Sunday, March 23, 2008
Lower GI Bleeds
Ah, the sweet smell of melena. Inhale deep the sharp, metallic odor. Feel it burn the sinuses as it wafts its way slowly down into your lungs. Nothing beats it. You won't ever forget it. As a surgeon, I am lucky to get not so infrequent reminders of its distinctive stench.
Little old lady (LOL) with renal failure, peripheral vascular disease, coronary artery disease, etc, presented to the ER with a liter of marroon currant jelly stool between her legs. She'd been diagnosed with "diverticular bleeds" twice in the past year; both episodes resolved without invasive intervention. She was also on Plavix (the evil anti-platelet medicine) and was due to be dialyzed that day. There's a playbook for these situations, like many things in medicine. No need to think; just run the scripted offense:
1. Airway, breathing, circulation. Get good IV's and run some fluid in.
2. You drop an NG to make sure the source of bleeding isn't proximal to the ligament of trietz.
3. Good rectal exam/anoscopy/proctoscopy to rule out hemorrhoids or bleeding rectal mass.
4. If bleed is slow or has stopped, consider colonscopy
5. Massive bleeds that stabilize--- Tagged red blood cell scan. This is done in nuclear medicine and is pretty sensitive for blood loss that exceeds .1cc/min. It helps localize the bleed to general areas, i.e "right side" or "left side". Most of the time it's useless, but you see it in the algorithms in all the textbooks.
6. Mesenteric Angiography---- Usually the tagged cell scan is done first to help direct the angiogram to the bleeding vessel. This isn't as sensitive a test; the bleeding has to exceed 1cc/min.
7. Surgery---- hemodynamic instability or unrelenting transfusion requirements mandate the cold knife. The tricky part is deciding what sort of surgery to do. Total colectomy? Left colectomy? Hopefully some information can be gleaned from the above studies to help guide your decision.
When I saw this particular lady, she looked like hell. BP 70/30. A continuous lava flow of red stool emanating from between her legs. The bed was saturated. She looked pale and ghostly. He husband sat in the corner reading a magazine. The nephrologist was running the show, giving blood via her dialysis catheters. That's a lot of blood, I told him. Yes, he said. We may need to visit the OR this evening, I said. He said, she's not a candidate for surgery; her coronary artery disease is quite severe. That may be, I replied. But bleeding to death is also problematic.
With three units of blood she sort of, kind of stabilized and went for tagged red blood cell scan. Left colon, about as positive a scan as I've ever seen. Now it's eleven at night and I'm the only one left. I hear the nurse taking an order from the GI consultant for a mesenteric angiogram. I call him back and tell him that the patient is dropping her pressure again and the bloody river continues to flow from between her legs. It's time to stop screwing around and fix the problem. Angiography is notoriously user-dependent in these situations. It's not always successful. Moreover, there's complications; some quote a 20% incidence of colonic ischemia. Plus, it was snowing and the radiologist said it would take two members of his team 45-60 minutes to arrive. No thanks. I'm the one sitting at the patient's bedside. I'm the one whose ass is on the line if she crashes in the angio suite. I booked the case and did the extended left colectomy. She actually did well. I was even able to perform a primary anastomosis.
The point is, playbooks are for shit if you don't look at the patient. I know it's fun to mess around with all these new-fangled tests and non-invasive procedures, but sometimes a timely operation isn't the worst thing in the world.