“The way the dogs are trained is as rigorous as anything else we do in the diagnostic lab,” says Marthe Kenny Charles, MD, a medical microbiologist at Vancouver General Hospital in British Columbia. “The goal is for us to identify the areas in the patient environment that are causing the promulgation of this organism. We wanted to have a way to ensure that the environment is being cleaned adequately.”

According to a recently published paper1 by Charles and colleagues, “In a series of examinations that assessed odor recognition and search capability, the canine team demonstrated 85.7% sensitivity and 94.7% specificity.”

After initial success with one team, a second dog and handler have been added. No one wants C. diff in the patient environment, but Hospital Infection Control & Prevention asked Charles if the dogs sometimes detect trace amounts incapable of causing infection.

“That’s a good question,” she says. “As you can tell in some of our recent publications, that is something that we are starting to look into in terms of trying to [quantify] the amount the dogs are able to smell. In terms of how that translates into the infectivity of [C. diff], I don’t think we have the answer at this moment.”

Likewise, there are some signs of C. diff infections dropping, but as is often the case with infection control, it is difficult to ascribe a multifactorial effect to a single intervention.

While the dogs are identifying C. diff that may otherwise linger in the environment, the effect goes beyond that in those who witness the dogs at work. The detection rounds typically include an infection preventionist (IP) along with staff nursing and environmental cleaning.

“It opens communication and conversation,” Charles says. “The IP sometimes provides guidance on how to clean whatever items have been alerted by the dog. I think there is something really visual about the dogs alerting in different areas.”

In that sense, the dog detection event resonates as a reminder to the nurses and staff of the important of infection control and environmental cleaning. The dogs also are used on units that have experienced recent cluster events with antibiotic-resistant organisms like methicillin-resistant Staphylococcus aureus (MRSA).

“Because we know that MRSA, for example, is the type of microorganism that can be transmitted by contact, the belief is that if there has been a breach in terms of infection control practices, then there might be C. diff and other organisms that could be transmitted,” Charles says.

Teresa Zurberg, RCT, is one of the dog handlers tracking C. diff in the hospital. She is a validated explosives and narcotics canine detection handler and co-authored a recent article on the Vancouver program.2 Zurberg works with an English Springer Spaniel named Angus. Hospital Infection Control & Prevention asked about her unusual job in the following interview, which has been edited for length and clarity.

HIC: How did you get involved in the C. diff detection program?

Zurberg: In 2013, I had C. diff. I had a wound on my leg that became infected and I had to go in for some IV antibiotic therapy. I ended up with C. diff and was in the hospital for about a week. I lost about 20 pounds, and when you are only 5-foot-1, 20 pounds is a lot of weight. I have never been so sick in my life. At that time, I was a professional canine handler with bomb dogs and drug dogs. We had a Springer Spaniel puppy that I knew was going to be a working detection dog, but we hadn’t decided what for. My husband, who is a nurse in quality and patient safety at the hospital I work at, came across an article about a dog in the Netherlands who had been trained to look for C. diff in patients. He asked me, “Can you train a dog to do this?” I said, “Yes; if it’s got an odor I can train a dog to find it.” That’s when we approached the quality and patient safety executive team and proposed training a dog to do this. But instead of looking at patients — because you always test patients — [we proposed] to look at the environment. There is no logistically feasible way of testing the environment for reservoirs of C. diff.”

HIC: The reports indicated that dogs are detecting C. diff with a high sensitivity and specificity. Do you, as an expert, feel that the dogs are performing this task sufficiently to justify the program?

Zurberg: Definitely. No detection dog in the world is ever going to be 100%. You have to factor in so many things, like handler error. The dogs are the easy ones to train. It’s the people who take longer to train.

HIC: Some may assume anyone could walk a dog around the hospital. What happens in your interactions with the dog? Do you pick up cues?

Zurberg: To be a canine handler is a profession. It’s not just any person off the street, even though we have people who try to do it all the time. First, you have to have the right dog. If I go look at 100 dogs, I might shortlist 10. Out of that 10, I might pick one. All dogs can sniff, but not all dogs can work, especially in the environment that we work in. In the world of detection, we work in environments that are fairly extreme. We have people around us all the time, we have to navigate hazards, we have to talk to people while we are still looking and working the dog. Then, answer questions if the dog alerts.

