“Organizational silos,” and how they prevent effective zoonotic disease tracking

It appears that the agencies that we rely on to track disease outbreaks need to start tracking disease, not just their own jurisdiction.

An article in Sociology of Health and Illness piqued my interest this last week that reveals the amount of segregation different government agencies have when dealing with zoonotic disease. The understanding of the goals and connections between livestock, wildlife, and human health among these agencies are often apathetic at best, and antagonistic at worst.

The author of the article took it upon himself to interview several government agencies with different species and regional jurisdictions, and was able to reveal what he calls “organizational silos” that develop when the values and cultures of these different agencies prevent them from working with outside groups. When attempting to monitor emerging infectious disease (EID), identification of cross-species movement is critical to predicting and preventing pandemics. Unfortunately, while they may be able to acknowledge the geographical movement of EID’s, many organizations are blinded by their specific oversight of humans or animals.

Copied from the article: Diagram showing the crossover between domestic animals, wildlife, and human EID. Important emergence factors for each circle are listed on the outside.

There are many telling comments contained in his interviews, and I encourage you to read the article to get the whole scope of the problem, but I’ve chosen to list a few of my favorites here:

From the Director of Animal Health Division at a state Department of Agriculture:

“‘We got a positive [flu result] on one of our routine surveillance tests’ of a poultry farm, Spencer complained, and ‘we were required to contact the USDA right away because of the pandemic Asian strain’. Spencer added, ‘It seems a little silly because there was no clinical illness on the property, and the strain came back something pretty common…’ In Spencer’s eyes, it was ‘hard to justify’ reporting the flu strain to the USDA… These days, Spencer said he passes on information about disease events to the state DOH and leaves it to them to tell local health boards. ‘If somebody screws up’, he shrugged, ‘at least we can blame the [Department of Health]’.”

Not an uncommon perspective for many organizations, or even coworkers! Let’s hear from another director at the USDA Animal and Plant Health Inspection Service (APHIS):

 “Clinton argued that the ‘single biggest threat for disease’ comes from ‘wildlife intermingling with domestic livestock’. He told me, ‘You can’t control the birds’ and he rightly pointed out that ducks are flu incubators. If the bird flu – which Clinton called the top priority of his agency – becomes pandemic in humans, he told me, it will come from waterfowl.”

Interesting, I might argue that we have much more interaction with domestic fowl (can’t remember the last time I handled a wild duck), but let’s see what others had to say about this viewpoint.

“Nina Marano, a zoonotic disease expert at the CDC, told me that ‘most of the outbreaks have occurred through interaction with domestic poultry’. Another example: though poultry farmers singled out wild birds called cattle egrets as the source of a 2004 flu outbreak in California, the egrets tested negative – it turned out that contaminated egg containers circulating between farms were the culprit (McNeil 2004).”

Finally, one last example of how a zoonotic disease often isn’t treated as such by human health agencies. From a Director of the Infectious Disease Bureau of a city Public Health Commission:

“When I asked Sanders to describe a zoonose that she responded to, she mentioned a recent outbreak of salmonella…and she believed that the pathogen came from two live poultry markets in Chinatown. What I found telling was that, in Sanders’ lengthy discussion of this outbreak, she did not mention any communication with veterinary medicine agencies.While the Disease Bureau’s response to salmonella followed protocol, it did not turn to the Department of Agriculture, the USDA, or any other agencies involved in animal health for help or information. Nor did it share information with them.”

Clearly here the city health board considered this a food safety issue, but payed no attention to the implications of getting meat from an approved source (a domain which definitely belongs to the USDA), or the fact that other agriculture agencies may be interested in a salmonella outbreak. There are many other telling quotes within these interviews, and I again encourage you to check out the article.

The author of the study concludes that the only examples we get of harmonious collaboration are for those diseases which are in the public eye such as rabies and influenza (H5N1 and H1N1), though we still have lines drawn even when the public is asking for action (“‘we have enough H1N1 to worry about without worrying about turkeys’. He
concluded that turkey infection is ‘a Department of Agriculture issue’”). The most shining example of the failure to communicate by these institutions in the article is the discovery of Bird Flu in the US.

