There is no single ICD-10 code for GI bleed with anemia. In practice you use two codes: one for the bleed and one for the anemia. The usual pairing is K92.2 (gastrointestinal hemorrhage, unspecified) with either D62 (acute posthemorrhagic anemia) when the blood loss is sudden, or D50.0 (iron deficiency anemia secondary to chronic blood loss) when the bleeding has been slow and ongoing. When the source of bleeding is known, a more specific combination code replaces K92.2.
Below I walk through how those codes fit together, how the clinical picture drives the choice, and what the bleeding actually does to the blood count.
The Core Codes at a Glance
ICD-10-CM separates the bleeding event from its consequence. The GI code describes where and how the patient is bleeding; the anemia code describes what that bleeding has done to the anemia picture. Coding only one of them leaves the record incomplete.
| Code | Description | Typical use |
|---|---|---|
| K92.0 | Hematemesis | Vomiting blood, source not yet identified |
| K92.1 | Melena | Black, tarry stools, source not yet identified |
| K92.2 | Gastrointestinal hemorrhage, unspecified | GI bleeding with no documented site or cause |
| K62.5 | Hemorrhage of anus and rectum | Bright red rectal bleeding localized to the rectum or anus |
| D62 | Acute posthemorrhagic anemia | Anemia from sudden, significant blood loss |
| D50.0 | Iron deficiency anemia secondary to blood loss (chronic) | Anemia from slow, ongoing loss that has drained iron stores |
| D50.9 | Iron deficiency anemia, unspecified | Iron deficiency where blood loss has not been documented as the cause |
Choosing Between D62 and D50.0
The anemia code depends on the tempo of the bleed, and the clinician’s documentation has to make that clear. A coder cannot infer acute versus chronic from a hemoglobin number alone.
Acute blood loss: D62
D62 fits the patient who arrives with hematemesis or melena, a falling hemoglobin, and signs of volume loss such as a fast heart rate or dizziness on standing. The note should state something like “acute blood loss anemia” rather than just listing a low hemoglobin.
Chronic blood loss: D50.0
D50.0 fits the patient whose bleeding has been slow and hidden, often found through a positive stool test or unexplained iron deficiency. Here the blood loss has outpaced iron intake over weeks or months, and the lab picture shows small, pale red cells with low ferritin.
Can both apply?
Yes. A patient with a long-standing slow bleed from a colon lesion can present with an acute bleed on top of it. If the documentation supports both chronic iron deficiency and a new acute loss, both D50.0 and D62 can be reported.
When a Specific Source Replaces K92.2
K92.2 is a placeholder for “bleeding from somewhere in the GI tract.” Once endoscopy or imaging identifies the source, ICD-10-CM expects a combination code that captures both the condition and the bleeding in one entry. Using K92.2 alongside a combination code is redundant.
| Source of bleeding | Example code |
|---|---|
| Gastric ulcer, chronic or unspecified, with hemorrhage | K25.4 |
| Duodenal ulcer, chronic or unspecified, with hemorrhage | K26.4 |
| Esophageal varices with bleeding | I85.01 |
| Diverticulosis of large intestine with bleeding | K57.31 |
The anemia code is still added separately. So a bleeding duodenal ulcer with acute blood loss anemia would be coded K26.4 plus D62. Always confirm the exact code and any fifth or sixth characters against the current code set, since sub-codes differ by acuity, perforation, and site.
Why the Bleed Causes Anemia
Red blood cells carry most of the body’s iron inside hemoglobin. Every milliliter of blood lost from the gut takes red cells and their iron with it. The bone marrow normally replaces red cells at roughly 2 million per second, and it can speed up considerably, but only if iron is available.
In an acute bleed, the problem is sheer volume. The hemoglobin may look near-normal in the first hours because plasma and cells are lost together; it drops as the body pulls fluid into the circulation and as IV fluids are given. In a chronic bleed, the volume lost each day is small, but iron stores slowly empty. Once ferritin runs out, the marrow cannot make normal hemoglobin, and iron deficiency anemia follows.
Common causes and risk factors
- Upper GI: peptic ulcers, gastritis, esophageal varices in liver disease, and tears at the junction of the esophagus and stomach.
- Lower GI: diverticulosis, colon polyps and cancer, hemorrhoids, and abnormal blood vessels in the bowel wall.
- Risk factors: NSAID or aspirin use, anticoagulant and antiplatelet drugs, heavy alcohol use, and chronic liver disease.
How the Diagnosis Is Documented
Good coding depends on good documentation, and good documentation depends on a clear workup. The core steps are consistent across most hospitals.
- Complete blood count: hemoglobin and hematocrit quantify the anemia. Normal adult hemoglobin is roughly 13.5 to 17.5 g/dL in men and 12.0 to 15.5 g/dL in women. The platelet count also matters, and our guide to low hemoglobin and platelets covers what it means when both are down.
- Iron studies: low ferritin and low transferrin saturation point to chronic loss and support D50.0.
- Red cell indices: a low mean corpuscular volume (MCV) suggests established iron deficiency; a normal MCV is typical early in an acute bleed.
- Endoscopy: upper endoscopy (EGD) and colonoscopy locate the source, which decides whether K92.2 can be replaced with a specific code.
- Other tests: stool testing for hidden blood, CT angiography, or capsule endoscopy when the source remains elusive.
Treatment Links Back to the Codes
Management addresses both halves of the diagnosis, which mirrors the two-code structure. The bleed is controlled with endoscopic therapy, acid suppression with proton pump inhibitors for ulcers, drugs and banding for varices, or surgery and interventional radiology when needed. Offending drugs such as NSAIDs are stopped where possible.
The anemia is treated according to its tempo. Acute blood loss with instability may need red cell transfusion; most guidelines use a restrictive threshold around 7 g/dL for stable patients, with higher targets in heart disease. Chronic iron deficiency is treated with oral or intravenous iron, and ferritin is rechecked to confirm stores have refilled. For the wider picture of anemia types and treatments, see our anemia guide.
Key Takeaways
- Code the bleed and the anemia separately: typically K92.2 plus D62 or D50.0.
- D62 is for acute blood loss; D50.0 is for chronic loss that has caused iron deficiency.
- Replace K92.2 with a combination code once the bleeding source is documented.
- The physician’s words, not the lab value alone, determine which anemia code is supported.
Frequently Asked Questions
Is there one ICD-10 code that covers both GI bleeding and anemia?
No. ICD-10-CM has no single combination code for GI hemorrhage with anemia. You report a code for the bleed (such as K92.2 or a site-specific code) and a separate code for the anemia (such as D62 or D50.0).
Can I use D62 if the hemoglobin is low but the note does not say “acute blood loss anemia”?
Generally not. A low hemoglobin on its own is a lab finding, not a diagnosis. The treating clinician needs to document the anemia and its link to blood loss; if it is unclear, a coding query to the provider is the right step.
What code applies to melena without a known source?
Melena alone is K92.1, and hematemesis alone is K92.0. If the documentation describes GI hemorrhage more broadly without a site, K92.2 is used. Once a source is found, a specific combination code takes over.
Does a slow GI bleed always cause anemia?
Not always. If the bleeding is minimal and iron intake and stores keep up, hemoglobin may stay normal for a long time. Ferritin often falls first, which is why iron studies can reveal a slow bleed before anemia appears.
