Visit any online forum dedicated to cats and dogs and it won’t be long before some member posts a comment about vaccination. It is a topic that never fails to open up passionate – and often heated – debate between those that ‘do’ and those that ‘do not’. Well, this article is not about the pros and cons of vaccination per se.
Rather, it is about a paper recently published in the journal Animal Conservation called “To vaccinate or not to vaccinate: lessons learned from an experimental mass vaccination of free-ranging dog populations” (Belsare and Gompper, 2015).
This paper gives some fascinating insights into the epidemiology of vaccination that every pet owner concerned about the pros and cons of vaccination can learn something from. The paper reports the results of a study carried out on a population of free-ranging dogs living in and around a group of villages in rural India. And, as the title of the paper suggests, the aim of this study was to determine how effective a mass vaccination program would be against canine adenovirus, canine parvovirus and canine distemper virus. These viruses, of course will be very familiar to dog owners living in affluent countries where preventive veterinary care is the norm and routine vaccination comes as part of the package.
To make the most of this paper, you need to understand some basic concepts about the epidemiology of infectious diseases, natural immunity and vaccination. Specifically, the mathematics of how natural immunity and vaccination are supposed to work in preventing the spread of infectious diseases through large populations of individuals – that is, how vaccines help to prevent epidemics, and the importance of understanding the role of natural immunity in mass vaccination programs.
Herd immunity, vaccination and the magic number ‘67’
The concept of a population of individuals developing a natural ‘herd immunity’ to a given infectious disease has been around for almost as long as vaccinations themselves. It did rather fall into disuse, but has re-emerged in the last few decades when scientists and accountants started speaking to each other. On the one hand, the scientists were interested in the benefits of mass vaccination programs to help manage, or ideally eradiate infectious diseases, while on the other, the accountants were interested in the costs / benefits of such programs.
Herd immunity arithmetically describes how the ratio between (the number of individuals that are immune to a given infectious disease) and (the number of individuals that are not immune, i.e. susceptible, to the same disease) alters the spread of that disease through the population. How an individual developed their immunity to the disease, for example through previous infection, or vaccination, is irrelevant.
The equation for describing herd immunity is (R0 − 1) ÷ R0 (Fine at al., 2011). This can be best explained using an example with a diagram, and its derivation is also shown at the end of this article for those interested in knowing the maths.
The case of Charlie and an outbreak of disease X
Imagine all the dogs living in London and let’s suppose that there are one million; the actual number is immaterial here. Now, suppose that 1 London dog, we’ll call him Charlie, is infected by a virus (called virus X) while on holiday overseas one summer with his family overseas. X does not make dogs particularly unwell and, in fact, his owners don’t even know he has it. Back Charlie comes into the UK and, soon enough, he’s out and off on his usual daily walks where he meets all his doggie friends. The X virus is rather infectious, like the common cold in humans, and easily passes from one dog to another. Once infected though, the dog’s immune system launches a response to the virus and consequently produces antibodies that will protect the dog from reinfection with the same virus for the rest of the dog’s life.
However, X is a new virus to the UK, thanks to Charlie, and all the dogs in London and beyond have no immunity to and are therefore susceptible to contracting the disease. On average, each day on his walk, along with his doggie friends that he sees on a regular basis (they have all already caught the virus of course), Charlie meets and greets another 4 other dogs he has NEVER MET BEFORE, and all of them become infected with the X virus too. So, just considering Charlie’s role is spreading the disease here, on day 1 there is 1 case, Charlie himself. On day 2 there are 4 cases and on day 3 there are 16 cases and so on and on and on. Of course, in general, infectious diseases spread over a longer period of time depending on the incubation period of the virus and how infectious it is, but regardless of the time course of the infection, the numbers remain the same. Remember also that all the infected dogs are now spreading virus X to all their friends. In addition, every day, they are also meeting another 4 other dogs they have NEVER MET BEFORE and infecting them as well.
From this you can see how quickly virus X spreads around the dog population in London. As time goes on, and as more and more dogs become infected and then recover, because all the dogs develop an immunity to the virus, they can’t be infected for a second time. We now have three different populations of dogs:-
- POP-1. a population that have already had virus X and recovered, they are not now spreading the disease to other dogs.
- POP-2. a population that currently have virus X and are infectious, they are spreading the disease to other dogs
- POP-3. a population of dogs that have not yet been exposed to and caught virus X, they are susceptible to the disease.
