Professional Documents
Culture Documents
Michael 2011
Michael 2011
a r t i c l e i n f o a b s t r a c t
Article history: We studied pre-hospital practices of 72 consecutive snake bite victims at a hospital in north-
Received 2 March 2010 central Nigeria. The primary outcome assessed was death or disability at hospital discharge.
Received in revised form
Victims were predominantly male farmers, and in 54 cases (75%) the snake was identified
23 September 2010
as a carpet viper (Echis ocellatus), with the remainder unidentified. Most subjects (58, 81%)
Accepted 23 September 2010
Available online 28 October 2010 attempted at least one first aid measure after the bite, including tourniquet application
(53, 74%), application (15, 21%) or ingestion (10, 14%) of traditional concoctions, bite site
incision (8, 11%), black stone application (4, 5.6%), and suction (3, 4.2%). The majority (44,
Keywords:
rural health 61%) presented late (after 4 hours). Most (53, 74%) had full recovery at hospital discharge.
envenomation Three deaths (4.2%) and thirteen (18%) disabilities (mainly tissue necrosis) occurred. The use
viper of any first aid was associated with a longer hospital stay than no use (4.6 ± 2.0 days versus
Africa 3.6 ± 2.7 days, respectively, P = 0.02). The antivenom requirement was greater in subjects
first aid who had used a tourniquet (P = 0.03) and in those who presented late (P = 0.02). Topical
epidemiology application (Odds Ratio 15, 95% CI 1.4-708) or ingestion of traditional concoctions (OR 20,
95% CI 1.4-963) were associated with increased risk of death or disability. Ingestion and
application of concoctions were associated with a longer time interval before presentation,
a higher cost of hospitalization, and an increased risk of wound infection.
© 2010 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd.
All rights reserved.
0035-9203/$ – see front matter © 2010 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.trstmh.2010.09.005
96 G.C. Michael et al. / Transactions of the Royal Society of Tropical Medicine and Hygiene 105 (2011) 95–101
Table 2
Risk of death or disability for various first aid measures and characteristics
Presenting time
Early (≤4 hours after bite) 28 3 (10.7) 1
Late (>4 hours after bite) 44 13 (29.5) 3.5 0.90-14 0.06
First aida
Tourniquet 53 14 (26.4) 4.7 0.58-212 0.16
Concoction applied to bite 15 8 (53.3) 15 1.4-708 0.01
Concoction ingestion 10 6 (60.0) 20 1.4-963 0.009
Incision 8 2 (25.0) 4.3 0.18-275 0.53
Blackstone 4 2 (50.0) 13 0.39-823 0.11
Suction 3 0 0 0-182 1.0
None 15 1 (6.6) 1.0
a
More than one first aid measure may have been taken
of black stone was also significantly associated with older for wound infection with antibiotics and wound dressings.
age (52.8 ± 29.6 years for users versus 21.7 ± 13.0 years for Disability was present in 13 subjects (18%) at discharge,
nonusers, P = 0.02). Application of a traditional concoction including limb ulceration and limp, sublingual haematoma
on the bite site was significantly associated with female with restricted tongue movement, and knee joint stiffness.
gender (46.7% females for users versus 21.1% females for Twenty subjects (27.8%) did not return for the one week
non users, P = 0.04). follow up appointment, but all 13 with disability at hos-
pital discharge returned. One had residual limb ulceration
3.3. Presentation to the hospital and one had residual joint stiffness that was improving.
The mean duration of hospitalization was 4.4 ± 2.2 days,
The majority of victims (48, 66.6%) presented for their ranging from 1-12 days. The cost of hospitalization ranged
first treatment at the Comprehensive Health Centre. Most from 200 naira (US$1.54) to 33 710 naira (US$259.30) with
(45, 62.5%) presented late (more than 4 hours after the a median of 1549 naira (US$11.92).
