TRANSACTIONS OF THE ROYAL SOCIETY OF TROPICAL MEDICINE AND HYGIENE (1999) 93,177-179 The role of thromboelastography in the management of children with snake-bite in southern Africa G. P. Hadley’, P. McGarr’ and M. Mars2 Durban, South Aji-ica Departments of ’ Paediatric Surge y and ‘Physiology, University of Natal, Abstract In the absence of a direct laboratory test of envenomation, there is a need for an alternative mechanism for the early recognition of envenomation following snake-bite in children. A severe clinical diathesis may result either from envenomation or from the release of an inappropriate tourniquet applied as ‘first-aid’ often several hours before presentation to hospital. Abnormalities of clotting are associated with both events. A normal thromboelastogram (TEG) provides early recognition of patients in whom the clinical course is likely to be benign (sensitivity = 94%). An abnormal TEG identifies patients of whom 50% will develop a severe clinical diathesis. A TEG is a more accurate predictor of disease severity than International Normalized Ratio alone. The TEG does not supplant clinical observation in the management of snake-bite in children but allows stratification into high- and low-risk categories. Keywords: snake-bite, envenomation, children, prognosis, thromboelastography, South Africa Introduction Snake-bite in children is common throughout the tropics and subtropics. The small mass of a child reduces the volume of distribution, and therefore increases the likelihood of clinically significant toxicity of a given dose ofvenom. All snake-bites cause alarm but it is sometimes forgotten that not all snakes are venomous and that not all attacks by venomous snakes result in envenomation of the victim (TIBBALS, 1992). The situation is com- pounded for the clinician by the fact that the exact circumstances of injury are often unclear and evidence of fang marks may be obscured by scarification per- formed by a traditional healer, or by oedema. Further, tourniquet application is still regarded as appropriate first-aid in many communities and identification of the snake is rare (h4~~s et al., 199 1; YERZINGATSIAN, 1997). This situation pertains throughout much of sub-Saharan Africa, yet the potential toxicity, scarcity and cost of antivenoms make it desirable that envenomation be precisely diagnosed. Snake populations differ in different parts of the world and therefore patterns of envenomation will also differ. Tests of envenomation and venom detection kits based upon the species of snake prevalent in one area may not be relevant to any other area. No specific tests for envenomation by the snakes prevalent in southern Africa are currently available. Thus direct laboratory contirm- ation of envenomation is not possible and reliance must be placed on clinical observation and indirect indices such as clotting dysfunction. Snake venoms are traditionally divided into cytotoxic, haemotoxic and neurotoxic groups depending upon their principal action. However, venoms are complex mixtures of active agents and few, if any, will have a single ‘pure’ effect (ISEMONGER, 1983). In addition, snake-bite pro- vokes a tissue response with inter alia the release of vasoactive cytokines. Antivenom may restore some ele- ments of the clotting process whilst having no effect on others, e.g., thrombocytopenia, suggesting that the latter may not be a direct venom effect (BOND & BURKHART, 1997) but a secondary response. Based on these observa- tions we have postulated that any significant envenom- ation will have a measurable effect on the clotting process which, even though the effect is subdominant, will be detectable by a dynamic assessment such as a thrombo- elastogram (TEG). This postulate is not accepted by all opinion and some maintain that certain venoms, parti- cularly neurotoxic venoms, have a single pure action. Resolution of the controversy must await the assessment Address for correspondence: Professor G. P. Hadley, Depart- ment of Paediatric Surgery, University of Natal, Private Bag 7, Congella 4013, Natal, South Africa; phone +27 312604227, fax +27 31 2604455, e-mail hadley@med.und.ac.za of coagulopathy using TEG in a series of patients following neurotoxic envenomation. From a practical point of view an Australian series showed that all envenomated children demonstrated coagulopathy when assessed by prothrombin index (PI), partial thromboplastin time (FIT), fibrinogen levels and fibrinogen degradation products (FDP), in- dependent of the species of snake (TIBBALS, 1992). No child in whom envenomation was not confirmed showed evidence of coagulopathy. The study concluded that coagulopathy is a highly sensitive, specific and reliable indicator of envenomation (TIBBALS, 1992) indepen- dent of the species of snake. It has also been shown that the ischaemic reperfusion injury associated with prolonged tourniquet use results in disturbances of coagulation (PETAJA et al., 1987). We have therefore looked at the TEG, as an indicator of envenomation and as a predictor of a severe disease diathesis. The TEG is a dynamic mechanical study of the coagulation process from fibrin formation through plate- let aggregation to fibrinolysis. Various relevant para- meters including the time to initiation of clot formation (r-time), the rate of clot formation (k-time) and the strength of the resultant clot (ma or maximal amplitude) are measurable from the TEG tracing (Figure). The TEG has been extensively used in the management of clotting dysfunction in cardiac and transplantation sur- Rerv (KANG et al., 1985.1989), and has been shown to be a sensitive indicator of early sepsis in the neonate (GRANT & HADLEY, 1997). Both hyper- and hypocoa- gulability are significant and the precise TEG abnorm- 4 +y ! a . .._- .._.” ..-.. . . 20 mm ma .,,...........,.,.. ii T - r k Figure. Normal thromboelastogram showing the time to initia- tion of clot formation (r-time), the rate of clot formation (k- time) and the strength of the resultant clot (ma or maximal amplitude).