The first human cases of H5N1 in the US were wrongly diagnosed with St. Louis encephalitis, resulting in the deaths of 3 patients. A veterinary pathologist at the Bronx zoo observed neurological symptoms in some of the zoo’s birds and suspected a link, however encephalitis would not have killed her birds. Both the CDC and local DOH would not accept new information from her, instead keeping the encephalitis diagnosis. She then sent specimens to a friend at an Army Medical Research Institute of Infectious Diseases, who revealed the etiology of the disease and I’m sure had a hilarious conversation with the CDC and DOH (could you please explain to us why this veterinarian is doing your job casually on the side, and doing it better?). By the time the CDC received/accepted this information, H5N1 was endemic in the area.

Nothing against the CDC, it’s a fantastic organization, but this highlights the closed lines of communication that exist between human and animal agencies the author discusses. In order to prevent the next EID crisis, rigorous epidemiology is critical. Refusing to acknowledge the importance of cross-species movement to the virulence and emergence of a disease that falls under your agency does not only prevent you from identifying the next source of infection, but leaves you with nothing but reactive measures catered to a epidemic that you refuse to fully appreciate.


Jerolmack, C. (2012). Who’s worried about turkeys? How ‘organisational silos’ impede zoonotic disease surveillance Sociology of Health & Illness DOI: 10.1111/j.1467-9566.2012.01501.x

Staphylococcus aureus diversity and subclinical mastitis

This is the first study I’ve found that was interested in cataloging bacterial diversity among subclinical (or asymptomatic) infections. While they may be less threatening to the animal’s overall health, these infections have great significance in the world of animal agriculture, where they restrict growth (or in this case, milk production), and encourage the use of medicated feeds which in turn motivate people to purchase organic products. Identifying the risk factors and causes of these infections could therefore impact both the management of food animals, and any legislation defining how and when medications can be used. With that in mind, let’s jump back into mastitis, and everyone’s favorite gram-positive, S. aureus.

It’s because of my plasmids, people can’t help but stare.
Image from http://cellimagelibrary.org/

S. aureus is one of many bacteria that cause mastitis, however it is of additional importance as it often causes chronic or recurring cases of mastitis that result in unusable milk and discomfort of the animal. In this study, the authors investigated 11 dairy farms where they expected to find S. aureus, based on previous culture findings at each farm. They defined cows that they took milk samples from as having new or chronic infections based on somatic cell counts (SCC) in the milk. If values were >200,000 cells/mL for the month of collection the infections were considered new, whereas if cell counts were  >200,000 cells/mL for more than 2 months, those infections were considered chronic. They took a single milk sample from each teat of the infected cows, for a total of 1,354 mammary glands from 350 cows.

Pulse field electrophoresis was used to identify the different subspecies/serotypes/pulsotypes (pick your word), and to identify the genes coding for enterotoxin production that had been amplified by PCR. An ELISA test was used last to detect the presence of several enterotoxins.

As the majority of exposure to enterotoxins produced my S. aureus is through milk and dairy products, subclinical infections of S. aureus are very important as a food safety control point. Unlike cows with clinical cases that are removed from production, cows with subclinical infections continue to contribute milk that makes it to the consumer, provided that the SSC is <750,000 cells/mL. The authors were unable to detect a large amount of enterotoxin in their samples, but many of the pulsotypes contained the genes coding for their production. Other studies cited by the author report the common presence of these genes in S. aureus  samples, but expression rates are inconclusive or unexplored. This means that theoretically, subclinical cows could be introducing these bacterial toxins into consumer milk in small amounts.

It’s difficult to tell how significant these amounts might be. Toxic doses of one of the enterotoxins, “Toxic Shock Syndrome Toxin 1”, has been found to be as low as 100 micrograms/Kg in miniature pigs. The concentrations that may be introduced through contaminated milk, and the bioavailability when ingested, should be explored. Takeuchi et al. (1998) were able to detect the presence of TSST- 1 in bulk milk tanks, but no one has yet to quantify the amounts of TSST- 1 potentially present in pasteurized milk.

