Zinc, Pigs and the Meaning of Antimicrobial Stewardship

Bacteria are smarter than we give them credit for.
Or maybe we’re not a bright as we think we are.

Antimicrobial stewardship is sometimes (wrongly) assumed to simply be the practice of ‘using fewer antimicrobials’. It’s more complex than that, because the issue is complex. At face value, using overall reduction in antimicrobial use is a logical target, and it’s true that it is a big aspect of stewardship.

However, what is an ‘antibiotic’?

“Raised without antibiotics” and “Antibiotic-free” are good marketing terms, but what do they mean in terms of antimicrobial resistance? That’s less clear.

In pig production, control of post weaning E. coli diarrhea is a big problem. Prophylactic antibiotics are effective for this, but that’s not ideal. The main thing that’s done to replace antibiotics is to add a lot of zinc to the piglets’ diets at that age.

  • The reason….zinc kills bacteria.
  • The problem…bacteria don’t care whether we call it an antibiotic, just that it’s trying to kill them. So, they try to resist it.
  • The bigger problem….the way the resist it can be linked to the way they resist conventional antibiotics.

We (and others) have previously shown that addition of high levels of zinc to the diet of piglets selects as well for MRSA (methicillin-resistant Staphylococcus aureus) as tetracycline, the commonly used antibiotic for prevention of post-weaning diarrhea. If zinc selects for MRSA but not other resistant bugs, while tetracycline selects for a broader range of resistance, zinc use still might be of concern, but it would be a better option than tetracycline. However, that’s not the case and a recent paper in PLOS ONE (Ciesinski et al 2018) provides more information on the impact on other important resistant bacteria.

In that study, they took a typical approach of feeding groups of pigs either low levels (dietary requirements) or high levels (antibacterial levels) of zinc, and they investigated what happens with E. coli. They found significantly higher levels of multi drug resistant E. coli in association with feeding high levels of zinc: 5.8-14% in the control group compared to 29-30% in the high level zinc group. This appeared to be because the resistant strains persisted better than susceptible strains, as numbers of E. coli didn’t increase, but the proportion of resistant strains did.

Does this mean we shouldn’t be feeding high levels of zinc to piglet?

  • I don’t know. Prevention of disease is important for various reasons, including piglet welfare, reduced need to use therapeutic antibiotics (which are often more important drug classes than those used for prevention) and the need for economic production of safe food.
  • Whether antibiotics or zinc are better (or less worse) for resistance in piglets and the corresponding human risk is still unclear.
  • Another unanswered question is the impact of high levels of zinc in manure, since that ultimately makes its way into the ecosystem (just like some antibiotic residues).

However, it provides more evidence that ‘common sense isn’t evidence’ when it comes to antimicrobial resistance. We can’t assume things will have positive or negative aspects because ‘it makes sense’. We need proper research to figure out the best ways to optimize and improve antimicrobial use, minimizing resistance while maximizing the care of people and animals.
That’s antimicrobial stewardship.



Show Me The Data (But Also Do Something With It)

Surveillance is an important aspect of antimicrobial stewardship. Knowing where we stand, whether it’s the amount of drug used, the types of drugs that are used, the prevalence resistance or various other types of data, is important. That’s why there’s a lot of effort put into surveillance.So good

However, data need to be used to have a real impact. That can involve setting targets, identifying areas where interventions are needed, developing and testing interventions and similar things that aim to reduce or optimize antimicrobial use, not just tell us what’s happening.

Sometimes the action component gets lost, since analyzing data can be a lot easier (and less expensive, and less time consuming) than effecting change.

So, linking those areas is important, as is figuring out how to get better data (and…not-so-subtle CANresist plug here….something we need to do more of.)

  • A commentary in the Veterinary Record (Coyne et al, Dec 2017) entitled Antimicrobial use in dairy cattle; what gets measured gets improved highlights some important issues.. A research paper in the same edition of that journal (Hyde et al) reports on changes in antimicrobial use on UK dairy farms.

A few highlights:

They looked at a few different ways to assess antimicrobial use, including mass (mg) of antimicrobial active ingredient per population correction unit (mg/PCU), defined daily doses (DDDvet) and defined course doses (DCDvet).

  • It’s a challenge to accurately capture use on farms (and other places) because of the ways antimicrobials are dispensed and used, but with a standard approach, changes over time can be assessed with a good degree of confidence. It also highlighted some of the problems we have with surveillance, including how to report data. Reporting kg of drug used is pretty crude. Defined daily doses and other metrics used in humans are hard to accurately apply in animals, with markedly different weights and limited animal-specific use data. In dairy cattle, intramammary antibiotics are commonly used, something that defies accurate capture using some reporting approaches.