As all the dogs in London continue on their daily doggie business, at some point during the spread of the infection, there comes a time when there are no longer enough susceptible dogs left in the population (POP-3) for it to be possible for the dogs still spreading the disease (POP-2) to meet 4 of them. Instead, each day, the infectious dogs meet 4 other dogs they have NEVER MET BEFORE, but only 1 of these is susceptible (POP-3) and catches virus X. The other 3 dogs have already contracted the disease and are either spreading the virus around themselves (POP-2), or have already recovered and are immune (POP-1).
This point during the spread of virus X through the London dog population is really important because it means that the number of NEW CASES of dogs with the disease and that ARE INFECTIOUS has plateaued, as shown in the diagram.
To start with, R0 is defined as the number of new cases per day that Charlie infects right at the start of the outbreak while all the dogs in the population are susceptible. R0 is the Basic Reproduction Number of the virus in the dog population, and in this outbreak, R0 = 4 (or 400%). As more and more dogs are infected and develop an immunity, the number of new cases starts to fall below R0.
When the number of new cases reaches 1 (or 100%), then each INFECTIOUS dog is INFECTING just 1 other dog. In other words, replacing itself as the INFECTIOUS dog as he recovers from the virus. Now, let’s put the numbers into the equation above:
(R0 − 1) ÷ R0
When each INFECTIOUS dog is only infecting 1 in 4 new dogs he meets, then:
(4 – 1) ÷ 4 = 0.75 (or 75%)
This number is the Herd Immunity Threshold (HIT) for virus X, and it tells us that for this disease, 75% of the population of dogs need to be immune to the disease in order for the disease to stop spreading. How they got that immunity – through catching the disease or through being vaccinated – is irrelevant.
Now consider what would happen if virus X had an R0 of 10. Now, when the number of new cases reached 1, the equation, (R0 − 1) ÷ R0, would be (10 – 1) ÷ 10 = 0.90 (or 90%). As you can see, for an R0 of 10, the HIT is 90%, so 90% of the population of dogs need to be immune to the disease in order for the disease to stop spreading. R0 thus reflects the virulence of the disease.
Here are some well-known human diseases along with their initial reproduction numbers, R0 (Rodpothong and Auewarakul, 2012). You can easily work out the HIT for these diseases for yourselves:
|12 – 18
|5 – 7
|5 – 7
|5 – 7
It is much harder to find R0 values for feline and canine diseases in the literature, but here are a few I have managed to dig up:
|Canine Distemper Virus
|Nouvellet et al., 2013
|Rabies (Dogs, Tanzania)
|Hampson et al., 2009
|Rabies (Dogs, Hong Kong)
|Hampson et al., 2009
|Rabies (Foxes, Italy)
|Nouvellet et al., 2013
|Common feline infectious diseases
|Horzinek and Thiry, 2009
Now, here’s a really important point about HIT. Let’s go back to Charlie and disease X for a moment, which started with an R0 of 4. We know that in this outbreak, HIT is reached when for every FOUR dogs each infected dog comes into contact with, only ONE dog contracts the disease. That is, the new R0 = 1. In order for the disease to start to die out in the dog population, R0 needs to fall below 1, i.e. R0 < 1.
It is on this principal that the logic of vaccination is built – the goal of vaccination is to achieve R0 < 1 for the target disease. As you can see from the figures shown above, the initial R0 values are generally less than 3 for common infectious diseases in people and in cats and dogs. A good starting point for effective vaccination against these diseases is to set a goal of achieving the HIT based on a Basic Reproduction Number greater than it actually is. This is why for many common diseases in all animals, vaccination programs are based on an arbitrary assumption that the Basic Reproduction Number is 3. This number sets a target of achieving effective immunisation through the vaccination of at least 67% of the population (Horzinek and Thiry, 2009; Keeling and Rohani, 2008). This is the magic number 67.
It must also be remembered that vaccinations are rarely 100% effective in a given population for all sorts of reasons. For example, some humans and other animals are poor responders and do not develop adequate immunity as a result of vaccination. Some individuals may decide against vaccination for themselves or their pets for various reasons. Most vaccines are relatively unstable with a short shelf life so improper handling and storage can easily destroy them, for example being left out of the refrigerator for an length of time. Assuming a HIT of 67% helps mitigate against some of these failures.
