Canine Rabies Outbreaks, Vaccination Coverage, and Transmission in Humans: Greater Accra Region, Ghana- A Retrospective Study-2006-2011

Rabies is a highly fatal, viral, zoonotic disease caused by a Lyssa virus. It is acquired through the bite of an infected animal, mostly dogs. Estimated annual global human mortalities from rabies is 61,000, over 99% of which are from developing countries where about US$583.5 million is spent on its control. Rabies is enzootic in Ghana. Vaccinating about 70% of the dog population leads to a reduction in rabies transmission to humans; however, percentage dog anti-rabies vaccination coverage in Ghana is unknown. This study therefore investigated the magnitude of rabies in dogs and humans in the Greater-Accra region, and assessed annual percentage dog anti-rabies vaccination coverage from 2007-2011. Secondary data on rabies vaccination, post mortem and dog quarantine records for all ten districts in the region, and human rabies records were analysed. Means and percentages were calculated, graphs drawn and trends analysed. The number of animal samples which tested positive out of 309, was 283. Predictive value positive was 91.6%. Of the positive cases, 97.5% (276/283) had no previous vaccination history, and 96.1% (272/283) were from dogs. Fifty-five out of 174 (31.6%) samples were from dogs which bit more than one person. Human exposures to dog and other animal bites are not differentiated. Average annual number of outbreaks was 31, whereas percentage dog vaccination ranged from 10.26-17.56. Current annual percentage dog anti-rabies vaccination coverage is very low, whereas the incidence of dog rabies in the region is high. The number of humans affected may be underestimated. Government should immediately facilitate annual mass vaccination of pets.


Introduction
Rabies is a highly fatal, viral disease with an insidious onset, caused by a Lyssa virus of the family Rhadoviridae. It is a multi-species, zoonotic disease which affects mainly warm-blooded animals [1][2][3]. Infection usually follows the bite of a rabid animal [1,3].The virus is neurotropic, travelling by peripheral nerves through the central nervous system, and finally ending up in the brain where it causes encephalopathy [1][2][3][4]. Rabies occurs in two forms; furious and dumb rabies [1,[4][5][6]. The former is easier to diagnose clinically based on symptoms. However, in dumb rabies, due to the absence of characteristic aggressive symptoms often associated with the disease, diagnosis may be missed, [1,[4][5][6], making the disease even more dangerous. Moreover, once symptoms of central nervous system involvement manifest, rabies is nearly always 100% fatal [1,4,7]. It has been established that over 90% of human rabies exposures are from dogs, and of these, 99% are from free-roaming dogs. Additionally, 99% human rabies deaths are from rabid dogs [7]. All these facts therefore, suggest that control of rabies in humans is hinged on its control in dogs. This has been adequately proven by examples from a number of countries which have eliminated canine rabies through control of the disease in dogs [10,11]. In yet some other countries, significant strides have also been made in canine rabies reduction through vaccination, with a concomitant reduction in human rabies. Despite all these available evidence, rabies caused an estimated 61,000 human deaths globally in the year 2010, with over 95% of these occurring in Asia and Africa alone [11,13]. A more recent study put the annual global estimates of human rabies deaths at 59,000. In Ghana, 25 cases of human rabies were reported from 2009-2011. From an economic perspective, between US$6-8.6 billion is estimated to be spent on rabies control globally each year [11,14], whereas other studies suggest that about US$124.2 billion could be saved annually with canine rabies control through dog vaccination. In Africa and Asia, an estimated US$583.5 million is spent on rabies control annually. This is in sharp contrast to developed countries such as the United States, where an estimated $300 million is spent annually on rabies control [7,8]. In the later, though, the greater part of this expenditure is concentrated on dog vaccinations [7,8]. Furthermore, high stray dog populations in developing countries means that a single stray, rabid dog becomes a source of infection for several others as they mingle freely with each other [17,18]. This poses high risks of rabies infection to humans if such uncontrolled gatherings of dogs are unvaccinated or have low percentage vaccination coverage. As in any infectious disease, adequate herd effect is required to break the cycle of transmission [20,21]. Different studies have different postulates on the percentage vaccination coverage (PVC) which would be adequate to achieve this with respect to rabies. A study has shown that, when a percentage vaccination coverage (PVC) of 70 of the dog population was achieved in two successive years, followed by at least a PVC of 50 for 5 consecutive years, it was possible to break the canine cycle of rabies transmission to humans [22]. The World Health Organization (WHO), also recommends a PVC of at least 70 in rabies endemic areas. Another study, however, suggests that canine PVC of 39-57 is adequate to break the cycle of transmission. Finally, global health partners such as the WHO, World Animal Health Organization (OIE), the Global Alliance for Rabies Control (GARC) all agree that elimination of dogmediated human rabies is possible, since all the tools needed for this are available. Clearly then, human rabies elimination is inextricably linked to elimination of the disease in dogs. Canine and human rabies have thus been targeted for elimination by the year 2030, and Ghana, being a part of the international community, must work towards this. With this background, an analysis of rabies data in the Greater-Accra Region (GAR) of Ghana was carried out to determine the magnitude of animal rabies from 2007-2011, to establish the trend of rabies outbreaks in animals and humans, and to assess the percentage canine anti rabies vaccination coverage for the same period.

