How to code for gastrointestinal bleeding

Robert S. Gold, MD

Gastroenterologists and other physicians who admit patients to the hospital for work-up or treatment of gastrointestinal (GI) bleeding occasionally leave holes in their documentation.

Hospitals may admit patients because of weakness or "near-syncope."� Staff may encounter very low hemoglobin, leading to transfusion and the initiative to work up the patient for possible occult gastrointestinal bleed.

• If the physician does not tie the presentation with the low hemoglobin and all you see is a GI work-up starting, ask the physician whether the symptoms are due to the anemia of chronic GI blood loss (sometimes, in premenopausal women, it'll be chronic menstrual blood-loss anemia)
• When you see that an upper and/or lower GI endoscopy has found lesions, look for the tie-in between the chronic blood loss and the lesion
• When documentation supports all of these, document the lesion found with hemorrhage as the principal diagnosis

Patients may be admitted because of acute bleeding, upper or lower gastrointestinal tract in origin. It is most important to know the terms for presentations of blood coming out of the bodily GI orifices. Consider the following examples:

1. Hemoptysis"”Refers to the coughing up of blood. The blood may be bright red, streaky, foamy, or just flecks. Patients with chronic cough or chronic inflammation of the respiratory tract may cough blood. Patients with acute irritation problems of the respiratory tract may cough blood. This is seen in nose bleeds, posterior pharyngeal lesions, bronchiectasis, tuberculosis, pneumonias (klebsiella, for example), acute pulmonary edema, and pulmonary embolism. There aren't a lot of GI diagnoses there.

2. Hematemesis"”Refers to vomiting blood. First, you must determine whether you're dealing with a fast bleed or a slow bleed. Is the patient vomiting coffee-ground emesis or bright red blood?

Remember this"”blood exposed to gastric acid turns black. If the bleed is slower, it creates the opportunity for stomach acid to turn blood black. Black vomit comes from an upper GI bleed (gastritis, duodenitis, esophagitis, ulcer of any of these, tumors, angiodysplasias, etc.) It's from the esophagus, stomach, or duodenum.

If a patient vomits red blood, it means the bleeding is so fast that there is no time for the blood to turn black. This occurs in Mallory-Weiss syndrome (laceration of the esophagus from severe retching), fast-bleeding gastric ulcers or gastritis or duodenal ulcers or duodenitis, esophageal (and gastric) variceal bleeds and aorto-duodenal fistula, among others.

The following represent upper GI bleeding:

3. Melena"”Represents black blood from the rectum. It is black, we now know, because the blood reacted with gastric acid. A GI bleed from well below the duodenum will not turn black. When there is melena described, it is an upper GI bleed, caused by the same things as described in item 2 of this list, above.

Can an upper GI bleeding problem cause other than black blood from the rectum? Unfortunately, yes. When a patient has rapid bleeding from a duodenal ulcer or from an aorto-duodenal fistula, that may appear as red or maroon blood from the rectum. This patient will be unstable, requiring rapid transfusion and immediate intervention to prevent death. Also, some patients, especially the elderly or patients post-ulcer surgery, don't make gastric acid. There's nothing to act on the blood to turn it black.

4. Hematochezia"”Designates the bright red blood from colonic angiodysplasia. Either that or "bright red blood per rectum."� Hematochezia can be bright red or maroon. Blood from the rectum can be spotting on the surface of the stool or it can be streaking the surface of the stool (meaning that the bleeding site is the rectum or anus), or it can be mixed in with the stool (meaning that it came from above the rectum). Rapid bleeds yield bright red blood and no stool.

You now know how to find the source of bleeding by looking at the description of the patient's presentation. Now, let's take a look at some of the coding aspects.

Coding Clinic (CC) tells us that 578.9 (hemorrhage of gastrointestinal tract, unspecified) is used when you don't know where a GI bleed is coming from.

As soon as the bleeding site is determined through documentation (CC 2Q 1992), you drop the 578.9 and use the code for the diagnosis "with hemorrhage."� That's a breeze. A patient comes in vomiting blood. The physician says that it's from acute gastritis, so the code is 545.01 (acute gastritis with hemorrhage). A patient comes in with hematochezia. The physician says it's from the patient's diverticula, observed bleeding on colonoscopy, so the code is 562.12 (diverticulosis, colon, with hemorrhage).

The principal diagnosis may also depend on the circumstances of the admission. When the code for the disease process has alternatives of a digit to represent "with hemorrhage,"� then that code is used (CC 4Q 1991).

When the code does not have a corresponding digit to represent "with hemorrhage,"� you may have to sequence the bleed as principal and the diagnosis as secondary if you only work up and treat the bleed (CC Sept/Oct 1985).

But what happens when the physician doesn't draw a straight line from here to there"”from the bleed to the finding? ICD-9-CM tells us that, when a GI bleed leads to admission and a cause of the GI bleed is found, you can make the connection"”usually. Consider the following three examples:

• A patient comes in having had bright red blood per rectum and a negative colonoscopy, but an upper GI study shows gastritis. Can these possibly be related? Not a chance. This must be coded 578.9, GI bleed with no known cause and gastritis without hemorrhage.
• A patient comes in having vomited bright red blood. An upper GI endoscopy is negative, but a colonoscopy shows angiodysplasias of the cecum. Can they possibly be related? Not a chance. This must be coded 578.9, GI bleed with no known cause and angiodysplasia (acquired, by the way) without hemorrhage.

On the other hand, if the patient's presentation lines up with the disease found, then it's likely a slam-dunk to have the diagnosis with hemorrhage assigned. The physician should not object to a query to line up the GI bleed with the pathology.

• Finally, what do you do when a patient arrives for work-up of a GI bleed, and three or four possible causes are found, none of which are bleeding? If the physician tells you which possible cause he or she believes to be the culprit, you're okay.

If the physician doesn't know and believes that it's probably one of them, I would recommend assigning 578.9 as the principal diagnosis and list the pathologic findings "without hemorrhage."� This follows the "comparative or contrasting diagnosis rule,"� wherein the symptom is the principal diagnosis and the possibles or probables are listed as secondaries.