bite). Timing of presentation after the bite ranged from Late presentation was not significantly associated with
1-308 hours with a median of 5.0 hours. Reasons for late disability or death (Table 2). Among the first aid measures,
presentation included a long distance to the hospital (18, topical application (OR 15, 95% CI 1.4-708) or ingestion
40%), lack of transportation (7, 15.6%), no money (5, 11.1%), of concoction (OR 20, 95% CI 1.4-963) were significantly
father’s absence (5, 11.1%), treatment by traditional healer associated with increased risk of death or disability. The
(4, 8.9%), lack of awareness of available treatment at the median (interquartile range [IQR]) time to presentation in
hospital (2, 4.4%), treatment at the church (2, 4.4%) and those who topically applied a traditional concoction was
belief that the snake was not venomous (2, 4.4%). Most 29 (5-141) hours compared with 5 (3-8) hours in those
patients (52, 72.2%) rode a motorcycle for transport to the who did not. Similarly, the median (IQR) time to presen-
hospital. The median time to presentation was 5 (4-20) tation in those who ingested a concoction was 16.5 (6-42)
hours in those who used any first aid and 4 (3-6) hours hours compared with 5 (3-10) hours in those who did
in those who did not (P = 0.13). not. In a logistic regression analysis, late presentation was
All subjects had features of snake bite on admission, no longer significantly associated with death or disability
and nearly half presented with spontaneous bleeding when controlling for the application or ingestion of tradi-
(Supplementary Table 1). A total of 23 (32%) had anaemia tional concoctions.
(haematocrit <35%), and two received blood transfusions. The proportion of subjects whose cost of hospitaliza-
tion was less than 1000 naira (US$7.70) was significantly
3.4. Outcomes greater among those who did not use first aid than in those
who did (57.1% versus 25.9%, respectively, P = 0.02). As
Most (61, 84.7%) subjects received antivenom prior to shown in Table 3, the median cost was significantly greater
discharge (Supplementary Table 1). Of these, 52 (72.2%) among subjects who applied or ingested traditional con-
required antivenom on admission, and 9 (12.5%) developed coctions than in those who did not. A significantly greater
coagulopathy later in the hospital course. Only 6 (9.8%) median cost was associated with delayed presentation than
subjects had antivenom within four hours of the bite, and early presentation to hospital (P = 0.04).
the median interval between bite and commencement of The use of any first aid was associated with a longer
antivenom was 11 hours. The median antivenom quantity duration of hospital stay than non use (4.6 ± 2.0 days versus
used by subjects was 20.0 mL. 3.6 ± 2.7 days, respectively, P = 0.02). The mean duration
Most (53, 73.6%) subjects had full recovery at hospi- of hospitalization was significantly longer in those who
tal discharge, but 3 (4.2%) subjects left against medical used a tourniquet than in those who did not (4.6 ± 2.0
advice without full recovery. Death occurred in 3 subjects, days versus 3.7 ± 2.5 days, respectively, P = 0.04). The mean
resulting from spontaneous bleeding from multiple sites or duration of hospitalization was significantly shorter in
seizures and coma, probably due to intracranial bleeding. those who incised the bite site than in those who did not
During hospitalization, 30 subjects (41.7%) were treated (2.9 ± 1.6 days versus 4.6 ± 2.2 days respectively, P = 0.03).
G.C. Michael et al. / Transactions of the Royal Society of Tropical Medicine and Hygiene 105 (2011) 95–101 99
Table 3
Comparison of cost and duration of hospitalization of snake bite victims
Tourniquet
Yes 53 1745 0.27 5 0.04
No 19 1080 3
Concoction ingestion
Yes 10 7690 0.02 4 0.84
No 62 1375 4
Incision
Yes 8 1255 0.55 2.5 0.03
No 64 1549 4
Blackstone
Yes 4 1255 0.65 2.5 0.09
No 68 1549 4
Suction
Yes 3 1178 0.68 6 0.70
No 69 1500 4
Presenting time
Early (≤4 hours after bite) 28 1085 0.03 4 0.75
Late (>4 hours after bite) 44 1975 4
a
More than one first aid measure may have been used
b
At the time of the study $US1 = 130 naira; c As determined by Mann-Whitney test
However, there was no significant difference in the mean infection. A greater antivenom requirement and longer
duration of hospitalization between early and delayed pre- hospital stay were associated with tourniquet use.