Study Area and Study Population
The study area was the GAR of Ghana, which is located in the south-east of the country along the Gulf of Guinea, between latitude 5 0 33 ′ 0″ North, 0 o 13′ 0″West. It occupies a land surface of 3,245 square kilometers, and an estimated population of 4,010,054. The region is divided into ten administrative and political districts. A survey of the dog population during polio vaccination exercises carried out by the Ministry of Health in 2009 estimated the number of dogs in GAR at 82,684.

Study Type and Sources of Data
Secondary data on official monthly veterinary records on rabies incidence in GAR for the period 2007-2011, were reviewed. These comprised data on dogs quarantined for suspected rabies, rabies tests carried out at the Accra Veterinary Laboratory, and monthly reports from the ten districts in the GAR. Human rabies morbidity and mortality data from 2008-2011 from the Public Health Division of the Ghana Health Services, GAR, was also reviewed.

Data Analysis
Data was analysed and the results expressed as means, frequencies, and percentages, and the trend in animal vaccinations over the period determined. The number of human involvement with suspected rabid dogs were also analyzed.

Ethical Approval
Permission was sought and obtained from the Director of Veterinary Services and the School of Public Health, University of Ghana, Legon, prior to the study.

Number of Suspected Samples Submitted for Rabies Testing and Animal Species Involved
A total of 319 samples from different animal species were submitted to the Accra Veterinary Laboratory on suspicion of rabies from 2007-2011. The highest number of samples submitted was from dogs; 95.6% (305/319), contributions from other species were cats and horses, 2.5% (8/319), 0.9% (3/319) respectively. The rest were 1% for goat, pig and rabbit (1/319) each. Out of the 319 samples submitted, 3.1% (10/319) were decomposed so were not tested. Out of the number of samples tested for rabies, 283/309 were positive; predictive value positive (PVP) was 91.6%. The proportion of dog samples out of the total number of positives was 96.1% (272/283). During the period of review, a total of 1,146 dogs were quarantined on suspicion of rabies, with the highest and lowest numbers, 182, and 254 being in 2009 and 2008 respectively. An analysis of the vaccination status of animals from which samples were submitted for rabies testing is shown in Table 1. Of the 283 samples that tested positive for rabies out of 309, 97.5% (276/283) were unvaccinated. All seven samples, which came from vaccinated dogs, however, also tested positive for rabies.

Number of Rabies Outbreaks and Percentage Vaccination Coverage in GAR, 2007-2011
From  Dog vaccinations were mostly carried out based on individual requests by owners, however, there were also mass vaccinations carried out in 6/10 districts. Mass vaccinations were not syynchronised across the region, but rather, each district carried out its own schedule of vaccinations. Additionally, they spanned several weeks, sometimes months, and were limited in scope, as shown by the PVC. All vaccinations were on owner-pay principles, with some limited support from the local authorities.

Human Involvement with Rabies Suspected Dogs for Which Samples Were Submitted for Rabies Test
Of the samples submitted to the laboratory for rabies test, 174/319 was from dogs which bit 324 humans; of these, 55/174 (31.6%) bit more than one person. Further, twenty percent (64/324) of victims were bitten by their own dogs.

Human Morbidity and Mortality Data from Rabies
From the Public Health Division of the Ghana Health Service, GAR, a total of 15,896 reported animal bite cases involving humans were recorded for the period under review. However, there was no data on what proportion of this figure was from potentially rabid dogs as all animal bite cases were lumped together. There was also, no available data on the number of humans who were given post exposure vaccination for rabies; animal bite victims had their wounds sanitized, and were given anti-tetanus injection and antibiotic cover.

