Supratherapeutic INR in the Anticoagulation Clinic
A 64-year-old man with a mechanical mitral valve on warfarin comes to a hospital-based anticoagulation clinic for routine INR monitoring. His INR is 5.8 (therapeutic range 2.5–3.5). He denies bleeding but reports three days of dark, tarry stools and fatigue. Vital signs are BP 118/72, HR 88, and Hgb 9.8 g/dL, down from 11.5 g/dL six weeks ago. The provider documents “elevated INR, anemia, possible GI bleed,” holds warfarin, gives oral vitamin K, and refers to GI. The clinic note does not state whether the anemia is acute or chronic or link it to the GI findings. The patient is discharged home with follow-up. The hospital outpatient department bills a clinic E/M. The CDI specialist is asked to review the encounter for documentation specificity and query opportunity.
1. Does the documentation support coding “possible GI bleed” and “anemia” in the outpatient setting? Explain the coding guideline and its practical effect here.
Read the worked answer
“Possible GI bleed” cannot be coded in the outpatient setting. ICD-10-CM outpatient guidelines prohibit coding diagnoses described as possible, probable, suspected, questionable, rule out, or working diagnosis. The phrase “possible GI bleed” is a differential, not a confirmed condition, so the coder should not assign a GI bleed code from this note. The anemia is different: “anemia” is a confirmed diagnostic statement, but it is nonspecific. A nonspecific anemia code may be assignable, but the record does not say whether it is acute blood loss anemia, chronic anemia, or anemia of another type. The Hgb drop from 11.5 to 9.8 g/dL and the melena are clinical indicators that could support acute blood loss anemia, but the provider has not made that link. The practical effect is that the encounter may be coded with an unspecified anemia and an abnormal INR, while the GI bleed - the clinically important event - is lost. That affects code accuracy, risk stratification, and the clarity of the follow-up plan.
2. Draft a non-leading query that addresses the anemia and the GI finding. Include the clinical indicators you would cite and the options you would offer.
Read the worked answer
Query: “The note documents an INR of 5.8, a Hgb decrease from 11.5 to 9.8 g/dL over six weeks, and three days of dark, tarry stools. You documented ‘anemia’ and ‘possible GI bleed.’ Please clarify: (1) Is the anemia acute blood loss anemia, chronic anemia, or anemia of unspecified type? (2) Is GI bleeding confirmed, suspected, or not present? Clinical indicators include the melena, the Hgb drop, and the supratherapeutic INR on warfarin. You may document the confirmed diagnosis and clinical basis, or state that the diagnosis remains uncertain.”
This is non-leading because it does not tell the provider to document acute blood loss anemia or GI bleed. It presents the objective findings, asks for a clinical determination, and includes an “uncertain” option. The query should be attached to the specific encounter and should not suggest that a particular answer is needed for reimbursement. If the provider confirms acute blood loss anemia and GI bleeding, the coder can assign those codes; if the provider says the anemia is chronic and the GI bleed is not confirmed, the coder must follow that documentation.
3. The provider responds, “Anemia likely due to GI bleed; no endoscopy yet.” Can the CDI specialist code acute blood loss anemia and GI bleed? What additional documentation is needed, if any?
Read the worked answer
The response is still uncertain. “Likely due to GI bleed” is a probable linkage, not a confirmed diagnosis. In the outpatient setting, the CDI specialist cannot code GI bleed from “likely,” and should not code acute blood loss anemia merely because the Hgb dropped. The provider needs to state the confirmed diagnosis, if one exists. A follow-up query could ask: “You documented that the anemia is likely due to GI bleeding. For outpatient coding, please clarify whether acute blood loss anemia is confirmed and whether GI bleeding is confirmed, suspected, or not confirmed. If confirmed, please document the clinical basis.”
No endoscopy is not required to diagnose a GI bleed; melena and a Hgb drop can be sufficient clinical evidence. But the provider must make the diagnosis, not the CDI specialist. If the provider confirms acute blood loss anemia and GI bleeding, both may be coded. If the provider continues to document only “possible” or “likely,” the coder should use the confirmed anemia code that is supported - commonly an unspecified anemia - and should not assign a GI bleed code. The CDI specialist should also avoid pressing for “acute blood loss anemia” as the answer; the query must allow chronic anemia or uncertain etiology.