The risks of not having effective immunity in local populations of dogs is made all too clear in this example. In February of 2010, there was a massive and devastating earthquake and tsunami in Chile. Along with all the human suffering brought on by this disaster, there was an outbreak of canine distemper (CDV) in the pet dog population of the small town of Dichato. Investigators of the outbreak found that only 35.5% of these dog had been vaccinated against CDV (Garde et al., 2013).
In addition, we need to remember that rabies still kills around 20,000 people every year(Belsare and Gompper, 2013)
With this information on-hand about herd immunity and vaccination, there is more we can learn from the paper being reviewed in this article – “To vaccinate or not to vaccinate: lessons learned from an experimental mass vaccination of free-ranging dog populations” (Belsare and Gompper, 2015).
There are many populations of free-ranging domesticated animals all over the world which act as reservoirs for a number of infectious diseases that put wildlife in surrounding areas at risk, especially those that are endangered species because of loss of wild habitat. For example, all canids are susceptible to diseases such as rabies virus, canine adenovirus(CAV), canine distemper virus (CDV) and canine parvovirus (CPV), and mass vaccination programs have been the cornerstone of the management of such diseases in both domesticated and wild species all over the world for many decades. Targeted vaccination programs against rabies in wildlife using food treats baited with vaccine such as Raboral baits has been very successful in some parts of the world.
This is an interesting study because its roots lie in a previous study by the same authors (Belsare and Gompper along with others) in 2007 of an epidemic of CDV in Indian foxes (Vulpes bengalensis) living within the 1,222 square kilometre Great Indian Bustard Wildlife Sanctuary (GIB WLS), Nannaj in central India (learn more about this wonderful place here http://kadambarid.in/nannaj.html). Antibody titre testing showed they had low titres for CDV. This was because of their isolation in relatively small numbers, leaving the foxes vulnerable to the disease which was killing off significant numbers.
This area also has a large population of free-ranging dogs living there too, and it was suspected that these dogs were responsible for infecting the foxes. In 2011, as a result of these findings, the conservation area authorities decided to undertake a mass vaccination program of the dogs (not the foxes!), and this was organised and undertaken by Belsare and Gompper. While they were capturing and vaccinating the dogs, the researchers also took blood samples and tested them for CAV, CPV and CDV. The results showed that the dogs had high titres for all these diseases which meant that they were already endemic (naturally occurring and stable over a long period of time) within the dog population.
As discussed earlier, the value of any mass vaccination program is based on the assumption that within the target population of animals, there are many susceptible individuals. In other words the Herd Immunity Threshold (HIT) for the disease has not been reached and R0 is still above 1 at the time of vaccination. If this was not the case and many animals within the population had already caught the disease, recovered and were now immune, then vaccinating them would be pointless.
Belsare and Gompper decided to investigate this further by carrying out a small-scale, village-level study during the 2011 mass-vaccination program, and it is this study that is the subject of the review you are now reading. The working hypothesis they wanted to test was this –
“…a mass vaccination program against endemic (enzootic) pathogens would reduce the number of seronegative (susceptible) dogs in the population, thereby contributing to the herd immunity as well as reducing the occurrence of clinical and subclinical cases caused by these pathogens.”
NOTE: the key word here is endemic.
The 1,222 square kilometres of the GIB WLS is divided up between people and their livestock and wildlife conservations areas. About 52 square km of the land consists of many small village communities and agricultural land used by the villagers for growing sugar cane, grapes and seasonal crops, and for grazing their domestic livestock. Among the human settlements there are protected areas for wildlife totalling about 32 square km. In common with many third world human communities, dogs are an ubiquitous part of life and the concept of ‘owning a dog’ is very different to ours. In the GIB WLS, dogs are free-roaming and puppies have never been socialised and habituated to human contact as we understand it. There is also no such thing as pet insurance, veterinary health care, or routine vaccination.
How the research was done
The researchers began by selecting 6 villages where they estimated the local human and dog populations to be similar. They then set about identifying all the dogs associated with the villages by capturing them, marking and photographing them, then releasing them. Selected dogs were categorised by their estimated age based on dentition, body size and weight, and whether or not the dogs had descended testicles or developed teats:
- Puppies: 0 to 4 months old.
- Juveniles: 5 to 12 months old.
- Adults: > 12 months old.
