Epidemiology of snake bites in selected areas of Kenya

Introduction Snake bites are a silent public health problem in Kenya. Previous studies on snake bites in the country have mainly focused on identifying offending snake species, assessing the severity of envenomation and testing the efficacy of antivenom. Factors associated with snake bites in the country are yet to be fully understood. The aim of this work was to determine pharmaco-epidemiological factors associated with snake bites in areas of Kenya where incidence, severity and species responsible for snake bites have been reported. Methods Kakamega provincial hospital, Kabarnet, Kapenguria and, Makueni district hospitals were selected as study sites based on previous findings on incidence, severity and species responsible for snake bites in catchment areas of these hospitals. Persistent newspaper reports of snake bites in these areas and distribution of snakes in Kenya were also considered. Cases of snake bites reported between 2007-2009 were retrospectively reviewed and data on incidence, age, site of the bites, time of bite and antivenom use was collected. Results 176 bites were captured, 91 of which occurred in 2009. Individual incidence was between 2.7/100,000/year and 6.7/100,000/year. Bites peaked in the 1-15 year age group while 132/176 bites were in the lower limb area and 49/176 victims received antivenom. Most bites occurred during the dry season, in the bush and in the evening. Overall mortality was 2.27%. Conclusion There is a need to sensitize the Kenyan public and healthcare personnel on preventive measures, first aid and treatment of snake bites.


Introduction
Snake bites are a neglected emergency in Kenya. This is because there is low awareness of snake bites as a public health problem in the country. Few studies have been carried out to evaluate the magnitude of the problem of snake bites in Kenya. Coombs and coworkers reported the incidence of snake bites in Kakamega and Western Kenya, Lake Baringo and Laikipia, Kilifi and Malindi as well as Northern Kenya to be between 1.9/100,000/year and 67.9/100,000/year [1]. Additionally, they reported that the mortality rate of snake bites in these areas was 0.45/100,000/year [1].
Furthermore, Kihiko reported that snake bites at Kitui District hospital were majorly characterized by compartment syndrome and focal gangrene [2]. Puff adders, black spitting cobras, black mambas and the boomslang have been reported to be behind a majority of the snake bites in Kenya [3][4][5][6]. Recently, Harrison and others reported that some of the snake antivenoms in Kenya were not pre-clinically effective against the medically important snakes in the country [7]. However, there is a paucity of information on the pharmaco-epidemiological factors associated with snake bites in Kenya. The aim of the present study was to determine the pharmaco-epidemiological factors associated with snake bites in selected areas of Kenya where previous reports of the incidence, severity and species responsible for snake bites have been documented.

Methods
Study sites: A five-point criterion was used in selecting the study sites. I) Findings of a report by Coombs and co-workers on the incidence, severity, and species responsible for snake bites in the catchment areas of the study sites. II) Findings of a report on the distribution of medically important snake species in catchment areas of the study sites [6] (Table 1). III) Multiple newspaper reports of snake bites in catchment areas of the study sites [8][9][10][11][12][13][14]. IV) Reports that environmental conditions similar to those experienced in the catchment areas of the study sites favor snake inhabitation [15][16][17].
V) The fact that the sites were the major health facilities within their Distribution of snake bite victims based on gender: Ninety (90) victims were male while 86 were female. All clinicians interviewed opined that males were more likely to be bitten as compared to females. Moreover, all clinicians interviewed opined that the lower limbs were the most likely part of the body to be bitten.

Distribution of antivenom use based on the number of bites:
Kapenguria had the least (14/176) number of bites but had the highest rate of antivenom use (71.43%) while Kakamega had the highest number of bites (77/176) but with only 29.87% antivenom use. Moreover, despite Makueni having the second highest number of bites, the rate of antivenom use was only 6.25% (Table 4).
Distribution of snake bite victims based on gender, age and antivenom use: Gender and age distribution of bite victims who received antivenom was as shown in Table 5 This group also registered the highest number of victims who did not receive antivenom ( Table 6).
Distribution of snake bite victims based on age, part of the body bitten and antivenom use: Victims bitten on the lower limbs received the most antivenom while those bitten on other parts of the body received the least antivenom ( Table 7).

Appraisal of informant age and experience:
The mean age of the informants was ~37 ± 11 years with a mean of 9.8±8.5 years of experience.

Other factors
Location, time of the day when the bites occurred and occupation of bite victims: All clinicians interviewed were in agreement that the bush was the most likely place to be bitten by a snake. Additionally, all informants with the exception of the informant in Kabarnet, considered residential areas to be the least likely place of occurrence of snake bites. Seventy-five percent (3/4) of the clinicians interviewed identified evening as the most likely time for snake bites to occur. All clinicians interviewed were of the opinion that manual laborers were the most likely group of people to be bitten by snakes.
Season of the year when bites are likely to occur: All clinicians interviewed considered bites to be higher in the dry season compared to the wet season.
The period of time taken by victims to seek treatment: All clinicians interviewed were of the opinion that victims were least likely to report to the hospital within two hours of having been bitten. Fifty percent (2/4) of the clinicians were of the opinion that victims of snake bites were most likely to present to the hospital between 2 and 6 hours after a bite. According to 75% (3/4) of the interviewees, bite victims delayed in presenting to the healthcare facilities owing to the long distances they had to travel to get to the hospital. One interviewee held the belief that poor infrastructure and poverty contributed to delays in seeking treatment for snake bites.

Cultural beliefs of victims:
Seventy-five percent (3/4) of clinicians held the belief that victims of snake bites sought treatment from traditional healers before visiting the hospital. One interviewee was of the opinion that this observation was due to the easy accessibility of traditional healers. Interviewees in Kapenguria and Makueni ascribed this to the belief that only bewitched people were bitten by snakes. One (25%) of the interviewees opined that culture was the reason behind the use of herbal medications in managing bites.

