GI Infections: GI Infections: Diagnosis to Surgery

0:00Welcome to Board Pearls. This is episode one of five of the GI Infections chapter, in the Special Populations and Acute or Supportive Care module. This episode follows one Clostridioides difficile infection from the first loose stool to the operating room. The first patient is a woman in long-term care who finished clindamycin, and her case shows who should be tested, why the test takes two steps, how two laboratory numbers decide severity, why fidaxomicin is now the first drug and metronidazole the last, why intravenous vancomycin does nothing for the colon, and when the surgeon is called. The second patient is a man after hip surgery, and his case shows why a positive test can mean less than it seems. The episode ends with why she isn't retested once she's well.

0:45Here's the patient. An eighty-one-year-old woman in long-term care finished clindamycin for a dental infection three weeks ago. For four days she's had six to eight watery stools a day, with cramping and a low-grade fever. She takes a daily proton pump inhibitor and no laxatives. Her white count is seventeen thousand and her creatinine is one point two.

1:06She most likely has Clostridioides difficile infection, CDI, a toxin-mediated colitis that typically follows antibiotic exposure within the prior eight to twelve weeks. Her clindamycin falls inside that window and belongs to the highest-risk classes, along with fluoroquinolones, third-generation cephalosporins, and broad-spectrum penicillins. She also carries three of the other risk factors: advanced age, long-term care residence, and proton pump inhibitor use. The rest of that list, hospitalization, severe comorbid illness, inflammatory bowel disease, immunosuppression, malignancy or chemotherapy, and prior CDI, is worth knowing because each one raises suspicion in the next patient. CDI causes about five hundred thousand infections and thirty thousand deaths a year in the United States, with hospital-onset and community-onset cases roughly equal, so her facility isn't the only setting where it's found.

2:04Her symptoms fit how CDI usually looks: watery diarrhea, sometimes with mucus, abdominal cramping, low-grade fever, and leukocytosis. Gross blood is unusual in CDI. The presentation that's easy to miss is a right-sided variant centered on the cecum, with fever of unknown origin, leukocytosis, and right lower quadrant pain or tenderness but little or no diarrhea. Had she come in with fever, a high white count, and right lower quadrant tenderness and only a few stools, C. difficile would still belong on her list. If she were scoped, pseudomembranous colitis, yellow plaques with abrupt transition zones, would be essentially diagnostic, with necrotic surface epithelium, fibrin and mucin caps, and a neutrophilic infiltrate on biopsy.

2:52Before she's tested, she has to meet the testing threshold, because a test sent on the wrong patient creates the wrong diagnosis. Testing is done only with a high pre-test probability: three or more unexplained watery stools a day, in a patient not taking laxatives, with classic risk factors. Testing formed stool or asymptomatic patients has been documented as a major cause of overdiagnosis, with false-positive results leading to unnecessary antibiotics. She has six to eight stools a day, takes no laxatives, and has clindamycin and her age and facility behind her, so she's tested.

3:28Her test has two steps because no single assay is both sensitive and specific. Nucleic acid amplification testing, NAAT, for the toxin genes tcdA and tcdB is highly sensitive for the organism but can't tell active toxin production from harmless colonization. Glutamate dehydrogenase antigen, GDH, is similarly sensitive and similarly nonspecific. Toxin enzyme immunoassay, EIA, is specific for active disease but misses cases. So the algorithm screens with NAAT or GDH and confirms with toxin EIA. Two positives establish CDI, two negatives effectively exclude it, and a positive NAAT with a negative toxin needs clinical correlation, because it often means colonization. When pre-test probability is high, as hers is, NAAT alone is acceptable; when it's lower, the two-step algorithm prevents overdiagnosis. Her NAAT and toxin EIA are both positive, so she has CDI.

4:31The second patient shows what that discordant result looks like at the bedside. A fifty-two-year-old man is on day three after a hip replacement. He received perioperative cefazolin and takes a stool softener and senna for opioid constipation. He's had three loose stools today, his NAAT is positive and his toxin EIA negative, and his team wants to treat. He probably doesn't have CDI. His laxatives explain his stools, so he never met the testing threshold, and a positive NAAT with a negative toxin often means colonization. The right step is to stop his laxatives and reassess his stools before giving any C. difficile therapy. Treating him would expose him to antibiotics for an infection he most likely doesn't have.

5:16Once her diagnosis is confirmed, the next decision is how sick she is, because severity sets treatment intensity. Severity rests on two numbers. Nonsevere disease has a white count below fifteen thousand and a creatinine below one point five. Severe disease meets either threshold, a white count of fifteen thousand or more or a creatinine of one point five or more, so one abnormal number is enough. Fulminant disease adds hypotension or shock, ileus, or megacolon, and it's the stage where the surgeon is consulted at presentation rather than after treatment fails. Her white count of seventeen thousand makes her severe, even though her creatinine is normal. Without hypotension, ileus, or megacolon, she isn't fulminant.

