Infectious Disease Compendium



Anaerobic gram positive rod; includes C. botulinum, C. difficile, C. novyi, C. perfringens, C. ramosum, C. septicum, C. sordellii, C. tetani. Some are spore formers and exist forever in the environment.

Epidemiologic Risks

Most Clostridia are found in soil and gi tracts.

Eating Clostridia botulinum contaminated food: home canned is most common. Also occurs from drinking Pruno,(aka hooch (PubMed) an illicit fruit based alcohol brewed in prison toilets. I can't make this stuff up (PubMed). Black tar heroin is associated with wound botulism (PubMed) and some will get it more than once. Can you say heroin is addicting? And home fermented tofu (PubMed). Relatively common in Alaska, and often associated with aquatic game foods, including "foods that had been collected from decomposing whale or seal carcasses found on beaches or floating at sea (PubMed)."

C. difficile is driven by antibiotics, with clindamycin and quinolones being the most likely suspects, but any antibiotic can do it. However, cumulative exposure to antibiotics is also a risk (PubMed).

But tetracyclines are protective (PubMed).

Asymptomatic carriage rates are about 12% (PubMed).

Hospitals may till the soil, metaphorically speaking, with antibiotics, but the source is usually NOT the hospital, with 35% of cases genetically linked (PubMed) and in another study 75% of hospital associated C. difficile is not hospital acquired (PubMed); they must have brought it in with them. 15% of patients will have C. difficile colonization on admission (PubMed).

They are not getting it from us, ie health care workers, where in one study zero, nip, nil, nada were colonized with C. difficile (PubMed).

Curiously, doxycycline may be protective when given with other antibiotics (PubMed) as is rifampin.  Being on metronidazole for another reasons decreases odds by 80% (PubMed).

Probiotics are beneficial in preventing C. difficile, in one meta analysis by 66% (Reference) are probaby cost effective (PubMed).

The closer the probiotic is given to the first dose of antibiotic, perhaps the better the prevention (PubMed).

C. difficile can be found in up to 20% of ground meat from stores (PubMed), salads and veal. I bet it is found in damn near everything we eat, since everything we eat has a fine patina of human and/or animal stool. It is more a food borne infection than most people suspect.

And Spanish sandboxes. Along with Salmonella. Sandboxes are a disgusting soup of pathogenic fecal flora. Which defines children when you think about it.

Obesity is associated with a different bowel microbiome than is seen in thin people and may be why obesity is a risk (PubMed).

C. difficile can be nosocomial and, pay attention here me bucko, IS NOT KILLED BY THE ALCOHOL FOAM. Sorry I yelled. But you have to wash your hands. Since the turn of the century there is a new strain of C. difficile being driven by quinolone use, both for inpatient and outpatient disease (PubMed). This strain causes severe disease, with toxic megacolon and death not infrequent.

Diarrhea is common in the ICU, it is usually neither infectious or C. difficile. Not that C. difficile isn't an infection(PubMed).

PPI use is a risk for disease and relapse (44% chance of relapse); as the hippies knew, acid is good, man (PubMed). PPI use is a risk in every study (PubMed).

C. difficile can cause diarrhea with no classic risk factors. In Maryland, as an example, 43 (3.9%) of outpatients had C. difficile as a cause of their diarrhea. 7 had no recognized risk factors, and 3 had neither risk factors nor co-infection with another enteric pathogen (PubMed). "Nearly half of Medicare beneficiaries admitted with CACD have no recent antibiotic exposure (PubMed)."

The incidence of multiple recurrences is going up (PubMed).

Risks to family low, 5 of 1061 spouses and 3 of 501 children developed C. difficile within three months of the index case (PubMed).

Nurses will often say that stool smells like C. difficile, and I have always poo-pooed the idea. All stool stinks (but mine) but if a dog can be trained to sniff out C. difficile (poor beast) why not a human (PubMed)?


- C. botulinum: botulism. A and B is contaminated food (besides canned goods has occurred with fresh carrot juice (PubMed), that's the problem with being a vegan) and wounds (has occurred in cocaine users, growing in a sinus), F is from fish. There have also been cases with the cheap and the vain (that would make a good movie title) using laboratory grade botulism toxin, which is far more concentrated than the commercial preparation. Progressive weakness usually starting with ocular muscles.

There is also the newly discovered (2013) H for which there is no anti-toxin (PubMed).

Also occurs from drinking Pruno, illicit fruit based alcohol brewed in prison toilets. I can't make this stuff up (PubMed).

