In general, anticoagulant rodenticides have had a good reputation for safety. This reputation is based on their widescale use by amateurs and professionals with relatively few serious incidents of exposure to non-target species, despite numerous exposure incidents. Human poisoning records indicate that anticoagulant poisonings are substantially less than poisonings from medicines, alcohol and other household chemicals. Regarding animals, in the first three years (September 1978 to August 1981) of HOTLINE calls to the Animal Poison Control Center at the University of Illinois Urbana, 4.4% of total calls related to anticoagulants. In 1982, anticoagulants accounted for 8% of HOTLINE calls and ranked fourth in concern, behind insecticides, toxic vegetation and certain household products [14, 15]. In 1983, the number of calls for all poisonings had increased, as did the percentage of anticoagulant-related calls, which were more than 10% . For the year July 1982 to June 1983, about 0.8% of all calls to LAMARPIC (Los Angeles Medical Association Regional Poison Information Center) related to anticoagulant exposures of all species . This represented about 8% of all their pesticide calls; 41% of all anticoagulant calls involved dogs, a fact also found in other countries .
Considering that more than 25 million pounds of anticoagulant bait are estimated to be used each year in the United States, the safety record is impressive but hardly surprising. After all, such baits contain low concentrations of toxicant and their slower mode of toxic action and the availability of an antidote make death of non-target domestic animals unlikely, particularly when veterinary intervention is available. A survey of 483 dogs treated by veterinarians for warfarin poisoning in England showed that the majority (81%) recovered, although the number that succumbed was significant and the costs incurred for veterinary care were considerable . Similar results were noted in a survey of United States veterinary institutions: 35 dogs (22%) died of the 158 poisoned with warfarin (or associated anticoagulants generically termed as such), where the outcome was known . Fortunately, permanent effects from sublethal intoxication with anticoagulants are rare.
The past good safety record of anticoagulants is no reason for complacency. Recent events indicate that more care in their use by both professional and non-professional applicators is essential because a wider variety of anticoagulant rodenticides is now available, some of which are widely used and differ markedly from warfarin in toxicity and effects on rodents and non-target species [28, 43].
The anticoagulants first marketed in the 1950’s could be described as multiple-dose or multiple-feeding anticoagulants. Warfarin, pindone and isovaleryl indandione are examples of such first-generation anticoagulants. These products, as formulated into baits, are only moderately toxic to rodents and most non-target species, and normally achieve their lethal effects only when repeated feedings over several days produce an accumulation of the compound within the body. A single feeding by a rodent or non-target animal is usually sublethal. The challenge is to place these baits where they will be frequently consumed by rodents and not by non-target species.
Two baits introduced later in the 1950’s and 1960’s utilized more potent multi-feed anticoagulants: diphacinone (trade names include Ditrac®, Kaput®-D, Ramik®, TomCat®) and chlorophacinone (trade names include Borderline™, Rozol®).
Since the mid 1970’s, we have seen the introduction of second-generation, single-feed anticoagulants, which are based on three toxicants which are many times more acutely toxic to rodents than warfarin [9, 22, 34]. These are, brodifacoum (trade names include Final®, Havoc®, Jaguar®, Weatherblok®XT and Talon®) bromadiolone (trade names include BootHill®, Hawk®, Just One Bite®, Maki®, Resolv®, Revolver™) and difethialone (trade names include BlueMax™, FastDraw®, FirstStrike®, Generation®, and Hombre™). Even low concentration (0.005%) baits based on brodifacoum and bromadiolone toxicants and even lower concentration (0.0025%) baits with difethialone are capable of producing rodent kill after a single feeding; hence they are commonly referred to as single-feeding anticoagulants (although in practice rodents feed repeatedly and can accumulate much more than a lethal dose).
These three toxicants and diphacinone, mentioned above, are much more acutely toxic to non-target species like dogs and cats than the older anticoagulants such as warfarin. Of these, brodifacoum has appeared to be the most toxic to dogs and swine [5, 20, 31]. Indeed, in 1984, HOTLINE calls to the Animal Poison Control Center showed that the number of rodenticide-related calls had risen to first place, with 17% of total calls, ahead of calls related to insecticides and toxic vegetation. More than 92% of these rodenticide-related calls were due to anticoagulants and, of those calls where toxicosis or suspected toxicosis was assessed, 57% were due to brodifacoum . These estimates may be biased because only a few rodenticide product labels include the HOTLINE number. Tables 1 and 2 compare the acute oral LD50 (where known) of first- and second-generation anticoagulants for dogs and cats.
In practical terms, these differences in acute oral LD50 potentially mean that, in the case of the most toxic products, a single bait station or consumer packet contains enough product (a few to several ounces) to kill an otherwise healthy 22-pound dog which consumes the entire contents at one time. In contrast, the same dog may need to eat the contents of 15 or more bait stations or consumer packets containing more than 35 ounces of 0.05% warfarin bait before consuming a lethal dose. However, the differences between anticoagulants go far beyond differences in acute oral LD50 values. Some of the newer anticoagulants have longer or much longer biological halflives than warfarin and may remain in the body at a toxic level for many months.  The prolonged turnover may reflect differences in metabolic rates, tissue and blood release of compounds, binding to blood or other cells and plasma proteins, and genetic susceptibility or resistance. Compounds other than warfarin have a longer residue half-life in tissues . The residue half-life is clearly of importance both from the viewpoint of treating poisoned animals and in the potential for secondary poisoning when companion animals or wildlife consume poisoned rodents . A long biological half-life also increases the possibility of primary intoxication in non-target species such as dogs, which may repeatedly consume sublethal doses with an additive lethal outcome.
Considering these differences among anticoagulants, it is unfortunate that both amateur and professional users of rodenticides often use (and misuse) all anticoagulants as though they were as safe as first generation rodenticides such as warfarin. The result is an increasing number of severe or fatal poisoning incidents involving non-warfarin toxicants. The problem is exacerbated when, in the absence of information to the contrary, veterinarians treat the animals for generically assigned warfarin poisoning when, in the case of more toxic anticoagulants, the animal may require much more extensive antidotal therapy and supportive treatment [11, 17, 37, 38]. For instance, in many cases involving brodifacoum poisoning of dogs, the animals died after being sent home following veterinarian examination and treatment for anticoagulant poisoning. The majority of these animals could have been saved by extending antidotal therapy.
The case histories are representative of the range of dog poisoning incidents involving anticoagulants now being encountered and thus may be of use to veterinarians when designing treatment programs. While the focus in this guide is on anticoagulant poisoning, it is important that veterinarians understand that acquired or inherited hemostatic defects (e.g. disseminated intravascular coagulation, liver disease, quantitative and qualitative platelet defects, von Willebrand's disease, and the hemophilias) may produce symptoms that can be confused or concomitant with anticoagulant poisoning. The various coagulation tests and their limitations should also be borne in mind when making differential diagnoses [18, 24]. Dog poisoning case histories have been chosen because these represent a substantial majority of the companion animal poisoning incidents which are reported [14, 20, 33, 39, 41, 43, 51]. However, poisoning of cats, birds, horses and other animals are also reported and their treatment would similarly vary according to the type of anticoagulant to which they had been exposed.