Sep 282018

Jon-Emile S. Kenny MD [@heart_lung] with illustrations by Carla M Canepa MD [@_carlemd_]

My reputation’s never been worse, so …

-Taylor Swift


A 54 year old woman with lupus nephritis – treated with mycophenolate mofetil and prednisone – is admitted with pancytopenia and worsening confusion.  She was recently discharged after receiving levofloxacin and clindamycin for hospital acquired pneumonia.  On presentation to the intensive care unit [ICU], she is tachycardic and hypotensive following 2 days of multiple watery bowel movements with ‘horse barn odor.’  As she had been complaining of abdominal pain, a CT abdomen and pelvis was obtained which revealed diffuse edema and stranding about the entire colon.  She is with acute kidney injury, hypoalbuminemia, elevated lactate, mild acidemia and a WBC count of 2K.  She receives 3 L of lactated ringers fluid resuscitation and is started on norepinephrine via a peripheral IV.


I’ll take “Bad Reputation” for one-thousand, Alex …

Answer: Annual healthcare expenditure in the United States is increased by roughly 1.5 billion dollars by this vexatious gram positive, anaerobic, spore-forming, toxin-producing bacillus.

Question: What is Clostridium difficile?

It has been 8 years since the last Infectious Disease Society of America [IDSA] – Society for Healthcare Epidemiology of America [SHEA] practice-guidelines update on C. difficile infection.  This problematic diarrheal illness is the most common healthcare-associated infection in the United States, though about one-third of cases are acquired within the community.

C. difficile infection [CDI] is procured via the fecal-oral route, though it does so via exceptionally hardy spores that populate nearly all surfaces of the hospital. The virulent and epidemic strain of C. difficile known simply as ‘ribotype 27’ is more ceremoniously christened toxinotype III, restriction endonuclease analysis group BI, North American pulsed-field gel electrophoresis type NAP1, polymerase-chain-reaction [PCR] type 027 – or BI/NAP1/027. With aggressive restriction of fluoroquinolones, infection recognition and isolation strategies, ribotype 027 prevalence has declined in Europe since 2007 though it remains a formidable foe in North American locales.

The 2018 update contains an embarrassment of details spanning epidemiology, diagnosis, infection control as well as prevention and treatment of CDI for both pediatric and adult populations.  While these exhaustive guidelines should be thumbed-through by all clinicians dealing with CDI, this brief synopsis will focus and explore issues pertaining to diagnosis and management of CDI for clinicians – especially within the ICU.


A proper Bayesian analysis should be applied when approaching CDI.  Like all of medicine, a low pre-test probability for disease coupled with a sensitive test will likely yield a false positive rather than true disease.  How is this worded in the guidelines?

“Patients with unexplained and new-onset ≥3 unformed stools in 24 hours are the preferred target population for testing for CDI (weak recommendation, very low quality of evidence).”

The operative words above are unexplained and new-onset.   The authors make clear that patients with recent [i.e. within 48 hours] laxative use, and other common causes of diarrhea such as IBD, enteral tube feeding or chemotherapy should be considered carefully before reflexive C. difficile testing.  Further, the quality of the stool should be faithfully scrutinized prior to laboratory assessment; the stool should take the form of the container in which it is sent – in other words – Bristol Stool Type 6 and 7 [see stool chart].

The Bristol Stool Chart: stools of the type 6 or type 7 variety should be sent for C. difficile testing.

Institutional Criteria for Stool Submission Absent

Importantly, if an institution does not implement strategies to improve the pre-test characteristics of patients being tested for CDI – e.g. laboratories have no clinical data and accept all unformed stools for testing – then the guidelines recommend that both a sensitive and specific test be implemented [a ‘multi-step’ approach].   What are said tests?  While this is complicated as there are multiple commercially-available versions of these assays, the guidelines characterize the following as sensitive but less specific in an institution with relatively indiscriminant testing:

-GDH [glutamate dehydrogenase] which is a highly-conserved enzyme present in high levels in all isolates of C. difficile – in both toxigenic and nontoxigenic strains.

- NAATs [nucleic acid amplification tests] note that there at least a dozen of these tests of varying diagnostic accuracy directed against several genetic foci including the genes for toxin A and B; nevertheless, in the setting of relatively indiscriminant stool testing, the NAATs should be viewed as less specific, in totality.

To the above should be added the less sensitive, but more specific stool toxin A and B assays; these tests use monoclonal or polyclonal antibodies to detect C. difficile toxins – there are also numerous commercial assays available.

Institutional Criteria for Stool Submission Present

By contrast, if an institution appropriates prerequisite clinical criteria for stool testing [ideally, this avoids patients with a lower pre-test probability of CDI], then a NAAT alone, or a multi-step approach, as above, is preferable to a toxin test alone.


