Journal ofwilderness Medicine, 6, 183-188 (1995) ORIGINAL ARTICLE Bite marks are useful for the differential diagnosis of snakebite in Brazil SERGIO DE A. NISHIOKAl*, PAULO VITOR P. SILVEIRA!, and FERNANDO A. BAUAB2 lcentro de Ciencias BiomMicas, Universidade Federal de Uberltindia, 38400-902 Uber!tindia, Brazil; 2Faculdade de Medicina de Catanduva, Catanduva, Brazil Objective: To assess whether examination of the bite marks could be useful for the differential diagnosis between venomous and nonvenomous snakebite. Design:Assessment ofthe validity ofthe bite marks in the differential diagnosis between venomous and nonvenomous snakebite in a series of patients prospectively recruited in a 32-month period. Setting: Teaching hospital in a southeastern Brazilian town. Patients: Forty-twovictims ofsnakebitewho broughtthe captured or dead snake for identification and who had recognizable bite marks. Intervention: Inspection of the bite site, recording the aspect of the marks either as isolated fang punctures or as multiple scratch-like teeth marks. Results: The presence of isolated fang marks, when related to venomous snakebite, had a sensitivity of 100%, a specificity of 56%, and a predictive value of 89%. The finding of multiple, scratch-like teeth marks had a predictive value of 100% when related to nonvenomous snake bite. Conclusions: Examination of the bite marks showed to be useful for the differential diagnosis between venomous and nonvenomous snakebites in Brazil, which has implications for the option of empirical administration of antivenin to patients when the snake is not available for identification. Key words: diagnosis, snake, snakebite, venom Introduction Snakebite is a public health problem in South America and a hazard to travelers to rural and even some urban areas of this part of the world. In Brazil, an average of 20000 cases were reported yearly to the Ministry of Health from 1987 to 1990 [1], but the real figure is certainly higher than this. Most of the reported cases to the Brazilian Ministry of Health are of venomous snakebite, but several case series from hospitals show that up to 40% of the patients are bitten by nonvenomous species [2,3]. We report on the use of the examination of the bite marks for making the differential diagnosis between venomous and nonvenomous snakebite in Brazil. Patients and methods A prospective study on snakebite was carried out beginning January 1991 at the teaching hospital of the Universidade Federal de Uberlandia in southeastern Brazil. The descrip- * To whom all correspondence should be addressed at: Av. Teresina 1415, Umuarama, 38405-384 Uberliindia MG,Brazil. 1080-6032 1995 Chapman & Hall
184 Nishioka, Silveira and Bauab tion of the findings from the examination of the bite site of patients seen until August 1993 was annotated, and a comparison was made between bites of venomous (pit vipers) and nonvenomous snakes (from the families Colubridae and Boidae). Only the cases in which the snake was brought for identification were analyzed. The persons who had identifiable bite marks were classified into two groups: those with isolated fang punctures (Fig. 1) and those with multiple scratch-like teeth marks (Figs. 2 and 3). This division was made based on the premise that the former should be caused by pit vipers, which are solenoglyphs, that is, have front fangs capable of a wide range of movement, whereas the latter should be caused by nonvenomous [either aglyph (without fangs) or opisthoglyph (back-fanged)] snakes. The differences of presence and position of fangs between the snake groups are responsible for these aspects ofthe bite marks. The sensitivity, specificity, and predictive value of the bite marks for the identification of venomous and nonvenomous snakebite were calculated according to well-known formulas described elsewhere [4]. Fig. 1. Victim of Bothrops (lance-headed viper, a solenoglyph snake) bite, 24 h after the bite, showing two fang punctures, one of them covered by a blister.
Bite marks andsnakebite 185 Fig. 2. An aglyph snakebite, displaying multiple scratch-like teeth marks. The head of the respon sible snake, a specimen of Helicops modestus, is also shown. Results During the study period, 54 (58.7%) of92 victims ofsnakebit killed or captured the snake and brought it for identification. The snakes that were identified, 39 pit vipers and 14 nonvenomous snakes, are listed in Table 1. The one case of bite by a cora] snake (genus Micrurus), the only venomous snake in Brazil that is not a pit viper, was excluded from the analysis; that patient, who was described elsewhere [5], presented two clearly defined fang marks on a finger. Eight patients (five victims of nonvenomous snakes and three of pit vipers) had no recognizable sign of a bite; among the persons bitten by pit vipers, only one presented without any sign of envenomation. One patient made a deep incision at the bite site, attempting to suck the venom from the cut, and made it impossible to identifity the bite marks. For two cases, there was no appropriately recorded information on the aspect Fig. 3. An opisthoglyph (back-fanged) snakebite (of a specimen of Philodryas patagoniensis) with multiple scratch-like teeth marks plus two fang marks. (Photograph by F.A. Bauab of a patient not included in this series.)