Discussions
The highest annual PVC for dogs for the period was 17.56, in 2011. One of the reasons for the low PVC could possibly be due to "owner-pay policy", resulting in low patronage. This was consistent with findings from a number of studies on dog vaccinations in developing countries [9,[28][29]. In contrast, when canine vaccinations were free, patronage was higher, and consequently, higher vaccination coverage was achieved [9,28,29,30]. Interestingly, the highest number of rabies outbreaks in GAR was also recorded in the same year, 2011 when the highest PVC was achieved. This was contrary to the notion that increasing PVC resulted in an automatic decrease in rabies incidence. This can be explained by the fact that even the highest PVC of 17.5 was still inadequate to break the transmission cycle as it was clearly, far below even the minimum recommendation of 37.5. Another study also reported that even though a PVC of 50 could break the transmission cycle, it could not lead to rabies elimination in dogs. However, a PVC of 70 was sufficient to interrupt human exposure to the virus. All these studies agree that at least, a certain minimum percentage of dogs should be vaccinated in a given geographical location within a defined period in order to break the canine transmission cycle of rabies.
The rather high number of samples which tested positive to rabies, but had no previous vaccination history, as well as the high number of outbreaks are further confirmation of the finding of low PVC in the region. Additionally, the high PVP points not only to the high sensitivity of the test, but also, high prevalence of rabies in the dog population. Of particular concern is the finding that all seven samples from animals which were vaccinated within the year prior to the test were positive for rabies. This could probably be due to vaccine failure which could be attributed to several reasons, such as vaccination by quacks who use fake vaccines, or a break in the cold chain cycle. This finding is consistent with studies from both the developed and developing countries which have also reported rabies in dogs previously vaccinated against the disease [32][33][34][35].
The study shows that the majority of positive samples were from dogs. Furthermore, over, 98% of the reported human involvement with suspected rabid animals from which samples were tested, was also from dogs as found in similar studies [31,36,37]. Both of these observations support the World Health Organization's assertion that the majority of human exposures to rabies are from rabid dogs [8,10,12].
Since human exposures to dog bite cases were lumped together with other animal bite cases at the Public Health Unit, Ghana Health Service, GAR, it was not possible to glean exactly how many humans had had potential exposures to rabid dogs, nor whether they were followed up. Furthermore, it was also not clear whether the 324 dog bite victims recorded by the Accra Veterinary Laboratory were part of this figure. From literature, about 80%-90% animal bite cases involving humans are from dogs. This means out of the 15,896 animal bite cases reported at the Public Health Division of the Ghana Health Service, GAR, at least 12,717 could have been from dogs. This can be compared to studies in Asia where, out of 3,529,000 animal bite cases, 31,539 human rabies deaths were reported. Similarly in Africa, out of 802,000 animal bite cases, 23,823 human deaths were recorded [17,39]. Further, considering that rabies is endemic in Ghana and GAR for that matter, it is possible that some of the animal bite cases were from potentially rabid dogs. Hence there is the need for animal bite cases to be treated as potentially rabid cases and thoroughly investigated. In furtherance to this, all dog bite cases should be identified and reported as such by medical practitioners. In some jurisdictions of the developed nations, all animal bite cases are referred to the local veterinary authority who apprehend the offending animal, quarantine and observe it for at least 10 days. This is to ensure that the offending animal is not potentially rabid. Based on other epidemiological information, the exposed human is then given the post-exposure prophylaxis. However, it is not clear from available data how many of the dogs quarantined at the Veterinary Services Directorate, were referrals from human health hospitals, nor how many were self-reported. This would have helped to link the human victims to all animal rabies suspected cases. Limitations The dog population used was only from data captured from homes with children under five years of age during the polio immunization exercise. Additionally, stray dogs could have been missed in the count. This means that the actual dog population could be far in excess of what was used, which would have affected the PVC obtained. Further, the same population was used as the denominator for the analysis over the five year period. Since populations are not static, this could also have affected PVC obtained. Additionally, rabies cases captured reflect only those that were reported to the VSD.
We did not take into consideration other issues such as vaccination strategy, that is; central point as against door to door vaccination, or resource availability as a contributing factor to PVC.

Conclusions
The prevalence of rabies in the region is very high. On the other hand, the PVC is so low that it is not making any dent in the transmission cycle of rabies. A lot of potential human rabies cases are also not being captured by the Ghana Health Service. The Ministry of Food and Agriculture, through the Veterinary Services Directorate, should immediately facilitate mass vaccination campaign for pets in order to encourage participation and target a vaccination coverage of 70% or more. This should be continued annually for at least five consecutive years as recommended by the WHO. Additionally, the Public Health Department of the Ghana Health Service and the Veterinary Services Directorate should strengthen their collaboration with each other so as to improve rabies surveillance. This will ensure that potential human rabies cases are quickly intercepted.