Antimicrobial stewardship and infection control: November 2018

Two RTCs show that probiotics do absolutely nothing for children with acute gastroenteritis. I’m all for offer non-antibiotic therapies as a means of avoiding inappropriate antibiotic use, but also think we have an ethical obligation not to peddle treatments that we know to be inefficacious. On that note, this month’s NEJM includes two RTCs on the topic of giving Lactobacillus to kids with acute diarrhea. Both studies reached the same conclusion: it doesn’t do squat.

In the first trial, the authors recruited 886 children between 3 months and 2 years presenting to Canadian EDs with acute gastroenteritis. Subjects were randomized to receive either a fixed dose of two species of Lactobacillus or a placebo twice daily for five days. The primary outcome was moderate-to-severe gastroenteritis as assessed by a diarrhea severity score (the Vesikari score). Rates of moderate-to-severe gastroenteritis at two weeks were 26% in the probiotic group versus 24% in the placebo group. In multivariate analysis, receipt of probiotics did not protect against moderate-to-severe gastroenteritis, nor did it reduce the durations of diarrhea or vomiting. 30462939

In the second trial, authors recruited 971 children aged 3 months to 4 years presenting to US EDs with acute gastroenteritis. Again, the patients were randomized to five days of either a fixed dose of Lactobacillus or placebo taken twice daily, and the primary outcome of interest was moderate-to-severe gastroenteritis as measured by Vesikari score. And again, the rates of moderate-to-severe gastroenteritis were similar between groups (12% in the probiotic recipients and 13% in the placebo recipients) with no differences in durations of diarrhea or vomiting, daycare absenteeism, or household transmissions. 30462938

Lactobacilli have a lot to recommend them: they protect women from genitourinary tract infections, make yogurt, pickles, and sourdough, and team-tag with Saccharomyces to produce fine Belgian beers. But they don’t do much for acute gastroenteritis.


Fecal microbiota transplantation (FMT) shortens patient’s duration of colonization with multidrug-resistant pathogens. Who knew eating poop could be so broadly therapeutic? The authors performed a matched case-control study examining patients colonized with carbapenemase-producing Enterobacteriaceae and Acinetobacter (CPE/A) at a single French medical center who did or did not undergo FMT. Patients were ineligible for FMT for CPE/A decolonization if they were immunosuppressed, still needed antibiotics for their CPE/A infection, had ileus or gastrointestinal perforation, or if their CPE/A colonization was not deemed “a barrier to further medical care” (e.g. they were being discharged from the hospital). Controls were matched to cases 2:1 based on sex, age, the species of carbapenemase-producing bacteria, and the type of carbapenemase produced. The primary outcome was the time to a rectal swab culture negative for CPE/A.

Thirty patients (10 cases and 20 controls) were included in the analysis. At two weeks after the FMT, 80% of cases had clearance of CPE/A from surveillance cultures, whereas in the control group only 10% had cleared CPE/A from their cultures within two weeks of the original CPE/A detection. The overall success rate of FMT for CPE/A pathogen clearance was 54%; the first round of FMT was curative in only 40% of patients, and reached 80% after a second round for non-responders. Six months after the FMT, 90% of cases were negative for gastrointestinal CPE/A; one patient who did not respond to FMT later had spontaneous decolonization. Overall, the median time to decolonization in the FMT-treated patients was 3 days versus 51 days for the control patients.

These data strongly argue that FMT has a role in “turning back the clock” on MDRO carriage in patients after antibiotic exposure. Imagine if every patient with a hematologic malignancy banked stool prior to starting chemotherapy, then underwent autologous FMT before each round of chemo to undo the microbiologic damage of all those weeks of ciprofloxacin prophylaxis. Maybe we wouldn’t end up with so many wards full of people on ceftaz-avi and tigecycline. 30472293


The addition of UV light disinfection failed to reduce rates of colonization or disease attributable to C.difficile or vancomycin-resistant enterococci on a stem cell transplant unit. The authors performed a 20-month study involving the 25-bed bone marrow transplant unit at Sloan Kettering cancer center. The intervention was adding universal terminal disinfection with a UV-radiation device to the institutions’ standard protocol, which was disinfection with ammonium for all rooms except those of patients with C.difficile infection (CDI), which were disinfected with bleach. The primary outcome was colonization or infection with VRE or C.difficile. Colonization was assessed by weekly surveillance rectal swabs; infections were defined by clinical diagnosis.

