Slide 2

CDI is common in the United States. There were almost half a million cases in 2011 alone, as
well as a mortality rate of over 9%. CDI is also associated with high healthcare burden, including
healthcare expenditures of over $8 billion in a single year. Patients who develop CDI are likely
to get it again- 1 out of 4 patients are likely to develop another recurrence, and 1 out of 2
patients are likely to develop a second recurrence.

Slide 3

Antibiotics are a major risk factor for the development of CDI. Broad spectrum antibiotics, for
example, not only target infectious pathogens, but they can also target our normal gut flora.
This results in a favorable environment for opportunistic pathogens, like C. difficile, to
proliferate in our gut, potentially causing infection and negative health outcomes.

Transition: However, not all antibiotics are created equal in terms of how they affect the gut
microbiome.

Slide 4

A mouse study by Schubert and colleagues in 2015 showed antibiotic effects on the four major
phyla in our gut- Bacteroidetes, Firmicutes, Proteobacteria, and Verrucomicrobia.

Clindamycin and metronidazole, for example, showed a decrease in relative abundance

Bacteroidetes and Firmicutes compared to no antibiotic use. This could be due to clindamycin
and metronidazole’s activity against anaerobic bacteria, of which many belong to the
Bacteroidetes phyla. Ciprofloxacin, on the other hand, appears to have the least impact on
bacterial diversity compared to the other antibiotics.

Although not analyzed in this study, carbapenems are known for their broad spectrum of
activity including anaerobic coverage, so it can be deduced that they may have a profound
effect on our gut microbiome.

Notes:
 Clindamycin is not typically used to target anaerobes, but it does have good activity
against Bacteroides fragilis

Slide 5

On a clinical level, a meta-analysis by Brown and colleagues showed that, in the five studies
included, clindamycin was the strongest predictor of CDI risk, followed by CMCs and
fluoroquinolones.
Although FQ’s influence on the gut microbiome from the previous study showed small changes
in bacterial diversity, its impact on CDI risk can be explained through the discovery of a
particular CD strain.

Slide 6

The Center for Disease Control and Prevention developed an Emerging Infections Program that
monitors several infectious diseases, including CDI. One of their tasks was to look at various C.
difficile isolates in the United States. They saw that in 2016, ribotype 027 was the second most
common isolate in community-associated CD isolates, and the most common in hospital-
acquired CDI.

Slide 7

Transition: With the emergence of the 027 strain, it would make sense then, why
fluoroquinolones are a high predictor of CDI risk. However, a more recent analysis done by
Vardakas and colleagues in 2016 showed different results.

Slide 8

In this meta-analysis of 79 studies, although clindamycin was the strongest predictor of CDI risk
among those studied, fluoroquinolones were not.

We know that broad-spectrum antibiotics can have a large effect on our gut flora, so
theoretically, these antibiotics would provide greater risk in CDI development. With these
meta-analyses showing various results, our study sought to be the most recent analysis to
identify [objectives].

Notes:
 Vardakas analysis only included RCTs, while Brown included observational and RCTs.
Everyone in RCTs were already receiving antibiotics at time of study, while in
observational, it is believed that they would test patients on antibiotics more so than
those who aren’t on an antibiotic

Study analysis:
 Multivariable logistic regression model- testing individual antibiotic classes to see which
are at high-risk for CDI
 Can then analyze interactions between antibiotic classes to account for patients who
receive multiple antibiotics
 Add increasing combinations and build a Bayesian probability of risk for every
antibiotic/antibiotic class. First, assess patients receiving only one antibiotic/antibiotic
class, then patients receiving two antibiotics/antibiotic classes, and continue adding
until all combinations are accounted for