sentation (4.2 ± 1.8 days versus 4.5 ± 2.4 days respectively, Venomous snake bite still remains a preventable cause
P = 0.75). of morbidity and mortality. A retrospective study of 103
As shown in Table 4, the antivenom requirement was
significantly greater in subjects who had used a tourni-
Table 4
quet (P = 0.03) and in those who presented late (P = 0.02).
Antivenom requirement for the various groups
Both of these factors remained significant in a linear regres-
sion model. Topical application of traditional concoction to Groups Median (IQR) antivenom P value
requirement (mL)
the bite site and ingestion of traditional concoctions were
significantly associated with six-fold and four-fold greater Tourniquet
odds of wound infection, respectively (Table 5). Late pre- Yes 20.0 (20-40) 0.03
No 20.0 (10-20)
sentation was associated with three-fold greater odds of
wound infection. Incision
Yes 25.0 (0-35) 0.71
When we compared those with or without an identified
No 20.0 (20-35)
snake species, there were no significant differences in out-
comes. No deaths occurred in the group with unidentified Suction
Yes 50.0 (0-60) 0.45
snake species.
No 20.0 (20-30)
Black stone
4. Discussion Yes 30.0 (15-35) 0.71
No 20.0 (15-35)
We found that the commonest pre-hospital first aid Concoction applied to bite
in a Nigerian rural health centre was tourniquet applica- Yes 30.0 (20-50) 0.07
tion. Topical application of traditional concoctions to the No 20.0 (10-30)
bite site, ingestion of traditional concoctions, bite site inci- Ingestion of concoction
sion, suction and use of black stone were other measures Yes 30.0 (20-30) 0.13
employed. Ingestion and application of concoctions were No 20.0 (10-40)
both significantly associated with a longer time interval Late presentation (>4 hours after bite)
before presentation to the hospital, a greater risk of death Yes 20.0 (20-45) 0.02
No 20.0 (5-30)
or disability, a higher cost of hospitalization, and wound
100 G.C. Michael et al. / Transactions of the Royal Society of Tropical Medicine and Hygiene 105 (2011) 95–101
Table 5
Risk of wound infection for various first aid measures
Presentation
Early (≤4 hours after bite) 28 7 (25) 1.0
Late (>4 hours after bite) 44 22 (50) 3.0 1.1-8.5
patients admitted for snake bite in the same centre a hospital in north eastern Nigeria, 2.9% of patients who
recorded no deaths and 3.9% with tissue necrosis, but 52% ingested herbs presented with jaundice possibly induced
presented to the centre within three hours of bite.12 In by the unidentified herbs.18 We found that victims who had
contrast, in our study we found that 61% arrived late with ingested concoctions had a much greater risk of complica-
a median interval between bite and antivenom adminis- tions compared with those who did not, suggesting that
tration of nine hours. Despite an established treatment this is a potentially dangerous practice. However, none of
protocol, free antivenom and trained staff, death occurred the study subjects developed jaundice. The increased risk
in 4.2% and complications in 22% of victims. During the of wound infection associated with application of concoc-
study, the supply of antivenom was inconsistent, and fam- tions to the bite sites that we observed has been previously
ily members of patients had to travel long distances to documented.19 The association of traditional concoction
purchase antivenom before administration. Less than 10% ingestion with wound infection that we found may be
of victims received antivenom within four hours of the related to a longer delay before presentation among those
bite. Although not directly comparable to the cytotoxic who utilized traditional concoctions. Seeking care from a
effect of viper bites, patients systemically envenomed by traditional healer for snake bite is associated with delayed
Papuan taipans (Oxyuranus scutellatus canni) who received presentation to the hospital and exposes them to useless
antivenom within four hours had hastened resolution of or dangerous practices.8,14,20 However, we found that dis-
neurotoxic symptoms and shorter hospital stays.9 tance from where the bite occurred to the hospital was the
Attempts at first aid were a common practice after snake predominant reason for delayed arrival. Less than 10% of
bite, confirming the observations in a previous study.12 victims initially sought care from a traditional healer. The
Similar to a study in Nepal,8 we found that tourniquet causes for late presentation are not likely to be independent
use was the most common form of first aid used. Tourni- of one another.