6:01Because she's severe but not fulminant, she gets the same first-line drug as a nonsevere patient. Fidaxomicin, two hundred milligrams by mouth twice daily for ten days, is preferred for nonsevere and severe non-fulminant disease in the twenty twenty-one updates from the American College of Gastroenterology and from the Infectious Diseases Society of America with the Society for Healthcare Epidemiology of America. It cures about as often as vancomycin, but sustained recurrence falls by about fourteen percentage points, probably because its narrow spectrum spares Bacteroides and other commensal flora. Oral vancomycin, one hundred twenty-five milligrams four times daily for ten days, is the alternative when fidaxomicin is unavailable or unaffordable. It works because it's essentially unabsorbed and reaches the colonic lumen at therapeutic concentration. She starts fidaxomicin.

6:53Metronidazole would be the wrong choice for her, and the reason is where the drug goes. Metronidazole, five hundred milligrams three times daily for ten days, is reserved for the lowest-risk patients when neither preferred agent is available. Taken by mouth, it's rapidly absorbed, so the amount that reaches the colon depends on biliary and inflammatory secretion back into the lumen. That amount falls as the colitis heals, leaving the drug least active just when cure is being completed, and its disadvantage is greatest in severe disease like hers. It also carries a Food and Drug Administration boxed warning for carcinogenicity in mice and rats, and prolonged use is linked to peripheral neuropathy and other neurotoxicity, which limits how much of it a patient can receive over time.

7:35Then she worsens. On day two of fidaxomicin she becomes hypotensive and distended, stops passing stool, and imaging shows ileus. She now has fulminant disease, and her treatment changes to high-dose oral vancomycin, five hundred milligrams every six hours, plus intravenous metronidazole, five hundred milligrams every eight hours as an adjunct, with early surgical and infectious diseases consultation. Her ileus means oral drug may never reach the colon, so rectal vancomycin, five hundred milligrams in one hundred milliliters of saline every six hours, is added. Intravenous vancomycin is never appropriate in CDI, because it doesn't reach the colonic lumen at therapeutic concentration. The same property that makes oral vancomycin work, staying in the bowel, means the intravenous form stays out of it, so an intravenous vancomycin order written for her colitis is the trap.

8:29Her surgeon is already involved, so the question is whether she needs an operation and which one. Surgery is indicated for perforation, refractory hemorrhage, megacolon with progressive distention, or persistent sepsis despite optimal medical therapy. The two operations trade source control against keeping the colon. Subtotal colectomy with end ileostomy removes the toxin source entirely and suits the sickest patients with established necrosis or perforation, where leaving the colon means leaving the toxin behind. Diverting loop ileostomy with intraoperative colonic lavage and vancomycin instilled from above diverts the fecal stream while delivering high-concentration drug to the colon, letting it heal in selected stable patients. If her colon kept dilating and her sepsis persisted, she'd get a colectomy if she were unstable with necrosis, and a loop ileostomy with lavage if she were more stable with her colon intact.

9:22She recovers on medical therapy, and the last decision is not to retest her. Test of cure isn't recommended, because up to half of treated patients with resolved symptoms still carry detectable nucleic acid in stool, from dead organisms or persistent colonization. A positive test after cure would also tempt her team to call later symptoms a recurrence, when post-infection irritable bowel syndrome, which affects about a quarter of patients after CDI, is a common explanation. True recurrence, a return of disease within eight weeks of completed therapy, affects twenty to thirty percent of first episodes and forty to sixty percent after a second, and it's the subject of episode two.

10:01Here are the traps the boards set in these patients. A patient taking laxatives, or a patient with formed stool, shouldn't be tested for C. difficile. A positive NAAT with a negative toxin EIA often means colonization and isn't treated as infection on its own. The cecal variant can present with fever, leukocytosis, and right lower quadrant pain and almost no diarrhea. A single threshold, a white count of fifteen thousand or a creatinine of one point five, is enough to make disease severe. Metronidazole isn't chosen when fidaxomicin or oral vancomycin is available. Intravenous vancomycin never treats CDI. When ileus blocks oral delivery, rectal vancomycin is added. In fulminant disease, the surgeon is called at presentation rather than after failure. And a patient who has recovered isn't tested for cure.

10:51So, start to finish. Test only a patient with three or more unexplained watery stools a day who isn't taking laxatives. Screen with NAAT or GDH, confirm with toxin EIA, and read a positive NAAT with a negative toxin as possible colonization. A white count of fifteen thousand or more, or a creatinine of one point five or more, makes the disease severe, and hypotension, ileus, or megacolon makes it fulminant.

11:18Treat a first episode with fidaxomicin, two hundred milligrams twice daily for ten days, or oral vancomycin, one hundred twenty-five milligrams four times daily, and use metronidazole only when neither is available. Fulminant disease gets vancomycin, five hundred milligrams every six hours by mouth, and by rectum when there's ileus, plus intravenous metronidazole and an early surgical consult. Intravenous vancomycin is never used. Operate for perforation, refractory hemorrhage, progressive megacolon, or persistent sepsis, and don't test for cure.

11:52Episode two is recurrent C. difficile.

11:56For the full chapter, the practice vignettes, and the topic-tagged question bank, head to board pearls dot com. You'll find the rest of the series on Apple Podcasts, Spotify, or wherever you listen to podcasts. That brings us to the end of episode one of five of chapter thirty-three, and I'll see you in the next one.