If a tourniquet has been applied, it should not be removed until antivenom is at hand and the means for treating complications, including possible complications caused by the antivenom, are available.

If these requirements cannot be met within a suitable period of time, it is necessary to weigh the risk of the systemic effects of the venom that could occur after removing the tourniquet against the risk of progressive local tissue damage that may arise if the bandage is left in place.

The symptoms and degree of envenoming depend not only on the amount of venom injected and numerous other variables, but also on the time that has elapsed since the bite. This variable factor must be taken into account when making the following decisions:

• exclusion of envenoming (see below),
• the time interval between clinical examinations (see Therapy phase Hospital),
• emergency care (see below).

The fact that a patient has been bitten by a known venomous snake and the presence of bite marks do not automatically allow the conclusion that a clinically relevant injection of venom has taken place.

It is estimated that in 75% of North American crotalid bites no clinically relevant envenoming occurs (Minton 1987a); in contrast, in only 40–75% of Micrurus fulvius bites does envenoming not take place (Russell 1983, Kitchens and van Mierop 1987).

Following North American crotalid bites, strong local pain, swelling, discolouration of the skin and occasionally hyposensitivity in the region of the bite are early signs of a relevant injection of venom. If these do not occur within 15–30 min after the bite, it is highly likely that no venom was injected. Important exceptions to this rule are some populations of Crotalus scutulatus ("type A"), C. mitchelli, C. lepidus and C. tigris, which can cause systemic envenoming without significant local effects (Minton 1987a). Local symptoms may also be relatively insignificant following bites by C. adamanteus, even if systemic effects, including incoagulability of the blood, are present (Kitchens and Van Mierop 1987). The same applies to all bites in which direct intravenous injection of venom occurs (Minton 1987a). Within an hour after an effective crotalid bite the regional lymph nodes are generally swollen and painful. Nausea, vomiting and diarrhoea develop. Paraesthesias of the extremities and perioral as well as muscle fasciculations are characteristic for rattlesnake bites (Minton 1987a, Russell 1983) (see Biomedical database entry for C. horridus). Haemorrhagic blisters usually develop within less than 2 h after a rattlesnake bite, but can also occur after a long delay (Minton 1987a).

Progressive arterial hypotension, leading to loss of consciousness and cardiovascular failure, is one of the most significant complications of North American rattlesnake bites and is a certain sign of severe envenoming (Minton 1987a).

Apart from ecchymoses in the region of the swelling, bleeding, in particular life-threatening bleeding, is very uncommon with North American crotalid bites. This is so despite the fact that marked defibrinogenation and thrombopaenia can occur and although the venoms contain haemorrhagins. However, the haemorrhagins appear to be primarily locally active. In particular following bites by crotalids, whose venoms possess only direct fibrinogen-coagulating activity ("thrombin-like" activity), the haemostatic defect usually only manifests itself in the form of abnormal clotting test results and not clinically (Kitchens 1991, pers. comm.). For this reason coagulation should always be monitored if a crotalid bite has taken place or is suspected.

North American elapid bites (Micrurus fulvius) cause only minor local signs of envenoming. Bite marks may be so small as to be invisible. These bites are not particularly painful, there is no discolouration of the skin, and swelling, if it occurs at all, is insignificant. Moreover, systemic signs of envenoming, such as progressive paralysis that may ultimately lead to respiratory insufficiency and respiratory failure, may only appear several hours after the bite (Minton 1987a, Micrurus fulvius).

The absence of signs of envenoming in the first hours after the bite does not exclude the possibility that a relevant injection of venom has taken place. There may be a long delay before systemic signs of envenoming develop.

Following crotalid bites, whose venom is known to cause defibrinogenation, there may be a delay of several hours before this condition is detectable on laboratory tests. Also, the absence of local signs of envenoming does not exclude the later development of a haemostatic defect (Van Mierop and Kitchens 1980, Kitchens and Van Mierop 1983). The same is true for other systemic signs of envenoming, including arterial hypotension. Several patients developed cardiovascular failure following progressive arterial hypotension that first occurred hours after the bite (Hurlburt et al. 1988).

In a group of 45 patients, initially none or only one had minimal swelling and no other local or systemic signs of envenoming such as nausea, vomiting, perioral paraesthesia, alteration of consciousness, ecchymosis, blistering or marked pain. 24 (53%) later developed significant signs of envenoming such as moderate to marked swelling, thrombocytopaenia or cardiovascular failure. Of these 24 patients, 11 (46%) deteriorated within 4 h, 5 (21%) within 4–8 h and 6 (25%) within >8 h (no data for 2 patients) (Hurlburt et al. 1988).

Following Micrurus sp. bites the problem of assessing the risk of a clinically relevant envenoming is even greater. Minimal or no local signs of envenoming are the rule, and systemic signs of envenoming only develop after a long delay.

Symptomatic emergency medical treatment and antivenom treatment are complementary strategies.

The aim of antivenom treatment is neutralisation of the venom.

The success of antivenom treatment depends on the quality of the antivenom, the specific properties of those venom components relevant to envenoming and the time point at which antivenom is administered.

• Treatment of the haemorrhagic shock,
• antivenom.

Symptomatic emergency medical treatment and antivenom treatment are complementary strategies.

The aim of symptomatic emergency medical treatment is the rapid correction of critical parameters (fluid balance, blood pressure, oxygenation etc.) and the maintenance of vital functions (respiratory, cardiovascular).

Symptomatic measures help bridge the gap until specific treatment (antivenom) can be administered and starts being effective. If no antivenom is available or if the required effect is not achieved with antivenom, the goal is to employ symptomatic measures until such time as the venom naturally starts losing its activity. 
It is essential to correct arterial hypotension as quickly as possible. Following bites from North American crotalids, arterial hypotension is almost always the consequence of hypovolaemia and hypoproteinaemia caused by fluid sequestration in the bitten and subsequently swollen extremity. Large volumes of intravascular fluid replacement may be necessary (Hardy 1991). From animal experiments it appears that albumin is superior to colloidal and crystalline solutions for this purpose (Schaeffer et al. 1978). As haemostatic defects may develop at the same time, colloidal solutions are problematic, as they increase the risk of bleeding. If the hypvolaemic/hypotensive state is prevented or corrected in time, severe complications of North American crotalid bites, such as acute renal failure, non-cardiogenic pulmonary oedema, myocardial infarction and cerebral seizures, can also be prevented. These complications represented the most common causes of death following crotalid bites in Arizona between 1969 and 1984 (Hardy 1986).
Persistent signs of paralysis, including respiratory paralysis, appear to be preventable following Micrurus fulvius bites only if antivenom is administered sufficiently early. The prognosis of patients in whom respiratory paralysis is already established depends on timely endotracheal intubation and the success of artificial respiration.
Treatment with acetylcholinesterase inhibitors has been suggested but has not yet been investigated in controlled studies (Norris und Dart 1989).

The most common bacterial species isolated from rattlesnake venom contaminated with the snake’s oral flora are Pseudomonas aeruginosa, Proteus species and coagulase-negative staphylococci. Clostridium species and Bacteroides fragilis have also been found (Goldstein et al. 1979). Surprisingly, the incidence of bacterial infections following crotalid bites is low. One reason for this could be the antibacterial activity of crotalid venoms (Talan et al. 1991).