quet use in viper bite increased the antivenom requirement There was no association between late presentation
and duration of hospitalization. Tourniquet use is known and primary outcome (death or disability), suggesting that
to cause an increase in local complications by increasing there was no time at which antivenom could be considered
tissue anoxia and it may trigger severe systemic envenoma- ineffective after snake bite.9 Late presenters tended to have
tion after removal. Consequently, its use has been strongly more severe envenomation and required greater quantities
discouraged by experts.13,14 We found that one-quarter of of antivenom, which contributed to their increased cost of
subjects who used a tourniquet developed complications. hospitalization. Other investigators in west Africa found
The use of incision as first aid in this study was asso- that a delay in consultation did not affect the prevalence
ciated with shorter duration of hospitalization. Incision of bleeding, coagulopathy, or the number of administered
of bite site or limb as first aid has been associated with antivenom doses.21
adverse effects such as tissue damage, haemorrhage and Among the limitations of this study, we cannot exclude
infection.15 We do not believe that incision was respon- selection bias. Those who presented to the hospital with
sible for a shorter hospital stay. We suspect that subjects snake bite may differ in important ways from those who did
who incised the bite had other unmeasured confounding not present to hospital after snake bite. Those with dry bites
characteristics that were responsible for their shorter hos- or minimal envenomation may have not been recruited due
pitalization. Alternatively, subjects who used incision may to improvement with first aid. Unmeasured confounding
have avoided potentially more harmful practices, like use variables that are associated with both the exposure and
of a tourniquet. Suction was rarely used by victims in our the outcome may lead to spurious associations. The study
study. Suction use is controversial and has limited efficacy, was also limited by its sample size. The number of subjects
especially when applied five minutes after the bite, and in the subgroups that practiced some forms of first aid was
increases the risk of skin necrosis if applied after thirty small, which may account for a lack of statistical power to
minutes.15–17 find significant effects of these practices on outcomes.
While it is well documented that application of tra- Traditional first aid measures for viper bites, like use
ditional herbs, animal dung, mud and ashes to the bite of tourniquets and traditional concoctions, potentially
site or ingestion of alcohol have been used as first aid,7,8 contribute to morbidity and mortality in rural Nigerian
the ingestion of other traditional concoctions is not well communities. Delays in antivenom administration due to
described. In a four-year retrospective prevalence study at late presentation and erratic antivenom supply are also
G.C. Michael et al. / Transactions of the Royal Society of Tropical Medicine and Hygiene 105 (2011) 95–101 101
additive factors. Community education on appropriate 2. Pugh RN, Theakston RD. Incidence and mortality on snake bite in
actions after snake bite, including limb immobilization, savanna Nigeria. Lancet 1980;2:1181–3.
3. Habib AG. Tropical snake bite in Northern Nigeria: a clinical review.
cleansing the bite site with soap and water and rapid Niger Med PracM 1992;23:3–8.
transport of victims to the hospital could reduce the com- 4. Echitab (anti-snake venom) Study Group Nigeria-UK. Nigeria proto-
plications from viper bites in Africa. col. 1998/1999.
5. Gutierrez JM, Theakston RD, Warrell DA. Confronting the neglected
problem of snake bite envenoming: the need for a global partnership.