Five hundred and seventy nine patients had 704 admissions during the study period, among whom 22% had a positive test for C.difficile and 33% had a positive test for VRE. In the preintervention period, the prevalences of C.difficile and VRE were 8% and 17%; in the postintervention period, they were 7% and 13%. The acquisition rates of C.difficile did not differ before vs after the intervention (9.3 vs 7.1 cases per 1,000 patient-days; p>.05), nor did the acquisition rates of VRE (12.2 vs 9.7 cases per 1,000 patient-days). The authors mention that they tried to slice and massage the data in various ways to produce statistically significant differences between the two periods, all of which proved unproductive (but points for honesty!).

These UV disinfection papers pop up every few months and have done so for years – like cefazolin versus nafcillin for MSSA or steroids for pneumonia and sepsis, the debate about the added value of these machines is interminable. Probably the potential to capture benefit from adding UV disinfection to your current practice depends not only the device and protocol used, but on how well your environmental staff adhere to standard practices and what your residual rates of C.difficile and VRE transmission are despite that. For Sloan Kettering’s bone marrow transplant unit, which (assuming it operates at its full 25-bed capacity) is down to 1-3 episodes of CDI & VRE a month, it may simply not be possible to realize additional gains in infection control without reducing use of antimicrobials. 30226124


Dear vegans, here’s one more thing to feel self-righteous about: your antibiotic-susceptible gut microbiota. I really wish that vegan cuisine didn’t leave me feeling hungry and sad.  Not only is the vegan diet more sustainable, ethically less problematic, and associated with reduced incidences of obesity, diabetes, and malignancy versus omnivorous diets, but now it seems that it’s associated with reduced carriage of antibiotic resistant bacteria. Perhaps this shouldn’t be a surprise, given that the majority of the world’s antibiotic use is in the agricultural sector.

The authors recruited subjects who adhered to vegan (n=26), vegetarian (n=32), and omnivorous (n=43) diets. The stool microbiota of the subjects was investigated using 16s rRNA analysis, and the quantity of four antibiotic resistance genes (sul2, tetA, blaTEM, and strB, conferring resistance to sulfonamides, tetracyclines, beta-lactams, and streptomycin) were quantified. Pairwise comparisons showed that the loads of antibiotic resistance genes were lower in the patients with a vegan versus vegetarian or omnivorous diet (p=0.01 for both), though interestingly the difference in resistance gene load was no different between vegetarians and omnivores. However, the differences varied by individual gene: blaTEM was actually found less frequently among the omnivores than vegans or vegetarians, whereas sul2 and strB were found in greater quantities and tetA was found in similar loads between all groups. This suggests that some antibiotic resistance genes are more readily transferred by animal-based foodstuffs and others by plant-based foodstuffs, though on balance the animal-based diet leads to a greater acquisition of antibiotic resistance.

Something I found a little bit frustrating about this article is that they don’t describe the differences in taxonomic diversity of the microbiomes between groups. This is relevant because an alternate hypothesis is that, rather than resulting in transmission of antibiotic resistance genes from animal products, the omnivorous diet simply encourages proliferation of gut microbiota that happen to carry antimicrobial resistance genes, whereas the vegan diet enriches the gut microbiota with taxa that less frequently carry these genes. Anyway, I picked this article mostly because I thought it was fascinating, but here’s your clinical correlation: if you need an FMT for recurrent infection with multidrug-resistance organisms, maybe try to get it from a vegan? 30081136