186 Nishioka, Silveira and Bauab Table 1. Specimens of identified snakes classified as venomous and nonvenomous, and according to the presence and position of fangs in the mouth Snake Common name Family Number Presence/position offangs Venomous Bothrops altematus Urutu Viperidae 3 Solenoglyph Bothrops moojeni Jaracu«u Viperidae 21 Solenoglyph Bothrops neuwiedi Jararaca pintada Viperidae 5 Solenoglyph Bothrops sp "Jaracu«u" Viperidae 4 Solenoglyph Crotalus durissus Cascavel Viperidae 6 Solenoglyph Nonvenomous Apostolepis assimilis Coral Colubridae 1 Opisthoglyph Boa constrictor Jib6ia Boidae 1 Aglyph Eunectes murinus Sucuri Boidae 1 Aglyph Liophispoecylogims Cobra-capim Colubridae 2 Aglyph Mastigodryas bifossatus Jaracul} do brejo Colubridae 1 Aglyph Philodryas matogrossensis Cobra-cipo Colubridae 1 Opisthoglyph Philodryas olfersii Cobra verde Colubridae 1 Opisthoglyph Philodryas patagoniensis Cobra-copo Colubridae 1 Opisthoglyph Sibynomorphus mikanii Dormideira Colubridae 3 Aglyph Simophis rhinostoma Coral Colubridae 1 Aglyph Waglerophis merreni Boipeva Colubridae 1 Aglyph of the bite site. The distribution of the remaining 42 cases bitten by venomous and nonvenomous snakes, according to the bite marks, is shown in Table 2. The patients who had scratch-like multiple teeth marks were bitten by Boa constrictor, Eunectes murinus, Mastigodryas bifossatus, Philodryas matogrossensis, and P. patagoniensis. The presence of isolated fang marks related to venomous snakebite had a sensitivity of 100%, a specificity of 56%, and a positive predictive value of 89% (calculated from Table 2). The finding of several teeth marks related to nonvenomous snakebite had a sensitivity 56%, a specificity of100%, and a positive predictivevalue of100% (from Table 2). Table 2. Distribution of the cases bitten by venomous and nonvenomous snakes according to the bite marks Snake Venomous Nonvenomous Bite marks Isolated punctures 33* 4* Multiple, scratch-like o 5 Note: Nine cases with unrecognizable bite marks and two without available information were excluded (see text). 'Fourteen cases with a single mark, 18 with 2, and 1 with 3. *Two cases with a single mark and two with two marks.
Bite marks and snakebite Discussion The differential diagnosis between venomous and nonvenomous snakebite is obviously important because victims of the former can benefit from treatment with specific antivenin. Diagnosis is easier when the snake is captured and brought for identification, which occurs in 17 to 45% of the cases [2,3,6,7] in Brazilian series and when there is someone, preferably the doctor, who knows how to identify venomous snakes correctly. In South America, this is not very difficult, because more than 99% of cases of venomous snakebite are caused by pit vipers [2], which are easily identified by the presence of a pit between the nostril and the eye, bilaterally. The remaining (less than 1%) cases are caused by coral snakes (Micrurus species), whose identification is more complicated, as there are nonvenomous snakes with color rings that mimic the venomous species. In South America, it is widely accepted that the differential diagnosis between venomous and nonvenomous snakebite and also between different genera within the former can be made in most cases based only on clinical grounds, when the victim presents signs of envenomation when seen at the hospital. Nevertheless, there are situations when the differential diagnosis is difficult, particularly when the patient is seen very soon after the bite or is still asymptomatic, orwhen he uses a thigh tourniquet that delays the appearance of the signs of systemic envenomation. Although our figures are based on a relatively small number of cases-only 26% of the cases in our series were victims of bites by nonvenomous species, and among them only 64% (9 out of 14) had recognizable bite marks-our findings suggest that multiple scratch-like teeth marks are strongly indicative of nonvenomous snakebite. Conversely, when there are recognizable bite marks, the presence of isolated fang marks, although nonspecific, are highly predictive of a pit viper bite. This was so in our study because the prevalence of venomous snakes was high (79%); if the prevalence was lower, so would be the predictive value, but as stated earlier, the proportion of nonvenomous snakebite in other Brazilian series does not differ substantially from our findings [2,3]. The concomitant finding of one or two fang marks plus multiple teeth marks (Fig. 3), which was not a finding in our study, can also occur and is suggestive of an opisthoglyph bite. This information is of interest, as there are reports