It should be noted that none of the dogs selected for this study were vaccinated as part of the wider, mass vaccination program. In addition only adult dogs (> 12 months) were selected for titre testing because in the younger dogs, the existence of naturally-derived residual maternal antibodies could skew the results.
In all, 130 dogs were selected and divided into 2 randomly allocated groups – a treatment groups and a control group:-
- TREATMENT group: Were given a single vaccination of Canigen DHPPi/L and Rabigen mono rabies vaccine manufactured by Virbac Animal Health. Canigen DHPPi/L is a combination vaccine containing live CDV, CAV type 2, CPV and canine parainfluenza virus, along with inactivated whole organisms of Leptospira canicola and L. icterohaemorrhagiae.
- CONTROL group: Were given a single vaccination of Rabigen mono rabies vaccine only. This was done for the safety of the villagers and the researchers who had to capture all the dogs and take blood samples for antibody titre testing at several points during the study. Rabies is endemic in much of India.
Following vaccination of the dogs, 4 blood samples were collected where possible over the following year. The 4 samples collected were before vaccination, 6 months after vaccination, 9 months after vaccination and finally, 12 months after vaccination. Sampling was complicated by the fact that the dogs were not used to being handled and it required the consent and the co-operation of the villagers who ‘owned’ the dogs, and then their abilities to confine and hold them for sampling to safely take place. Inevitably, it was not possible to collect samples from all of the dogs on all 4 occasions.
Right at the start of the study, within the CONTROL group of dogs, 11 tested antibody-seronegative for one or more of CAV, CDV, or CPV. That is, they had no natural immunity to the disease and were therefore susceptible to it. By the end of the study, most of these dogs had converted to seropositive for that disease. This is a really important finding because it tells us that these dogs must have caught the disease, survived it and then recovered during the course of the year, and were immune by the end of the study. This is all without being vaccinated, remember.
Within the TREATMENT group of dogs, 11 also tested antibody-seronegative at the start of the study for one or more of CAV, CDV, or CPV. Out of these dogs, a year later by the end of the study, 1 of 4 was still seronegative for CAV and 1 of 6 for CDV. Again, this is important because it suggests that some dogs within this population of free-ranging dogs are non-responders. We know that in affluent countries where dogs are generally vaccinated routinely throughout their lives, there is a population of non-responders that runs at around 0.2%(Larson and Schultz R,, 2007). These are dogs in which their immune systems simply fail to recognise and respond to the challenge of the antigen introduced through vaccination.
What is really interesting here is that these non-responder dogs had not developed antibodies through natural infection either, even though the disease was endemic.
Within this study it was found that the natural exposure rate of ADULT dogs to disease, measured by their antibody titres, was really high at CAV: 76%, CDV: 83% and CPV: 95%. By contrast, the estimated cover provided through the mass vaccination program in 2011 was just 34%. The significance of these findings are obvious when they are compared with the standard target HIT of 67% in order for a vaccine program to be effective in controlling disease where the R0 is 3 or less, as discussed above.
In samples of the juvenile dogs tested, antibody titres to all the target diseases were lower than in the adults, and in the puppies tested the antibody titres were zero. This suggests, that these younger, susceptible dogs play a crucial role in the persistence and the ongoing spread of CAV, CDV and CPV in this population of dogs and that research on controlling such diseases should be targeted on these groups.
These results call into question the value of mass vaccination programs in populations of dogs where the target diseases are endemic. A more efficient and cost-effective approach would be to target either the younger individuals within a population, or the wildlife deemed at risk directly.
What do these findings mean to us?
Well, rural India where diseases like rabies, CAV, CDV and CPV are endemic within the dog population is not comparable with leafy Surrey where they are not. As discussed, the only way to avoid catching an infectious disease is to either remain in isolation from it, or to acquire an immunity to it, which might be through natural infection and recovery, or through vaccination.
This study does give us some valuable insights into the merits of mass vaccination programs in animal populations and it is also a timely reminder of the mathematics HIT and the magic number 67%. In parts of the World where diseases like rabies, CAV, CDV and CPV are not endemic, like the United Kingdom, the immunity of pet populations relies exclusively on effective vaccination programs. Those cats and dogs that are not vaccinated for one reason or another are in effect ‘free-riders’ relying on the HIT of the general population as a whole remaining high enough to protect them from disease.
