteria to be cited neutrally at the end of a template. It does not allow a yes/no query that introduces malnutrition from an RD note, and it does not allow “please document the MCC.” Use multiple-choice with clinically valid options and Other, please specify.
Original scenario. Mr. V. is on norepinephrine at 0.12 mcg/kg/min. The note says hypotension, fluids given. CDI may query for the type of shock if lactate, infection source, or cardiac failure indicators support a diagnosis. CDI may not assign septic shock from the infusion. If the provider documents cardiogenic shock after an anterior infarction, do not add septic shock just because a pressor is hanging.
Labs and imaging: the same rule, different pixels
Lactate, procalcitonin, white-cell count, bandemia, culture results, arterial blood gas, troponin, natriuretic peptide, ammonia, CD4 count, HIV RNA, sodium, creatinine, and bilirubin are all query fuel. A CT report of possible abscess, a chest radiograph of infiltrate, or an MRI of restricted diffusion is the same class of evidence: a diagnostic result. Pathology that the attending has not addressed is still a result looking for significance.
Section III of the Official Guidelines governs additional diagnoses in non-outpatient settings. Section III.B, Abnormal findings, states that abnormal findings (laboratory, x-ray, pathologic, and other diagnostic results) are not coded and reported unless the provider indicates their clinical significance. If the findings are outside the normal range and the provider has ordered other tests to evaluate the condition or has prescribed treatment, it is appropriate to ask the provider whether the abnormal finding should be added. The Guidelines note that this inpatient additional-diagnosis practice differs from outpatient coding of diagnostic tests interpreted by a provider. Do not import the outpatient test-coding habit onto an inpatient secondary diagnosis.
Signs and symptoms (Chapter 18 and other symptom codes) remain reportable when a related definitive diagnosis has not been established. That is a different rule from coding a raw sodium of 128 as hyponatremia with no provider statement. If the hospitalist writes “hyponatremia, 3% saline started,” significance is indicated. If the sodium sits at 128, nobody comments, and nobody treats it, you do not code the row. If the sodium is 118, 3% saline is ordered, and the problem list is silent, a query is appropriate.
Original scenario. An initial lactate is 4.2 mmol/L. The emergency department note never mentions lactate. No repeat is ordered. No bicarbonate is given. Do not assign lactic acidosis or severe sepsis from that cell. If a repeat lactate is trending and antibiotics plus fluids are underway for suspected infection, the lactate supports a sepsis query—it still does not assign the diagnosis.
Original scenario. Radiology reads “cannot exclude pulmonary embolism.” The team does not start anticoagulation, does not order a follow-up study, and documents that the filling defect is motion artifact. Do not code pulmonary embolism. If anticoagulation is started and the attending writes “PE,” the imaging was an indicator that became a diagnosis.
Abbreviations: standard ones versus dangerous ambiguity
CDI specialists live in abbreviated notes. Standard, unambiguous abbreviations that the facility recognizes—AKI, CKD, COPD, DKA when the context is diabetes, NSTEMI when the cardiologist has already said it—are ordinary medical writing. The exam still expects you to spell out a term on first use in educational material and to know when an abbreviation is not safe to interpret.
Dangerous ambiguity is the problem. MS can mean morphine sulfate, multiple sclerosis, or mitral stenosis. PT can mean physical therapy, prothrombin time, or patient. DC can mean discharge or discontinue. SS can mean sliding scale or Sjögren syndrome in the wrong specialty note. QD and QOD are on do-not-use lists because they are misread. U for units is misread as 0. MSO4 and MgSO4 are confused with each other. The Joint Commission Do Not Use list exists because these strings have harmed patients. For CDI, the harm is also a wrong diagnosis: coding multiple sclerosis from “MS improving” when the author meant morphine, or coding a coagulopathy from “PT tomorrow” when the author meant physical therapy.
| String | Plausible meanings | CDI action |
|---|---|---|
| MS | Morphine sulfate; multiple sclerosis; mitral stenosis | Query; do not pick the MCC reading |
| PT | Physical therapy; prothrombin time / INR; patient | Use surrounding orders to see the ambiguity; still query if diagnosis is at stake |
| DC | Discharge; discontinue | Confirm before acting on a stopped drug as underdosing |
| NPO, PRN, BID | Standard administration instructions | Usually safe in MAR context |
| AMS, CVA, CHF | Common but unfinished clinical labels | Treat as incomplete diagnoses; query for type |
Original scenario. A progress note says “MS improving, start PT in AM.” The problem list has no neurologic disease. A physical-therapy order is placed. Interpreting MS as multiple sclerosis would manufacture a diagnosis. Interpreting PT as a clotting time would manufacture a coagulopathy. Query if a diagnosis hangs on the string; otherwise treat it as morphine and physical therapy only when the surrounding orders make those meanings unmistakable and no diagnosis is being assigned from the abbreviation.
Unspecified and “not otherwise specified” abbreviations in diagnoses (CHF, CVA, CKD, PNA) are a different CDI task: they are real conditions written without required specificity. That is a query for type and acuity, covered in the body-system chapters—not an invitation to decode a slang abbreviation into the highest-weighted code.
Building a compliant indicator-based query
Source the indicator to a location in the record: MAR, laboratory date and time, radiology impression, pathology, or nursing flowsheet. State what you observed without diagnosing. Offer only options the numbers can support—the 2026 brief’s example is not to offer hypernatremia for a sodium of 122. Include Other, please specify. Leave judgment with the provider. Do not mention CC, MCC, DRG, HAC, SOI, or ROM. If the diagnosis is already written and looks unsupported, use a clinical-validation multiple-choice pattern (confirmed with support, ruled out, no longer valid, alternative diagnosis, other). Verbal queries still need date, time, people, indicators with source, a nonleading statement, options if any, and a response that reaches the permanent health record before coding.
Technology-generated queries follow the same standard. An electronic prompt that says “lactate 4.2 — add lactic acidosis?” is still a query. If it introduces a diagnosis from a lab row with a yes/no, it is the wrong format.
Exam-style habits
On test day you may open a drug handbook to confirm that megestrol is an appetite stimulant or that lactulose is used in hepatic encephalopathy. That lookup supports an indicator, not a code. You may open DRG Expert to see whether a named diagnosis is a CC or MCC. That lookup does not let you skip clinical validation. Practice one page of a chart: circle every drug, every abnormal number, and every abbreviation, then write which of them already have a provider diagnosis, which deserve a query, and which must be left alone under Section III.B. That exercise is the Domain II skill this section is built to teach.
Megestrol acetate is started for poor oral intake. The registered dietitian writes severe protein-calorie malnutrition. No provider has diagnosed malnutrition. What is the correct CDI approach?
An inpatient lactate is 4.2 mmol/L. The treating provider never comments on it, orders no repeat, and starts no treatment directed at that result. According to the Official Guidelines, what should happen?
A progress note reads “MS improving, start PT in AM.” The problem list has no neurologic disease, and a physical-therapy order is placed. What is the safest CDI interpretation?