There was a lot of variation, with use of farms ranging from 0.36 to 97.79 mg/PCU and 0.05 to 20.29 DDDvet.

  • That gives a snapshot of the farms that were studied, and it shows that there is at least a subset of farms that are ripe for an intervention. That’s step 1. Figuring out why their antibiotic use is so high is step 2, helping them drop it is step 3, and more surveillance to see if the rates actually drop is step 4. All are important and it’s critical to keep going after step 1.

25% of farms accounted for over 50% of antibiotic use.

  • Targeting those farms can obviously have a big impact overall.

The commentary states “This work emphasises the need for accurate baseline antimicrobial use data which is representative of the national dairy herd . Identifying such high antimicrobial use behaviours will allow the industry to better target antimicrobial use interventions to reduce use at a farm, veterinary practice and national level. Due to the wide spectrum of antimicrobial use recorded by Hyde and others, there is a need to approach antimicrobial reduction interventions at an individual farm level. For example, benchmarking can improve farmer awareness of their current antimicrobial use practices and may play a role in reducing high antimicrobial use behaviours and promoting prudent use.”

We can replace “national dairy herd” with every other species (including people), and add humans healthcare facilities and the community and this statement would still be accurate.



Antimicrobial Use in Food Animals: USA

Recently, we had a series of posts about the CAHSS report, including both human and animal antimicrobial use trends. Now, how about some US animal data to add to the mix?

The USDA has released their 2016 report on antimicrobial sales and use in food animals. It’s a big report that I assume most of you won’t want to read in full, so here are some highlights.

Antibiotic sales and distribution dropped by 10% from 2015-2016.

  • Pretty impressive.

Cattle and pigs battled for the biggest users, with moos edging out oinks.

Medically-important antibiotics encompasses pretty much anything used in humans, so it’s a broad group. Use of this group decreased in 14% from 2015 to 2016. Overall, 60% of antibiotics used in food animals were medically important. The majority of this (70%) was tetracyclines.

  • That’s not surprising, given how they are used in food animal species. It’s also a ton of drugs (figuratively….literally, it was many tonnes, with 5,866,588 kg used in 2016).
  • That a lot, but down 15%.
  • The good news is that tetracycline are in the lowest group of the medically important antibiotics.
  • Still, that’s a lot of antibiotics and probably a big area where further reductions could occur without negative impacts.

Cephalosporins and fluoroquinolones, two critically important drug classes, each accounted for <1% of use. Most of the cephalosporin use was in cattle.

  • Good news (although we can still be better).

As expected, chickens led the way in relative use of not medically important drugs.

  • There are a few reasons for that, including the diseases of greatest relevance and active efforts to curtail use of certain drugs.

In-feed use was most common, followed by in water use. Individual animal treatment accounted for a much lower percentage.

It’s hard to say whether this is all good news, bad new, or just news. Reductions are good. However, looking at kgs of drugs only tells a very small part of the story. Much more refined information is needed, along with efforts to reduce antibiotic use while avoiding impacts on animal health, welfare and productivity. It’s not an easy area to address, as it’s a complex problem that will require complex solutions.


Fossil Fuels and Antibiotics: an Analogy

I think most people buy into the concept of fossil fuels being finite resources. Someday, they’ll run out or logistics and cost of extraction will make them impractical. Accordingly, we’re thinking about ways to reduce and improve use (to delay the end of the fossil fuel era) and develop effective alternatives. People know all about this, but motivation is somewhat limited by the fact that we’re not going to run out of oil in our lifetimes.

There are some good comparisons to antibiotics. Unlikely other drugs, we have to consider them to be finite resources. There’s always some degree of “use them and lose them” because of antibiotic resistance. Like oil, we need to reduce and improve use to delay the end of the antibiotic era, and come up with alternatives. However, unlike oil, the timeframe is potentially much shorter. Antibiotic resistance isn’t a threat for future generations. It’s a threat for every one of us.

Modern medicine needs antibiotics. The 50th anniversary of Dr. Christiaan Barnard’s first successful human heart transplant was a little over a week ago. Without effective antibiotics, transplants (and many other routine procedures) would be a thing of the past

Putting things in that perspective should motivate us to put the required time, effort and money into antimicrobial stewardship. It’s not as flashy as a new Tesla, but it’s more important.


Global Antibiotic Use in Animals

We tweeted a question yesterday (#CANADAresist) asking people to guess which countries use the most antimicrobials in animals and which ones are expected to increase the most by 2030. Antibiotic use in animals is moving (and in some countries, very hard to hit) target, but a 2015 study (Van Boeckel et al, Global Trends in Antimicrobial Use in Food Animals) has some answers.