Authors’ contributions: GCM was involved in the concep- PLoS Med 2006;3:e150.
tion, design, data collection, analysis, interpretation and 6. Russell FE. When a snake strikes. Emerg Med Clin North Am
1990;22:20–43.
drafting of the manuscript of this study. TDT was involved
7. Cheng AC, Currie BJ. Venomous snake bites worldwide with focus on
in the conception, design, analysis, interpretation, drafting the Australia-Pacific region: current management and controversies.
of manuscript and revising of the paper and was the chief J Intensive Care Med 2004;19:259–69.
8. Sharma SK, Chappuis F, Jha N, Bovier PA, Loutan L, Koirala
supervisor of the study. MILS was involved in the plan-
S. Impact of snake bites and determinants of fatal out-
ning and supervision of the study and in critical revision comes in southeastern Nepal. Am J Trop Med Hyg 2004;71:
of the manuscript. GCM and TDT are the guarantors of the 234–8.
manuscript. 9. Trevett AJ, Lalloo DG, Nwokolo NC, Naraqi S, Kevau IH, Theakston
RD, et al. The efficacy of antivenom in the treatment of bites by the
Papuan taipan (Oxyuranus scutellatus canni). Trans R Soc Trop Med Hyg
Acknowledgements: The authors are grateful to Drs Bukar 1995;89:322–5.
A. Grema and Ahmed Loh for their support and advice 10. Baldwin M. The snakestone experiments. An early modern medical
debate. Isis 1995;86:394–418.
and to Dr A. Solomon for ensuring the availability of free 11. Chippaux JP, Ramos-Cerrillo B, Stock RP. Study of the efficacy of the
antivenom. We appreciate Ms O. Ifeoma for laboratory black stone on envenomation by snake bite in the murine model.
support and Mr Nambut for his helpful interaction with Toxicon 2007;49:717–20.
12. Madaki JKA, Obilom R, Mandong BM. Clinical presentation and out-
community leaders. We acknowledge Drs Ofakunrin and come of snake bite in patients at Zamko comprehensive health
Chidi for their efforts in the snake bite prevention campaign centre. Highland Med Res J 2004;2:61–5.
and in focus group discussions. 13. Gold BS, Dart RC, Barish RA. Bites of venomous snakes. N Engl J Med
2002;347:347–56.
14. Warrell DA. WHO/SEARO Guidelines for the clinical management of
Funding: Antivenom was provided by the Nigerian Fed- snake bites in the Southeast Asian region. Southeast Asian J Trop Med
eral Government-Echitab Study Group. Public Health 1999;30:1–85.
15. Ralidis PM. Medical treatment of reptile envenomation: a
review of the current literature. Top Emerg Med 2000;22:16–
Conflicts of interest: None declared. 36.
16. Hall EL. Role of surgical intervention in the management of
Ethical approval: Ethical Committee of Jos University crotaline snake envenomation. Ann Emerg Med 2001;37:175–
80.
Teaching Hospital. 17. Bush SP. Snakebite suction devices don’t remove venom: they just
suck. Ann Emerg Med 2004;43:187–8.
Appendix A. Supplementary data 18. Mustapha SK. Snake bite in Gombe. Highland Med Res J 2003;1:
22–7.
19. Bogdan GM, Dart RC, Falbo SC, McNally J, Spaite D. Recurrent coagu-
Supplementary data associated with this arti- lopathy after antivenom treatment of crotalid snakebite. South Med
cle can be found, in the online version, at J 2000;93:562–6.
20. Kalantri S, Singh A, Joshi R, Malamba S, Ho C, Ezoua J, et al. Clinical
doi:10.1016/j.trstmh.2010.09.005.
predictors of in-hospital mortality in patients with snake bite: a ret-
rospective study from a rural hospital in central India. Trop Med Int
References Health 2006;11:22–30.
21. Chippaux JP, Massougbodji A, Stock RP, Alagon A. Clinical trial of an
F(ab’)2 polyvalent equine antivenom for African snake bites in Benin.
1. Chippaux JP. Snake-bites: appraisal of the global situation. Bull World
Am J Trop Med Hyg 2007;77:538–46.
Health Organ 1998;76:515–24.