NEW research showing that reducing the use of broad spectrum antibiotics may help to cut rates of hospital-acquired vancomycin-resistant Enterococcus (VRE) was “intriguing”, say experts.

In a research letter published in the MJA, researchers from John Hunter Hospital in Newcastle reported that a global shortage of piperacillin/tazobactam in 2017 had prompted the hospital to shift to the use of intravenous amoxicillin/clavulanate (IVAC).

The researchers noted that the number of hospital-acquired cases of VRE almost halved, dropping from 191 in the 12 months before the shortage to 101 in the 12 months afterwards (down 47%). The rate of methicillin-resistant Staphylococcus aureus (MRSA) also fell from 53 to 31 (down 42%), but this was not statistically significant, the researchers said.

Lead author Associate Professor John Ferguson, infectious diseases physician and microbiologist at John Hunter Hospital, said that despite high levels of infection control compliance, the hospital had struggled to control transmission of VRE since late 2013, particularly in the intensive care unit.

“This [piperacillin/tazobactam] shortage represented a natural experiment that was really quite exciting to us. We realised, probably about 5 months into the shortage period, that VRE acquisitions had gone down and at that point we decided to keep [piperacillin/tazobactam] usage restricted … so we could examine the effects over a full 12-month period,” he said in an exclusive InSight+ podcast.

The researchers further reported that there were 24 sterile site detections of VRE and 49 of MRSA, and these fell to eight (down 67%) and 37 (down 24%), respectively, after the shortage.

Associate Professor Ferguson said these infections were “incredibly” difficult to treat, so this represented a “real gain for patients”.

No change was reported in the rates of Clostridium difficile.

In the 12 months after the shortage, mean piperacillin/tazobactam use dropped from 44 defined daily doses (DDD)/1000 occupied bed-days (OBD) to 5 DDD/1000 OBD; while IVAC use increased from 4 to 33 DDD/1000 OBD.

The researchers noted that as IVAC is a narrower spectrum antibiotic than piperacillin/tazobactam, the total broad spectrum parenteral antimicrobial use fell from 129 DDD/1000 OBD before the shortage to 91 DDD/1000 OBD.

They said that piperacillin/tazobactam had now been reintroduced on a restricted basis at the hospital but was used far less than before the 2017 shortage.

Associate Professor Ferguson said the current study had not been able to demonstrate that no harm had been caused to patients in the switch to IVAC.

“We would like to have individual patient data to look at the impact of that change and show that there wasn’t increased mortality or treatment failure from using a narrower spectrum agent in intra-abdominal infection,” he said.

Professor Karin Thursky, Director of the National Health and Medical Research Council National Centre for Antimicrobial Stewardship, said it was great to see a hospital review the impact of its antimicrobial stewardship program on the rates of multidrug-resistant organisms.

“This should be emulated by all hospitals,” she said.

Professor Thursky said the finding was “intriguing” and the mechanism remained unclear.

“We are aware that VRE rates in Australia are among the highest in the world, and we have certainly been looking for answers as to what needs to be done to reduce the rates,” Professor Thursky said.

“The answers lie in understanding how VRE is generated (ie, endogenously through antimicrobial pressure in a patient and through transmission in the environment) and how it can be controlled. This should include antimicrobial stewardship and infection prevention (cleaning and hand hygiene).”

Professor Thursky said selective pressure from antibiotic use could result in suppression of other normal gut flora and cause an overgrowth of enterococci in the intestinal tract.

“This occurs in the presence of antibiotics with potent anti-anaerobe activity (such as piperacillin/tazobactam and amoxicillin/clavulanate), as well as with other antibiotics such as third- and fourth-generation cephalosporins and vancomycin,” she said. “It is very interesting to postulate how piperacillin/tazobactam would differ from amoxicillin/clavulanate parenterally with regards to a difference in impact on gut flora, colonisation and persistence in patients.”

Professor Thursky said there was no information in the study about the patterns of use of other antibiotics strongly associated with VRE (eg, vancomycin and ceftriaxone) during this period.

She noted that another study had shown that both piperacillin/tazobactam and ampicillin/sulbactam were associated with a reduction in VRE when introduced instead of cefotaxime and clindamycin.

Professor Thursky added that the MJA research did not provide information about the clonality.

“How much of the change in VRE was due to infection prevention, and the reduction in transmission? We note [in the graph] that VRE acquisitions may have started to decline prior to the introduction of the IVAC,” she said, adding that factors that could account for background VRE acquisition included endogenous acquisition from antibiotic selection pressure and VRE transmission in the environment.

“A higher density of VRE faecal colonisation may significantly increase the risk of transmission of VRE strains to the environment and to other patients. Control of VRE requires more understanding about the mechanisms by which different classes of drugs promote gastrointestinal colonisation,” Professor Thursky said.

“We now have genome sequencing and microbiome studies to examine these relationships in more detail.”

3 thoughts on “Global antibiotic shortage sparks “intriguing” finding

  1. John Ferguson says:

    Thank you, I can clarify that the major VRE vanA strains are clonal by WGS at this hospital, implying cross transmission. VanB is less prevalent and has mixed clonality.

    The VRE trend was upward in the 12 months prior to shortage. Note that month to month variability is high. Of course the study only provides evidence of association rather than causation.

    As reported at the ASA meeting, 2019, glycopeptide and carbapenem use was unchanged pre/post shortage. There was a 17% increase in ceftriaxone usage from a baseline of 23 DDD/1000 pt-days.

    We know that pip/tazo is excreted heavily into the bowel via the biliary tract and is associated with a large change to faecal flora (Wilcox-MH et al). It was once thought that it might be less impactful on VRE, however this has not been shown in more recent clinical studies . e.g. Infect Control Hosp Epidemiol. 2004 May;25(5):380-3. Effect of the increasing use of piperacillin/tazobactam on the incidence of vancomycin-resistant enterococci in four academic medical centers. Stiefel U et al. and the Paterson reference within the paper.

    As to why amox/clavulanate usage appeared less able to drive the VRE problem remains uncertain and requires further study. As JHH remained a low cephalosporin use hospital despite the piptazo shortage, this may have allowed the reduction in VRE to be demonstrated as significant replacement with cephalosporins or meropenem would have fuelled the VRE issue further.

  2. Dr Philip Dawson says:

    They find this surprising? Why use a high powered antibiotic first line instead of following the guidelines for community acquired infections?

  3. James Hurley says:

    An interesting natural experiment. I have a manuscript accepted for publication in Journal of Antimicrobial Chemotherapy [Studies of SDD as a natural experiment to evaluate topical antibiotic prophylaxis and cephalosporin use as population level risk factors for Enterococcal bacteremia among ICU patients] which does not directly address the question of VRE. However, ICU’s that use a regimen of topical antibiotics (SDD) to prevent ventilator associated pneumonia (a common practice in Dutch ICUs) have a higher incidence of enterococcal bacteraemia compared to other ICUs that do not use SDD (or topical antibiotics).

Leave a Reply

Your email address will not be published. Required fields are marked *