The dubious winners: China, USA, Brazil, India and Germany.

  • Those make sense, as large countries with large food animal production systems.

Predicting the future of antibiotic use is hard, and hopefully the 2030 estimates will be high because of antimicrobial stewardship programs, but the authors predicted that the list in 2030 will be: China, USA, Brazil, India, Mexico

  • Again, those make sense. Lots of people. Lots of farms. Growing populations and economies.

Another interesting number is the predicted relative increase. I suspect most of us wouldn’t have guessed these.

  • Here, Indonesia takes the lead, followed by Myanmar, Nigeria, Peru and the Philippines.

Screen Shot 2017-11-22 at 8.10.50 AM

Data like these are just more examples of why antimicrobial stewardship programs need to develop and need to take on an international tone. Trying to attack a complex problem in isolation (either looking only at one country or looking only at animals or humans) can’t be the answer. Complex problems need complex solutions.


Canadian Trends in C diff, MRSA and Gonorrhea

As we continue with World Antibiotic Awareness Week, here are some more tidbits from the 2017 CARSS report.

Clostridium difficile

  • Hospital-associated C diff rates continue to decrease, with a fairly remarkable drop from 6.64/10000 patient days in 2011 to 4.04 in 2016.
  • Community-associated C. difficile infection is still a big unknown. In 2015, 37% of C. difficile infection patients admitted to sentinel hospitals were thought to have infections that originated in the community. That’s not surprising given what we’ve been learning over the past 10 years, and the epidemiology of C diff is much more complex than previously assumed.

C diff copy


  • There was a bit of an increase in MRSA rates in sentinel hospitals, from 2.84-3.13/10,000 patient dates between 2011 and 2016.
  • Similar to C diff, community-associated MRSA has become an increasingly recognized problem. When origin is considered, hospital-associated MRSA actually decreased from 1.93-1.69/10,000.
  • Pediatric hospital community-associated MRSA rates were higher than those in adult hospitals (1.56 vs 1.02). There was a big increase in pediatric MRSA bloodstream infection rates, something that’s obviously a concern given the potential severity of disease.

MRSA copy

  • In terms of strains, CMRSA10 (also known as USA100) continued its rise, while the old leader CMRSA2 continued to decline. Part of this is the result of more community MRSA cases (where CMRSA10 is predominant) and fewer hospital cases (when CMRSA2 has been the main strain).

mrsa strains copy


  • Along with the increase in numbers of cases of gonorrhea, resistance has also continued to increase. Azithromycin-resistance increased fro 1.3-47% from 2010-2015. Decreased susceptibility to cefixime and ceftriaxone increased from 1.1 and 1.9 to 2.9 and 3.5%, respectively, between 2014-2015.

gonorrhea resistance copy



CPE in Animals

Continuing our World Antibiotic Awareness Week blitz, let’s head back to the 2017 Canadian Antimicrobial Resistance Surveillance System report

Carbapenase-producing Enterobacteriaceae (CPE) are Gram negative bacteria (e..g E. coli, Klebsiella, Enterobacter) that are resistant to carbapenems, an important class of antibiotics. They’re a big concern because carbapenems are used for people with serious infections, and treatment options may be limited (since these strains are often resistant to many other drug classes too). Carbapenems aren’t used in food animals and are rarely used in companion animals, but resistance can be found in bacteria from animals. It’s probably driven by use of other antibiotic classes that are used in food animals that CPE are also resistant to. A big issue is that these bacteria are common inhabitants of the gut. Therefore, resistant strains can live happily in the intestinal tract for potentially long period of time, being passed in feces. (I probably don’t need to point out that cows poop a lot).

In Canada, screening of farm, slaughterhouse and retail samples was started in 2013, with a new technique added in 2016 to provide more sensitive detection.

  • The good news….no CPE were identified amongst over 13,000 screened surveillance isolates.
  • The somewhat concerning news….9/3,000 farm/retail/slaughterhouse samples were positive. That’s obviously a low rate but it’s not zero. All were seafood products. Three are stated as being imported, while my assumption is the other 6 were too (although I could be wrong…it’s not clear in the report). This highlights the fact that we don’t live in a bubble and bacteria could case less about immigration rules.

The fact that no CPE were identified in livestock or their associated products was great.

For now, at least.

It will probably happen but continued surveillance to identify and hopefully mitigate risks like this is important. We’re also continuing to look for CPE in companion animals. I’ve dealt with a few cases, but all were from the US. CPE in pets may end up similar to the situation with MRSA, whereby infections in pets are most often associated with transmission from infected owners, but food animals, food and the environment all remain potential ways that CPE could be spread from livestock.