All that being said, what good is this new information? It can be argued that because these infections are chronic and/or subclinical that these strains of S. aureus aren’t very pathogenic, but they’re still causing inflammation. By identifying common serotypes and factors leading to the subclinical infection of a herd, perhaps there are simple management changes that can prevent infection. Milking is an almost sterile procedure, with sanitation of the teats both prior and following milking, wearing gloves, and forestripping; but there could be other tricks that would target risk factors related to the spread of subclinical pathogens, especially those that are specific to a location.



Bulanda M, Zaleska M, Mandel L, Talafantova M, Travnicek J, Kunstmann G, Mauff G, Pulverer G, & Heczko PB (1989). Toxicity of staphylococcal toxic shock syndrome toxin 1 for germ-free and conventional piglets. Reviews of infectious diseases, 11 Suppl 1 PMID: 2928643

Oliveira L, Rodrigues AC, Hulland C, & Ruegg PL (2011). Enterotoxin production, enterotoxin gene distribution, and genetic diversity of Staphylococcus aureus recovered from milk of cows with subclinical mastitis. American journal of veterinary research, 72 (10), 1361-8 PMID: 21962279

Takeuchi, S., Ishiguro, K., Ikegami, M., Kaidoh, T., & Hayakawa, Y. (1998). Production of toxic shock syndrome toxin by Staphylococcus aureus isolated from mastitic cow’s milk and farm bulk milk Veterinary Microbiology, 59 (4), 251-258 DOI: 10.1016/S0378-1135(96)01253-9

Article Review: Vaccine Reactions

I’m apparently still on this immunology kick, because I seem to be finding it everywhere. Heck, I recently learned that we’ve cured the allergic response to peanuts and eggs in lab mice. Check out the link, the author is hilarious and the material is interesting.

These two articles offer a great look at the overall prevalence and risk factors associated with vaccine-associated adverse events. The components within the vaccine that cause these events are the antigen itself, adjuvants, preservatives, stabilizers, and residues from the tissue culture used to grow the vaccine (Moore, 2005). Vaccine reactions are similar to any acute allergic reaction, and can present with a variety of mild to severe symptoms. The mild being lethargy, anorexia, fever, edema (generalized or local to the injection site), pruitis, uticaria (hives or wheals), and pain at the injection site; the severe being vomiting, dyspnea (labored or shortness of breath), and anaphylaxis. There’s a lot of information about when certain symptoms tended to occur at intervals after the vaccines were given, but any reaction that isn’t within the first 3 days is pretty much never going to be life threatening. If anaphylaxis is going to occur, it’s going to be immediately following vaccination.

The really useful information was the breakdown of risk factors that can be used for client communication. I’ve decided to discuss them here, broken down into cat and dog categories.

Image from meow-cats.com

First, let’s start with cats. I almost like these numbers more because you don’t have to take into account bias based on animal or breed size, as most cats fall into the <20 lbs category. Nonetheless you still have to remember that a 4 pound kitten does way only a fraction of that 5 year old chubby (BCS >5 on a 9 point scale) longhair it will grow to be.

So the first two risk factors require a little bit of thinking in context to explain the numbers. It seems that cats weighing 2-4 Kg (4.4-8.8 lbs) and approximately one year of age are most at risk compared to other weights and ages. The high numbers for these groups can be explained by the number of first encounter events that occur. If you’re recording vaccine reactions, you will record less in older age groups and higher weights (low weight under 10 lbs is going to be suggestive of a young age rather than a smaller cat), because if an adverse event occurred at a young age, either the animal is no longer vaccinated or steps are taken to reduce its risk (medication, strict scheduling, vaccine selection). That being said, just because the numbers are higher by circumstance, this information is very relevant in a clinical setting. Knowing the epidemiology of these events can help technicians at clinics determine when the discussion of vaccine reactions is “routine” or “protocol”, or when it really needs to be a time to educate the client. Vaccine reactions may need to be just a bullet point when Schrodinger is there for his 4th rabies booster and a discussion when he’s receiving his kitten series.

