Although rabies is a reportable disease, the true incidence of the disease is greatly underestimated due to a lack of systematic surveillance and reporting by most developing countries [1]. In news published in September of 2011, the Global Alliance for Rabies Control estimates that 70,000 people die of rabies every year (http://rabiesalliance.org/media/news/annual-number-of-deaths-from-rabies-hits-70000-worldwide), mostly in developing countries from Asia (58%) and Africa (42%) [2]. Rabies is a vaccine-preventable disease. Since dogs are the main reservoir and source of infection for humans, a better control of rabies could be achieved through vaccination of dogs [3]. In Latin America, the mass vaccination of the dog population has led to a significant drop in the number of human and canine rabies cases [4].

Despite several mass campaigns of parenteral vaccination of dogs, canine rabies remains a serious health problem in Morocco and about 22 human deaths are reported every year [5]. Parenteral vaccination is the method of choice for owned dogs (i.e., dogs with a person that claims responsibility, according to the World Organisation for Animal Health (OIE) definition [6]), since approximately 70% of dogs in Morocco are considered accessible [7]. However, a proportion of dogs are stray (i.e., free-roaming with or without owner, or feral, according to the OIE definition [6]), and owned dogs may also be aggressive and difficult to handle, therefore oral vaccination could be helpful [8]. Rabies has been successfully controlled in several European countries through oral vaccination of wildlife with vaccine baits and considerable success has been achieved in the control of Arctic fox rabies [9] and raccoon rabies in Canada and in the United States [10]. International guidelines for rabies control in dogs and implementation of field trials using oral vaccines are available [11,12].

Moroccan authorities opted for evaluating oral rabies vaccine strain SAG2 in field conditions. The present study is a first trial undertaken in the region of Rabat (Morocco) to evaluate the acceptance of baits specifically designed for dogs.

Only two vaccines strains (SAG2 and V-RG) are currently recommended by WHO for oral vaccination of dogs [12]. SAG2 has been widely used in Europe and led to wildlife rabies elimination in several European countries [23]. It should be noted that no vaccine-induced rabies cases were reported in Europe after the distribution in the environment of more than 20 millions of SAG2 baits. This vaccine is registered for the control of canine rabies in India [16] and has been mainly evaluated in Tunisia [22], in Mexico [24], in South Africa [25] and in Indonesia [26], demonstrating its efficacy for dog vaccination in the field [27].

Numerous and extensive pathogenicity studies undertaken on more than 30 target and non-target warm-blooded species, including 5 from the canine family (of which dog), wild carnivores, rodents and non-human primates, demonstrated that SAG2 is an avirulent rabies vaccine strain [28]. The studies were performed according to the guidelines of WHO for safety testing of modified live-rabies vaccines [29] in species including rodents (mouse, rat, vole, squirrel, gerbil, jerboa, and meriones), carnivores (coyote, ferret, civet, mongoose, badger, and genet), non-human primates (Chacma baboons), other mammals (hedgehog, wild boar, domestic goat, and cow), and diurnal and nocturnal birds (crow, rook, buzzard, kite, and owl) [18,19,20,22,30]. Different studies showed that the persistence of the SAG2 virus in the oral cavity is very limited or even nil in non target species [14,19,20,21,22].

The immunogenicity and efficacy of SAG2 were evaluated in laboratory (mouse model) and in field conditions. Different trials demonstrated the protective effect of SAG2 vaccine bait in controlled laboratory conditions in different wildlife species and in dogs. These studies suggest that SAG2 induce humoral [15,16,17,31,32,33,34] and cell mediated immunity [31].

SAG2 vaccine baits appeared well-adapted to both owned and stray dogs living in Morocco. The shape of the bait is round (diameter=44 mm), facilitating the uptake by dogs of different sizes. The quantity of bait matrix was reduced to facilitate the puncture of the capsule to favour the contact of the liquid vaccine with oral mucosa and tonsils. It should be noted that capsules which were not pierced did not stick enough to the bait. It was observed that when the capsule was pierced before detaching from the bait, dogs tended to take again the capsule and lick the sweet content of the capsule. The baits have "cement/dirt" colour so that it merged with the colour of the soil and it does not attract children's attention. The chicken liver flavour has an attractive smell and a high palatability for Moroccan dogs. Several studies have shown that there were regional preferences for the bait flavours, such as poultry in Guatemala [35], dog biscuits in Mexico [24], chicken heads in Egypt [36] or fish in US Indian reservations (Navajo Nation and the Hopi Nation lands) [37].

Three different methods of distribution (central point, DDDM, and WIM) were previously tested in semi-rural areas in Tunisia with placebo baits containing sulfadimethoxine as a systemic marker [27]. When baits were given to dog owners at mobile vaccination centres with precise instructions for use, 79% of baits were fully consumed by the dogs [27]. The uptake reached 59% in owned dogs after door-to-door baiting and 42% in owned and ownerless dogs after transect line baiting [38]. In another study conducted in South Africa, the DDDM was tested. In total, 77% of the vaccine baits were at least partly consumed [39].

These methods of distribution were tested in this study. The DDDM, which reached primarily owned dogs and the hand-out model, which reached accessible stray dogs [11], were intended to evaluate bait palatability. The WIM adopted to distribute baits to stray dogs around the slaughterhouse of Tiflet gave encouraging results. More than 70% of the baits distributed along lines 2, 3, and 4 disappeared overnight despite the presence of abundant food and familiar smells in the environment. We assume that baits had been consumed by stray dogs, since 68% of capsules were found pierced with different characteristic canine teeth marks from dogs and because there were no feral animals in this area close to human habitations. No bait distributed along line 1 was consumed, probably because dogs could not hide in this fallow land. When this line 1 was identified on the previous year [13], the area was covered with bushes where stray dogs used to stay. This result illustrates that baits should not be distributed at random. Baits could be placed at selected sites where large numbers of stray dogs congregate such as slaughterhouses, garbage dump or public markets (souks) and along ways frequently used by dogs. Although the WIM is not specific and less safe than the DDDM and hand-out method, it allows the access to free-roaming and feral dogs, which constitute a high risk group in terms of rabies transmission [11]. The WIM requires the use of safe vaccines, fulfilling all criteria listed by WHO [11,12].

This pilot study showed that oral vaccination of both owned and stray dogs is feasible with baits specifically developed for dogs and adapted modes of distribution. These data will be confirmed with a larger field trial taking into account results of this study. The necessary 60-70% vaccination coverage could be achieved using vaccine baits to target stray dogs in addition to injectable vaccines used during mass parenteral vaccination programmes. Methods of distribution should be adapted to local conditions. For instance, in Asia, the "hand-out" model can be used because free-roaming dogs accept easily baits even when given by unknown people, whereas in Maghreb dogs wary of strangers because usually people prevent contacts with them by throwing stones in their direction, people worrying of dog bites and diseases (personal observations). Although the cost of oral vaccination is higher than that of the parenteral vaccination, the combination of both vaccination strategies may increase the vaccination coverage in the vector population and lead to rabies elimination. The cost of vaccination could be reduced by the use of aerial distribution in certain well defined and restricted areas.

Research on effective field methods for control and elimination of rabies, although less attractive than sophisticated molecular analyses, should remain the priority [40].