At this point, CPE is rare in people in the community, but if it successfully spills out into the community from hospitals, we’ll likely see further spillover into domestic animals. Yet another reason why a multisector approach to antimicrobial stewardship is needed (yes, that’s a plug for CANresist, but it’s true).



Antimicrobial Use Surveillance, Humans: Canada

The last post was about animals so let’s get to some human data from the 2017 Canadian Antimicrobial Resistance Surveillance System report.

In 2016, 247,014 kg of antibiotics were dispensed, most (>200K kg) through pharmacies, with ~40K kg through hospitals. Overall, ~92% of defined daily doses of antibiotics were dispensed through pharmacies.

  • This is another example why there needs to be more work on community-level antimicrobial stewardship programs. Hospitals are an important place to address antimicrobial use, but the community can’t be ignored.

A couple hundred thousand kgs of antibiotics isn’t cheap. The total cost in 2016 was ~$766 million.

  • How much of that was unnecessary isn’t known. However, even if we just think about antibiotic cost savings (ignoring resistance, complications and various other potential problems), reducing antibiotic use can have a big financial impact.

625 prescriptions were dispensed in the community for every 1000 people.

  • This number has been pretty stable over the past few years, but varies across the country.

When provinces and territories are compared, Newfoundland is gets the dubious distinction of leading the way in prescription rates per 1000 people, at 954. The Territories (a combination of the Northwest Territories, Yukon and Nunavut) had the lowest rate, at 280/1000.

provinces copy

The most common drugs have remained pretty stable. Amoxicillin was the most common drug, followed by azithromycin and cephalexin.drugs per 1000 copy

Prescription rates for kids < 14 yrs of age have been decreasing since 2011.

  • Maybe that’s due to increased awareness (both physician and parent) about how often some common problems like ear infections are viral and don’t need antibiotics.

Some good news. Some concerning news. Lots of room for improvement….and an indication of the need for more extensive antimicrobial stewardship programs.





Antibiotic Resistance Surveillance Report: 2017

The 2017 Canadian Antimicrobial Resistance Surveillance System report was just released, and as always it contains a lot of interesting data. Some are interesting, some are concerning, some are encouraging.

I’m not going to try to distill a 90 page report into a quick blog post, so I’ll cover some highlights separately. Let’s start with the Antimicrobial Use in Animals section. Here are some highlights (and some comments).

In 2016, ~1 million kgs of medically important antimicrobials were distributed for sale in Canada.

  • This misses some antibiotics that get used via loopholes, but those are being closed and this number is probably a pretty good overall estimate.
  • The great news…that’s 14% lower than 2007 and 17% lower than 2015

600,000 kg of ionophores and chemical coccidiostats were also distributed.

  • These are often combined with other antimicrobials (sometimes to make things look scarier) but they are irrelevant from antibiotic-resistance and public health standpoints. So, it’s good to see them separate (and to essentially ignore them).

99% were intended for food animals, on a per kg basis.

  • This is always hard to interpret and sometimes leads people to think that companion animals are irrelevant. We have to be a bit wary focusing just on per kilogram data (1 kg of antibiotic treats a lot more Chihuahuas than cattle). The main antibiotic classes used in companion animals were cephalosporins, beta-lactams and trimethoprim-sulfa, all drug classes of high importance.
  • What this shows to me is that we can have a huge impact on overall use focusing on food animals. However, the drugs that are used in pets are often the same as those used for serious infections in people, and we share bugs readily with our pets. So this numbers shouldn’t be taken as an indication to ignore them.

Fluoroquinolones decreased by 56% from 2015 to 2016.

  • Wow. That’s great, since this is one of the biggest classes we’re worried about. They’re important drugs (for both humans and some animal species) but are prone to overuse.

Most of the distributed antibiotics were those intended for use in feed, accounting for 76% overall. On the opposite end of the spectrum, intramammary drugs (used for mastitis in cattle) accounted for <1%.

When everything is put together on a per kg basis, 78% of antimicrobials distributed or sold in 2016 were for food animals, 20% were for humans, 1% for crops and 1% for companion animals.

  • Again, be somewhat wary of crude kg numbers (since the relevance of a kg of tetracycline is probably much, much less than a kg of a fluoroquinolone) and there are approximately 19 times more animals in Canada (excluding farmed fish) than humans, but they give some idea of how we use antimicrobials in this country and they give us numbers for comparison over time. The 2nd figure below is an interesting on to think about.

Some people will take these numbers and use them to spin certain agenda. However, we’re better off using them as the basis for more surveillance, more interventions and more research to reduce and improve use of antibiotics in Canada, whatever species they go into.

1 copy

2 copy