We know that for pet cats, where vaccination rates have been measured, they already fall well below 67% (Horzinek and Thiry, 2009). As far as dogs are concerned in the UK, canine parvovirus remains the most common cause of viral enteritis (Bird and Tappin, 2013). CPV has evolved and there are now 3 strains – CPV-2a, CPV-2b and CPV-2c – with a grim mortality rate of 90% without the intervention of immediate, aggressive treatment and intensive care (Bird and Tappin, 2013).
And what about rabies? Well, the UK remains free of this horrible disease, but in my opinion the possibility of a serious outbreak happening here is a question of WHEN, rather than IF. Interestingly, the prestigious journal of the British Medical Association, the BMJ, has recently published a very informative article for Doctors (Crowcroft and Thampi, 2015) which is available for free online.
|How (R0 − 1) ÷ R0 is derived
Using the case of Charlie the dog and virus X as an example, where R0 = 4.
The herd immunity threshold (HIT) is achieved for the population of dogs in London when of 4 new dog contacts made by each infected dog, 1 of those dogs catches the infection – that is, 1/4, or 0.25. Therefore R0 x 0.25 = 1. The 0.25 represents the proportion of dogs in the population that are susceptible. This also tells us that to achieve HIT, the proportion of immune, or resistant dogs in the population must be 0.75. We can now re-write the equation using the immune dogs like this:
4 x (1 – 0.75) = 1
Then, with a bit of algebra we can re-organise the variables like this:
1 – 0.75 = 1/4
0.75 = 1 – 1/4
Or, substituting the variables back into the equation
HIT = 1 – 1/R0
This can also be expressed like this, which is more usual in the literature
HIT = (R0 − 1)/R0
© copyright Robert Falconer-Taylor, 2015
This article is an original work and is subject to copyright. You may create a link to this article on another website or in a document back to this web page. You may not copy this article in whole or in part onto another web page or document without permission of the author. Email enquiries to firstname.lastname@example.org.
ILLUSTRATIONS: Herd Immunity, copyright COAPE.
Belsare AV, Gompper ME. 2013. Assessing demographic and epidemiologic parameters of rural dog populations in India during mass vaccination campaigns. Prev Vet Med. 2013 Aug 1;111(1-2):139-46.
Belsare AV, Gompper ME. 2015. To vaccinate or not to vaccinate: lessons learned from an experimental mass vaccination of free-ranging dog populations. Animal Conservation, Animal Conservation 18 (2015) 219–227.
Bird L, Tappin S. 2013. Canine parvovirus: where are we in the 21st Century?. Companion Animal, 18(4), 142-146.
Crowcroft NS, Thampi N. 2015. The prevention and management of rabies. bmj, 350, g7827.
Fine P, Eames K, Heymann DL. 2010. “Herd immunity”: a rough guide. Clin Infect Dis. 2011 Apr 1;52(7):911-6.
Garde E, Pérez G, Acosta-Jamett G, Bronsvoort BM. 2013. Characteristics of a canine distemper virus outbreak in Dichato, Chile following the February 2010 earthquake. Animals, 3(3), 843-854.
Hampson K, Dushoff J, Cleaveland S, Haydon DT, Kaare M, Packer C, Dobson A. 2009. Transmission dynamics and prospects for the elimination of canine rabies.
Horzinek MC, Thiry E. 2009. Vaccines and vaccination: the principles and the polemics. J Feline Med Surg. 2009 Jul;11(7):530-7.
Keeling MJ, Rohani P. 2008. Modeling infectious diseases in humans and animals. Princeton University Press. ISBN: 0-69111617-2.
Larson L. Schultz R. 2007. Three-year serologic immunity against canine parvovirus type 2 and canine adenovirus type 2 in dogs vaccinated with a canine combination vaccine. Vet. Ther. 8, 305–310.
Nouvellet P, Donnelly CA, De Nardi M, Rhodes CJ, De Benedictis P, Citterio C, Obber F, Lorenzetto M, Pozza MD, Cauchemez S, Cattoli G. 2013. Rabies and canine distemper virus epidemics in the red fox population of northern Italy (2006-2010). PLoS One. 2013 Apr 22;8(4):e61588.
Rodpothong P, Auewarakul P. 2012. Viral evolution and transmission effectiveness. World J Virol. 2012 Oct 12;1(5):131-4.