In totality, the ideal approach to CDI remains challenging as patients may shelter toxigenic strains in their colon and not have clinical disease and > 60% of patients may remain C. difficile positive even after successful treatment.  Accordingly, without distinct changes [e.g. new character of diarrhea or evolving and concerning clinical features], repeat testing should not be performed.  If, however, symptoms recur following resolution of CDI, then testing for recurrent CDI should include toxin detection.


A key takeaway for CDI management in the 2018 guidelines is that metronidazole should only be used:

  1. As an intravenous adjunct for fulminant CDI [see below],
  2. If there are absolute contraindications to vancomycin or fidaxomicin or
  3. If vancomycin and fidaxomicin are not available.

Thus, therapy for all CDI severities should now begin with oral vancomycin or fidaxomicin.  The dosing for oral vancomycin is 125 mg 4 times daily for 10 days for mild or severe disease.  If there has been a good clinical trajectory without complete resolution by 10 days, then extension to a fortnight of therapy is an option.  The dose of oral vancomycin should increase to 500 mg only if there is fulminant CDI [previously called severe, complicated - defined by hypotension or shock, ileus, or megacolon]; in this situation it is also recommended that IV metronidazole be added [see flowchart].

CDI Treatment Algorithm: this diagram is for teaching purposes and treatment regimens including doses must be cross-referenced.

First and second recurrences of CDI are beyond the scope of this brief post; importantly, for ICU practitioners, the dosing of vancomycin or fidaxomicin remains the same for acute recurrence; these patients may require a longer, tapered vancomycin course.

Lastly, there is no robust evidence for fecal microbiota transplant [FMT] in the acute, critical care setting and FMT should be considered upon a patient’s third episode of CDI.


In general I find the IDSA guidelines to be of excellent quality with fair and reasonable assessment of the evidence.  I find their approach to diagnosis – which is inherently Bayesian – to be an admirable synthesis of the current literature.  It seems prudent to set forth institutional guidelines for C. difficile testing bearing in mind the inherent imperfection of recipe-based medicine – especially in the immunocompromised.  Equally important is to know the diagnostic characteristics of the bioassays employed in one’s hospital.

The previous guidelines also relied upon risk assessment to guide treatment of C. difficile.  While the 2018 approach follows suit, treatment is now the same for mild and severe disease.  This follows – since publication of the last guidelines – the observed superiority of vancomycin in all disease severities; accordingly, oral metronidazole should rarely be considered.

In the ICU

Interestingly, the oft-quoted severity criteria used to guide treatment in the previous IDSA iteration borrowed from a study with a subpopulation of only 2 critically-ill patients.  Further, the recommendation for adding intravenous metronidazole in fulminant CDI [see flow diagram] in the current directives was based upon a 2015 CID paper with the following inclusion criteria:

“… 3 of the following 7 criteria had to be present within 24 hours of CDI treatment initiation: albumin <2.5 g/dL, heart rate >90 bpm, mean arterial pressure (MAP) <60 mmHg, WBC count ≥15000 cells/mL, age >60 years, serum creatinine ≥1.5 times baseline, or temperature ≥100.4°F (38°C).”

This begs the question as to whether the aforementioned criteria should be applied when initiating IV metronidazole in the ICU, rather than “hypotension, shock, ileus or megacolon.”  The authors of the 2018 guidelines fully disclose that their definition of fulminant CDI was based on expert opinion and requires further study.

Also of note, in the above-mentioned study of combination oral vancomycin and IV metronidazole in the critically-ill, 60% of patients in the combination group received oral vancomycin at 125 mg QID rather than 500 mg.  Despite this, the combination group was associated with a ~ 20% absolute risk reduction in mortality!  As the authors communicate, oral vancomycin 125 mg every 6 hours achieves fecal concentration in far excess of that needed to inhibit C. difficile. Currently, there is no evidence to support a higher dose for critically-ill patients with CDI and the higher dose is based on concern for impaired gut motility and drug delivery to the colon.

With respect to the use of fidaxomicin, its efficacy is similar to that of oral vancomycin; the cure rate for acute infection is approximately 90% for both.  As well, risk of recurrence was equivalent between both antibiotics amongst patients infected with the hypervirulent BI/NAP1/027 strain [38% of isolates].  Lastly, while there are case reports of rectal fidaxomicin, current evidence supports the use of rectal vancomycin when there is ileus.

Return to Case   

The patient’s condition progressively worsens on norepinephrine and vasopressin infusions.  She is intubated for deteriorating oxygenation with combined respiratory and metabolic acidoses.  Repeat CT of her abdomen and pelvis demonstrates ileus and megacolon.  She is started on rectal vancomycin as well as intravenous metronidazole; following surgical consultation and confirmation of CDI, she is taken to the operating theatre for urgent colectomy.  Hours after her colectomy her mean arterial pressure normalizes off of all vasoactive infusions and she is extubated the following morning.



Dr. Kenny is the cofounder and Chief Medical Officer of Flosonics Medical; he is also the creator and author of a free hemodynamic curriculum at

Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


2018 IDSA Guidelines for Clostridium Difficile Infection: fecal transplant gets its moment in the sun