There is a lot of rigor that goes into everything we do to make it a credible program. There is the sport of “nose work,” which I have judged and used to teach. A lot of people think it is just that easy; you do a three-minute search holding a leash and follow the dog around. But there is so much more going on. We have to have medical backgrounds and understand infection control practices and procedures. We have to be familiar with hospital environments and be able to do education in the moment. All this kind of happens at the same time, so it takes a lot to do it.

HIC: When you are inspecting a terminal cleaned before the next patient is transmitted, what is the process? Do you unleash the dog?

Zurberg: We never take the dog off the leash at any time. We do the entire hospital environment. If we go to a up to a unit, we are doing the hallways, everything in the hallways, the nursing stations, the clean supply rooms, the linen carts, shared sinks and showers. We do not generally do rooms that have patients in them. If you think about the kind of contamination that they are hitting on, when a dog tries to find drugs or bombs they are usually trying find a package that someone had tried to hide. They are concealed and there are only little pinpoint dots of odor coming out. Whereas, for C. diff contamination, something has rubbed up against something else to cause contamination. The odor is kind of in a smear pattern, and depending on how big it is and where it is, it will show a change of behavior in the dog. Our dogs will either sit, lay down, or pace if it is up somewhere high and is a large contamination. We spend thousands of hours learning the dog’s communication to be able to interpret that.

HIC: If something is detected, who is contacted?

Zurberg: When we go out to screen a unit, we have people who come with us. We have somebody from the environmental services team, and we also usually have a unit rep or an infection preventionist who comes with us as well. Say the dog hits on a [piece of equipment] that is the nursing staff’s responsibility to clean. They get them to have it cleaned. Everything we do is in the moment. As soon as the dog alerts, actions are taken to mediate that alert.

HIC: It sounds like the dogs really engage the staff in a unique way.

Zurberg: The dogs have been able to show us things we never considered. First, they make the infection control teaching tangible — people see why we tell them to wash their hands and to clean equipment after use. You see the dog working and then they sit and alert, and people are like, “Oh, now I get it.” A lot of times we have alerts in clean supply rooms. Why? Because people are not taking off their gloves. They are dealing with the patient, they have gloves on, forget something, and go grab it out of the clean supplies. They have now contaminated the clean supplies, and the dogs make these lessons tangible in the moment.

HIC: How do patients and families react to the dogs when they see them in a hallway or an empty room?

Zurberg: We have floppy-eared dogs for a purpose in the hospital. Both the other handler and I used to work German shepherds and malamutes when we did explosives and narcotics. But in the hospital, you need a softer-looking dog. We have people there who are scared of dogs, and there are some cultures where dogs are not always a good thing. We want to be respectful of that. The dogs we have are just generally cute and they open conversations that may have never taken place. It allows for us to have in-the-moment education with patients, families, and staff.

HIC: What is something unusual that you have found working with the dogs?

Zurberg: When the bathrooms are cleaned, everybody cleans the outside of the toilet paper dispenser, and we didn’t think about it until Angus’ nose went to the “inside” of the toilet paper dispensers. It makes sense, but it was something we hadn’t thought of before. Our environmental services team changed practices and let their cleaners know that this is a hot spot.

HIC: How does the dog decipher the signature of C. diff from fecal matter?

Zurberg: The dog’s nose-brain connection is different from humans. Dogs can scent-differentiate. In terms of senses, dogs are born nose, eyes, ears. A larger part of their brain is directly used for scent differentiation — that’s how they take in their world. Before we start working a dog in the hospital we go through a rigorous validation and certification process. We do double-blind, third-party testing on the canine teams before we start working. First, the dogs are only ever put on one odor. The only odor that is important to these dogs is C. diff — that is the only odor they are trained on. When they alert I can say, “That’s C. diff.” When they are tested and validated to work, we do odor recognition tests. They have to pick out the correct odor that is C. diff positive. They have to ignore the distractor odors. So, Angus has to pick a C. diff culture out of the lineup and also a C. diff fecal sample that is positive. At the same time, he has to ignore and walk past a fecal sample that does not have C. diff in it. They smell in parts per million and parts per billion.


  1. Charles MK, Wang Y, Zurberg T, et al. Detecting Clostridioides (Clostridium) difficile using canine teams: What does the nose know? Infection Prevention in Practice 1 2019;1:100005. Available at: https://bit.ly/2k9lmz1. Accessed Sept. 12, 2019.
  2. Li C, Zurberg T, Kinna J, et al. Using scent detection dogs to identify environmental reservoirs of Clostridium difficile: Lessons from the field. Can J Infect Control 2019;34:93-95.