Sex and neuter status have a large impact on reaction risk as well. Intact males actually have a lower risk of adverse events than neutered males and spayed and intact females. Apparently, estrogen has an immune boosting effect, while testosterone has an immune suppressing effect. This benefits intact males when it comes to all allergic reactions (and possibly auto-immune disorders).

Here’s the big one, and the one clinics have the most control over. With each additional vaccine given in a single visit, the risk of an adverse event increases by 28% in cats. That’s huge. Any cases of severe anaphylaxis or death recorded in the study were preceded by the animals receiving 3 or more vaccines in one visit. So clearly the biggest thing any clinic can do to prevent adverse events (or at least severe ones) is to adopt a vaccination schedule that prevents multiple vaccinations from occurring within the same visit. This can be difficult as clients will not want to end up paying for multiple exams throughout the year, but with boosters outside of rabies, exams shouldn’t be necessary unless an annual or other scheduled exam is due. As far as specific vaccines being more prone to adverse events, the only suggestive evidence was when both FVRCP and FeLV were given within the same visit. This is explained by both having two concurrent vaccinations given, and also the theory that vaccines containing multiple antigens or covering multiple serovars (multivalent) are more likely to illicit reactions. Interestingly, while clients are often scared by the potential for vaccine caused neoplasia from the rabies vaccine, it was among the lowest reaction rates observed with the administration of a single vaccine.

Image from peteducation.com

Dogs had much more biased data within the age and breed groups because there is an obvious relationship between body mass and the potential for reaction. When looking at the dog population, a chihuahua can be as little as 6% of the weight of a bullmastiff, yet they receive the same 1ml dose of vaccine. This means that an 8 lb Chihuahua is going to receive proportionally 15 times more vaccine than a 120 lb bullmastiff. Not surprisingly, this causes a bit of inflation in the number of reactions in groups that are smaller in size, such as toy breeds and puppies. The highest risk group in size was 0-10Kg (0-22lbs) and the highest risk age was approximately 2 years of age (with higher rates for <2 than the rates of >2).

Just like I mentioned before when talking about cats, the greatest risk factor for reactions in dogs was the amount of vaccines given in one visit. The difference though, is how the large weight distribution in dogs makes this even more important. Small dogs (<10Kg) are similar to cats in that their risk increases by 24% with every additional vaccine administered that visit, while large  dogs (10-45Kg) increase their risk by 12 percent. All 3 dogs in the study that suffered fatal reactions received 4 or more vaccines at once.

Breed dispositions were difficult to pinpoint, as the suspected breeds are all small breeds which suffer a higher rate of reaction already due to their size. There is suspicion that dachshunds may be predisposed to allergic reactions in general, but so far the evidence is inconclusive concerning vaccines. Only the Lyme vaccine appeared to carry a higher risk than any other, showing again that, with the exception of neoplasia concerns, rabies does not carry with it any additional risk. Spayed and neutered animals, as in cats, are more susceptible to reactions; however the difference between intact and spayed females is much larger in dogs than in cats (where they are nearly identical). Dogs do seem to display an interesting trend where vaccine reactions are more likely to occur on the 3rd booster in a series, likely catching clinicians and clients off guard as they have received the first two without incident. This just states again that the puppy and kitten periods (and new patients) are of much more relevance when discussing vaccine reactions with clients.

The articles are both great, and contain an excellent statistical analysis of millions of animals. They provide a great overall picture of the epidemiology of vaccine-associated adverse events, and are definitely worth a read for both veterinary doctors and staff. Knowing a couple of the more important statistics can reassure the client and lend credibility to technicians that are responsible for discussing these issues.

ResearchBlogging.orgMoore, G., DeSantis-Kerr, A., Guptill, L., Glickman, N., Lewis, H., & Glickman, L. (2007). Adverse events after vaccine administration in cats: 2,560 cases (2002–2005) Journal of the American Veterinary Medical Association, 231 (1), 94-100 DOI: 10.2460/javma.231.1.94

Moore GE, Guptill LF, Ward MP, Glickman NW, Faunt KK, Lewis HB, & Glickman LT (2005). Adverse events diagnosed within three days of vaccine administration in dogs. Journal of the American Veterinary Medical Association, 227 (7), 1102-8 PMID: 16220670