August 2019: Antimicrobial stewardship and infection control

What’s the current state of the literature on MDRO decolonization using fecal microbiota transplantation (FMT)?  The authors, who are out of the GI and ID divisions of the Mayo Clinic, provide a systematic literature review of papers published through February of this year.  They focused on articles that looked at the effect of FMT on MDRO colonization and prevention of MDRO infections; they ignored “partial decolonization” endpoints in the studies that reported those; similarly, they considered ongoing vs absent MDRO infection as their primary outcome of interest rather than a reduction in the incidence of MDRO infections after FMT.

In total, their search strategy returned 1661 records, of which 43 full texts were assessed for eligibility in the review and 21 were actually included.  These 21 tests included one RTC, seven uncontrolled clinical trials, two cohort studies, two case series, and nine case reports; in total, they include data on 192 patients.  Three studies evaluated FMT for MDRO infection prevention, 16 studies evaluated FMT for MDRO decolonization, and two studies examined both endpoints.  Of the studies looking at decolonization, 15 only looked at the gut, whereas three also examined other sites (e.g. skin, urinary tract, and respiratory tract).  MDROs examined included VRE, CRE, ESBL-producing enterobacteriaceae, MRSA, and miscellaneous antibiotic-resistant organisms.  Routes of FMT administration were mostly by NG or nasoduodenal tube (n=6 and n=8, respectively), or by enema (n=4) or oral capsule (n=1); endoscopic administration was used in only 4 studies.

After excluding patients who did not undergo FMT, had no proven baseline MDRO colonization, or had missing data, 151 patients were included in the final analysis.  Followup ranged from 14-1200 days; across all studies, the colonization and infection eradication rates ranged from 38-88%.  Similar rates were found when excluding case reports and case series, and when restricting the analysis to studies in which most or all patients had at least 6 months of followup.  No serious adverse events were reported. Eight patients died during followup, in only one of whom FMT could not be ruled out as a contributing factor; this person had a hematologic malignancy and underwent HSCT for that, then died 51 days later of what was thought to be either GVHD and/or CMV colitis.  No evidence of publication bias was found using funnel plot or LFK index.  In the one RTC examined, the MDRO eradication rate with FMT was 67%, but this was not significantly different than the 58% eradication achieved in the placebo arm.

So, FMT is safe, but it seems the real question is whether it hastens MDRO decolonization or prevents MDRO infection beyond what is achieved with simple passage of time. 30986562

Speaking of MDRO colonization, AAC has a paper out this month that examined 1455 patients hospitalized in Johns Hopkins ICUs over a 6-month period.  Of these, 8.2% had a positive screening test for CRE colonization (mostly rectal swabs), and 12 of these (10%) developed a clinical infection with CRE during their hospitalization.  Progression from colonization to clinical infection was much more common among patients hospitalized with carbapenemase-producing CREs than CREs with other mechanisms of resistance (36% vs 3%; p<0.05).  Interestingly, most of these were respiratory infections (67%); intrabdominal infections (10%) and bacteremias (2%) accounting for a minority of clinical infections. 31138574

Single-occupant patient rooms may slow the transmission of nosocomial pathogens.  Here’s an interesting study.  McGill University bought itself a new hospital, and in April 2015 moved all of its patients from a older 417-bed facility that had 3-4 occupant ward-type rooms to a newer facility with 100% single-occupant rooms.  The authors took it upon themselves to record rates of MDRO colonization and infection in patients at the old and new hospitals in the two years before and three years after the switch.  Their main outcomes of interest were colonization or infection with VRE, colonization or infection with MRSA, and infection with C.difficile, all measured as an rate per 10,000 patient-days.

The authors found that the switch from the old to the new facility was associated with a steep decline in VRE colonization (IRR 0.25 with 95% CI 0.19-0.34), VRE infection (IRR 0.3 with 95% CI 0.12-0.75), and MRSA colonization (IRR 0.57; 95% CI 0.33-0.96), but not MRSA infction (IRR 0.89 with 95% CI 0.34-2.29) or C.difficile infection (IRR 0.95 with 95% CI 0.52-1.76).  Aside from the point estimates and confidence intervals, I thought the continued trends in each outcome after the switch were interesting:  the VRE colonization rates dropped immediately and were flat thereafter, whereas the MRSA colonization rates dropped immediately and then began regressing toward the mean, and the CDI rates, which had already been steadily declining during the pre-switch period, continued to decline at more or less the same rate after the move to the new facility.  In any case, it seems that giving patients their own rooms may make some headway on VRE colonization and infection rates, but does less for MRSA and nothing for C.difficile. 31424489

Patients with ESBL-producing Enterobacteriaceae colonization on contact precautions are no more likely to transmit their MDROs when housed in multi-patient versus single-occupant rooms.  This was a cluster-randomized crossover noninferiority trial conducted across 16 hospitals.  Hospitals were assigned 1:1 to either implement contact precautions for patients with ESBL-producing organisms either exclusively in single-patient rooms or in multiple-bed rooms.  The primary outcome of interest was transmission of ESBL-producing MDROs to wardmates, as measured by rectal carriage of an organisms clonally related to that carried by the index patient.  The noninferiority threshold was set at 10%.

Over a three-year period, the 16 hospitals enrolled a total 693 index patients and 9527 wardmates, of whom 463 and 7093 were included in the per-protocol population.  Transmission of ESBL-producing MDROS to a wardmate occurred in 4% of index patients under the single-room strategy versus 7% of the multiple-bed strategy (risk difference 3% with 90% CI -0.3 to 7.1.  Again with these 90% confidence intervals that just happen to brush up against the noninferiority margin!  Torture the data long enough and you can get it to confess to anything, I guess. Anyway, the authors conclude that contact precautions are equally effective under a multiple-bed versus single bed rooming strategy.  I don’t know about that.  The big problem with the ESBL-producing organisms is that the most common EBSL gene, CTX-M, just doesn’t confer much of a penalty to an organism’s fitness. So, once it gets into a bacterial population, CXT-M doesn’t go away: put another way, the percentage of ESBL-producing Enterobacteriaceae on your hospital’s antibiogram will probably go up a lot faster than it will go down.  Given that, even if the benefit is small, single-occupancy rooms for patients colonized with these organisms may be worth the benefit to your antibiogram over the long haul. 31451419