of envenomation by species of South American colubrids, like Philodryas olfersii, P. patagoniensis, and Clelia delia plumbea, with local edema, simulating the local findings of a lance-headed envenomation [8-11]. Several other species of colubrids in other parts of the world can cause systemic and/or local envenomation [12]. The dentary and palatine teeth of pit vipers can sometimes inflict punctures in addition to those made by fangs; therefore, the specificity of multiple scratch-like teeth marks related to nonvenomous snake bite in larger series will probably not be 100%, as in this study. Victims of coral snakes, which are proteroglyphs (i.e. have immobile front fangs), may present isolated fang marks, like in our case [5], or fang plus teeth marks, as the bite of these snakes is more likely that of a colubrid than of a pit viper. The absence offang or teeth marks at the bite site is not always evidence that there is no envenomation, as we saw in two out of three cases of pit viper bite. One has also to bear in mind that many people have scratches, marks of insect stings, and other injuries on the bitten limb, which sometimes render difficult the identification of the fang or teeth marks. Thesame also occurs in casesofincisionson the bite site, as shown in one ofourcases. Our findings demonstrate that examination of the bite site can be useful for the correct 187
188 Nishioka, Silveira and Bauab diagnosis of snakebite in Brazil and that one can predict that it can also be true wherever venomous and nonvenomous snakes coexist. In many regions of Brazil and other countries of South America, the diagnosis of snakebite is made only on clinical grounds. Careful observation of the bite site can be extremely useful, together with the search for signs and symptoms of local and systemic envenomation, to provide the best possible diagnosis. Immunological diagnosis of snake envenomation by the detection ofvenom antigen in the blood or other body fluids has been developed and shown to be useful [13] but is still not commercially available in most parts of the world. When the correct diagnosis of the type of envenomation is made, the specific antivenin can be prescribed, therefore decreasing morbidity and mortality. Determination of whether the bite was by a venomous or nonvenomous snake and if there was envenomation or not may prevent unnecessary administration of antivenin, which sometimes can, by itself, cause untoward effects, including life-threatening hypersensitivity reactions [3]. Acknowledgment The authors thank Vera Lucia Campos Brites, from the Universidade Federal de UberHindia, who identified the snakes. References 1. Brasil. Acidentes ofidicos: distribuir,;ao dos casos confirmados por UF, Brasil, 1980-1991. Infonne Epidemiol6gico do SUS 1992; 1(1), 21. 2. Cardoso J.L.C. and Brando R.B. Acidentesporanimais pefonhentos. Clincia e tratamento. 1st ed. Sao Paulo: Livraria Editora Santos, 1982. 3. Silveira P.V.P. and Nishioka S.A. Non-venomous snake bite and snake bite without envenoming in a Brazilian hospital. Analysis of 91 cases. Rev Inst Med Trop sao Paulo 1992; 34(6), 499-503. 4. Hennekens C.H. and Buring J.E. Epidemiology in Medicine. 1st ed. Boston: Little, Brown and Company, 1987. 5. Nishioka SA., Silveira P.v.P. and Menezes L.B. Coral snake bite and severe local pain. Ann Trop Med Parasitol1993; 87,429-31. 6. Ribeiro L.A. and Jorge M.T. Epidemiologia e quadro clinico dos acidentes par serpentes Bothrops jararaca adultas e filhotes. Rev Inst Med Trop sao Paulo 1992; 32, 436-42. 7. Silveira P.V.P. and Nishioka S.A. South American rattlesnake bite in a Brazilian teaching hospital. Clinical and epidemiological study of 87 cases, with analysis of factors predictive of renal failure. Trans R Soc Trop Med Hyg 1992; 86,562-4. 8. Nickerson M.A. and Henderson R.W. A case of envenomation by the South American colubrid, Philodryas olfersii. Herpetologica 1976; 32,197-8. 9. Silva M.V. and Buononato M.A. Relato Clinico de envenenamento humano por Philodryas olfersii. Mem Inst Butantan, 1983/84; 47/48, 121-6. 10. Pinto R.N., Silva, N.J., Jr., and Aird S.D. Human envenomation by the South American opisthoglyph Clelia delia plumbea (Wied). Toxicon 1991; 29, 1512-6. 11. Nishioka S.A. and Silveira P.V.P. Philodryas patagoniensis bite and local envenoming. Rev Insti Med Trop sao Paulo 1994 (in press). 12. Minton S.A. Venomous bites by nonvenomous snakes: an annotated bibliography of colubrid envenomation. J Wilderness Med 1990; 1, 119-27. 13. Theakston, R.D.G. Snake venom in science and clinical medicine. 2. Applied immunology in snake venom research. Trans R Soc Trop Med Hyg 1989; 83, 741-4.