- black tar heroin is associated with wound botulism (PubMed) and some will get it more than once. Can you say addicting?

- C. difficile (2015 NEJM Review): pseudomembranous or antibiotic associated colitis.

The current molecular testing is at least 90% but "Considering that up to 50% of institutionalized individuals may be asymptomatically colonized by toxigenic and that diarrhea may have a variety of causes, false-positive results are likely, especially given the high sensitivity of molecular approaches. Because it is likely that asymptomatic colonization by toxigenic protects patients from CDI, inappropriate therapy under these circumstances may put the patient at greater risk for CDI at a later time. In addition, a significant proportion of patients who have been successfully treated for CDI may have persistent asymptomatic colonization for many weeks (PubMed)."

It is why, with the new assays, you DO NOT test on solid or semi solid stool: "An important side issue here is that many physicians insist that the laboratory test solid and semisolid specimens. It is imperative for laboratories to create and enforce specimen guidelines. Appropriate specimen collection is paramount to CDI detection in any clinical laboratory today. Because individuals can be colonized with C. difficile, testing of formed stool can result in false- positive tests, which may result in unnecessary treatments (PubMed)."

PCR positive, toxin negative do not have the disease.  The PCR by itself is overly senstive, leading to "over-diagnosis, over-treatment, and increased health care costs" (PubMed).

If you treat Clostridia difficile empirically it may result in false-negative PCR: " For PCR, 14%, 35%, and 45% of positive CDI tests converted to negative after 1, 2, and 3 days of treatment, respectively (PubMed)."

There are rare cases of extra-intestinal C. difficile (PubMed).

- C. novyi: soft tissues infections in heroin users, specifically Scottish heroin users laddie (PubMed).

- C. perfringens: Gas gangrene, one of the several necrotizing soft tissue infections. Skin popping drugs a major risk factor.

With gas gangrene: If you wait for gas in the diagnosis of gas gangrene, you might as well call the morgue. To make the diagnosis of gas gangrene, get a tissue gram stain. Carry it to the lab yourself, watch the tech stain it, look through the microscope and when you see a boxcar shaped, bipolar staining gram positive rod, swear softly to yourself and run, not walk, to the nearest phone for a surgeon, unless you are a surgeon. Then run like the Flash to the OR and get that patient debrided. This is a disease where you do not have a lot of time.

10% of outbreaks of gastroenteritis (PubMed); the classic being pig roasts.

- C. ramosum: Gas gangrene one of the several necrotizing soft tissue infections.

- C. septicum: sepsis from intra-abdominal source: often cecal cancer or lymphoma chemotherapy that kills the Peyers patches.

- C. sordellii: occasional bacteremia/sepsis esp after childbirth, abortion (? including after Mifepristone (RU-486), and in IVDA (PubMed). Usually fatal. A smattering of foot infections although not toxigenic strains and associated with bowel cancer (PubMed).

- C. tetani: tetanus aka lock jaw.


C. botulinum: Antitoxin therapy equine serum; in the United States can obtained from state health departments or the CDC (404-639-2206 workdays, 404-639-2888 other times). Wound botulism should be debrided PLUS penicillin, 10 to 20 million units qd OR metronidazole, but not proven to help.

Patients (in this case heroin users) who receive the antitoxin <12 hours after presentation do better.

Botulism can lead to permanent sequela (PubMed).

C. difficile (PubMed): metronidazole 500 mg po qid x 10 14 d is equal to vancomycin 125 mg po qid x 10 d for mild disease, vancomycin may be preferred for severe disease and dose at 500 mg qid (PubMed). With severe disease (toxic megacolon, WBC > 20,000) may want to start with vancomycin or even give both if the patient is NPO. Everyone (including me) thinks that vancomycin is superior to metronidazole, a fact that has yet to be definitively demonstrated in the literature, which tends to show they are equal.  In severe disease adding IV metronidazole to po vancomycin is of benefit (PubMed).

Vancomycin is also associated with a decrease in 30 d mortality (PubMed).

While there is a tendency towards using vancomycin, metronidazole may have less impact on the microbiome and colonization resistance (PubMed).

Fidaxomicin 200 po bid is equal to vancomycin with fewer relapses (PubMed)(PubMed) and may be superior to vancomycin when used to treat when you have to continue the inciting antibiotics (PubMed).

Nitazoxanide (250 mg 4 times per day for 10 days or 500 mg 2 times per day for 10 days) is as effective as metronidazole (PubMed). Nitazoxanide, 500 mg twice daily, for 10 days works in several studies for relapses (PubMed).

Vancomycin stool levels suggest that a loading dose of 250 mg or 500 mg qid during the first 24-48 hours, then 125 qid may be the optimal dosing (PubMed).

Adding tigecycline has been effective is a smattering of case reports (PubMed).

Resistance to metronidazole is being reported in some strains, one study as high as 25% of ribotype 1 (PubMed).

PPI use is a risk for disease and relapse (44% chance of relapse); as the hippies knew, acid is good, man (PubMed). Acid suppression is also associated with increased mortality (PubMed).

ASA decreases the risk, truly the wonder drug that works wonders

Interestingly, in a small series anti-motility therapy was helpful, not evil (PubMed).

Linezolid also effective in in vitro studies.

With relapsing disease, you usually switch to the other agent (magical thinking, no data to support a change is any better).

And here is a curiosity: having an appendix is protective against relapse (PubMed), so maybe an appendix transplant?

Stool 'transplants' work (I gag writing this), curing up to 90% of patients (Review) (PubMed). Fortunately they are developing synthetic stool, making acquisition easier. The source is the Senate, where they are full of, well, never mind. If you are going to transplant stool, give it from a thin donor; there is a case report of weight gain after a transplant from an obese donor (PubMed). This is why (Pubmed). And most curiously, a sterile fecal filtrate is also effective (Pubmed).

Stool transplants work best if preceeded by a vancomycin taper (PubMed).

For patients that have had several relapses I recommend a slow taper of one or the other agent as follows qid x 5 d then tid x 5 d then bid x 5 d then qd x 5 d then qod x 5d then stop. It seems to work in my experience. And may be even better if you finish up with a q 3 d dosing (PubMed). A vancomycin taper is no better than a stool transplant (Pubmed)

This disease may be getting worser: increased relapse and more severe disease being reported due to strains that are making 10 x the toxin as the old strains and is maybe refractory to metronidazole.

There is increasing data to suggest rifaximin at 400 mg 2 or 3 times daily for 10 to 14 days may be the best option for relapsing disease (PubMed); in several studies relapsing disease was treated with 2 weeks of vancomycin or flagyl then a 'chaser' with 2 weeks of rifaximin (PubMed) approximately halves the reoccurrence rate.

In one series IVIG for 3 days cured severe, refractory disease (PubMed); other series have shown no efficacy.

8ezlotoxumab is human monoclonal antibody against C. difficile toxin B. Use prevents recurrence. (PubMed) Costs 4 grand. No drug should cost the same as a fully loaded MacBook Pro.

In a small series of refractory disease, tigecycline was helpful (PubMed).

Toxin can be found days after the diarrhea is resolved, so do not check it.

If no active disease, then there is no transmission, so when fevers and diarrhea resolve, can take them out of contact precautions but you need to do a terminal cleaning of the room.

Preemptive (vancomycin) can prevent recurrent in patients who have had prior disease and are going to be exposed to antibiotics again(Pubmed).

Consider surgery if patient intubated or requires pressors; may be life saving (PubMed).

C. difficile colitis: do not give anti motility agents (well, maybe, if mild disease and treated with antibiotics one study found anti-motility agents to be safe (PubMed)); they can lead to toxic megacolon and death, generally considered a bad outcome. Same it true of adding rifampin; it kills (PubMed). Diarrhea is good and gets rid of the toxin. Plus, the patient can get some reading done.

The spores cannot be killed by alcohol, it is why you have to wash your hands. And it may have airborne spread (PubMed).

Having pateints wash their hands may decrease hospital acquired disease (PubMed) but with a p = 0.05 the effect is not likely real.

And there is also a post C. difficile reactive arthritis (PubMed).

C. perfringens: penicillin OR metronidazole PLUS clindamycin PLUS debridement. Do not delay surgery for hyperbaric oxygen (unless the patient is suffering from the bends as well).

C. ramosum: penicillin/beta lactamase inhibitor combinations, imipenem, metronidazole.

C. septicum: penicillin PLUS clindamycin PLUS debridement.

C. sordellii: penicillin PLUS clindamycin PLUS debridement.

C. tetani: see tetanus for details. Human tetanus immune globulin (HTIG) 500 IU. Active immunization must also be initiated; getting tetanus does not lead to immunity. Metronidazole preferred, penicillin, imipenem but only metronidazole has been shown to help. Treat with muscle relaxants especially diazepam; the cause of death is autonomic dysfunction.


Clostridium are often part of mixed infections or an unimportant bacteremia. The trick is deciding when it is important and when it is not. That is why there are ID doctors, he says in a self serving note.