Management of snake bites:
Seventy-five percent (3/4) of the clinicians interviewed reported that supportive care followed by the administration of antivenom was the most common course of action adopted in managing snake bites. Supportive care on its own was regarded by 50% of interviewees as the second most likely course of action to be adopted in managing snake bites. The interviewees unanimously agreed that reassurance of the victims was the least considered course of action in managing snake bites. Reassurance was meant to calm anxious patients who were victims of snake bites.
Availability of antivenom in the hospitals: All the clinicians interviewed opined that antivenom was not always available in the hospital. Seventy-five percent (3/4) of interviewees attributed this to delays in procurement of antivenom as well as a shortage in supply.
However, one of the interviewees attributed the non-availability of antivenom in the hospital to be due to absenteeism by custodians of antivenom in the health facilities. All clinicians interviewed were in agreement that despite antivenom being available sometimes, its use was not always routine. Moreover, the diagnosis that a bite was not from a poisonous snake was unanimously identified as the main reason for the non-usage of antivenom. According to 75% of the interviewees, the occurrence of a bite from a poisonous snake was ruled out if the patient had stayed for more than 24 hours without any symptoms of envenomation. Additionally, 75% (3/4) of the interviewees held the belief that visits by snake bite victims to traditional herbalists were the second most likely reason to hamper the use of antivenom. Seventy-five percent of the interviewees considered side effects as the least likely reason to hamper the use of antivenom.

Effectiveness of antivenom:
Seventy-five percent of the interviewees were of the opinion that the antivenom available in the health facilities were effective. One interviewee had a dissenting opinion based on her experience of a poor prognosis of snake bites even after the victims had received antivenom.
Challenges involved in management of snake bites: All interviewees opined that regular and timely supply of antivenom was the main challenge in the management of snake bite. Seventyfive percent of the interviewees were of the opinion that delayed assessment of patients who qualified for antivenom administration was the second most prevalent challenge they faced.

Policies that may improve management of snake bites:
Seventy-five percent of clinicians interviewed held the belief that timely supply of antivenom and proper training on management of snake bites would be the most significant policy measures in improving outcomes in the health facilities. One interviewee was of the opinion that urgency in the management of snake bites, timely supply of antivenom and continuous medical education programs would be the most important steps in improving outcomes.
Availability, diversity, cost of antivenom and referral of patients to purchase antivenom in private pharmacies: All nurses interviewed were of the opinion that antivenom was available in the health facilities but that there were periods of unavailability. Seventy-five percent of the interviewees reported that antivenom from only one manufacturer was available. One interviewee reported that her facility had antivenom from two different manufacturers. All interviewees reported that antivenom was offered free of charge in the health facilities. 75% of the interviewees were of the opinion that the health facilities did not refer snake bite victims to purchase antivenom from private pharmacies.
Availability of antivenom in private pharmacies: It was unanimously agreed amongst the pharmacy personnel interviewed that most private pharmacies did not stock antivenom. Forty-two percent (5/12) of the interviewees attributed the non-availability of antivenom to low demand and the high cost of antivenom to patients while 33% (4/12) of the interviewees attributed the nonavailability to be due to low demand only. Moreover, 92% (11/12) of interviewees reported that they received prescriptions from clinicians to supply antivenom to victims of snake bites. One interviewee, however, reported that he had never received any prescription from clinicians to supply antivenom. Ninety-two percent (11/12) of the interviewees were of the opinion that bottlenecks in the procurement of antivenom were a challenge in stocking antivenoms in private pharmacies.

Discussion
The number of snake bites over the 3 year period was 176 with an incidence of 6.7/100,000/year, 4.6/100,000/year, 2.7/100,000/year and 5.4/100,000/year for Kabarnet, Kakamega, Kapenguria, and Makueni respectively. These estimates were within the range of 1.9/100,000/year and 67.9/100,000/year reported by Coombs and colleagues [1]. However, these estimates were much lower compared to what was reported elsewhere in Africa [22]. The differing incidence of snake bites may be a reflection of the different rates of hospitalization across the catchment areas of these health facilities [22]. Individuals in the 1-15 and 16-30 year age groups were subject to more bites compared to any other age group.
Campbell and others reported that snake bites were more prevalent in children than in adults [23]. The curiosity of children, their size, the mobility of their arms and legs while playing and the increased exposure of appendages may prompt snakes to bite children more than adults [24]. Individuals in the 16-30 year age group mainly comprise of adolescents and young adults who are actively involved in herding, gathering firewood, and farming activities. Thus, individuals in this age group may be exposed to snake bites as they perform these duties. Snake bites on the lower limbs accounted for in the study area held the belief that available antivenoms were efficacious based on their experience of using them. However, it may not always be possible to vouch for the pharmacological efficacy of antivenom available in hospital settings [22]. The antivenoms we came across in the study area were manufactured in India and France respectively. The Indian antivenom was produced by Bharat Serums and Vaccines Ltd India, batch number A2608003, expiry date 07/11 while the French antivenom was manufactured by Sanofi Pasteur, batch number D5211, expiry 2/2011. Snakes of medical significance in the study area include the black mamba, puff adder, Jameson's mamba and Forest cobra among others [6] ( Table   1). The list of snakes on Table 1 may not be exhaustive as there may be a great diversity of snake species in Kenya and some are yet to be identified. Nonetheless, on account of the high survival rate, low uptake of antivenom and low mortality rate observed in our study, it may be suggested that victims may have been subject to "dry bites" or may have been bitten by non-venomous snakes.
According to Stewart [40], snake bites may not always be accompanied by envenomation. The mortality rate we observed was much lower than previous reports from other parts of the world [22,41]. It may be difficult to interpret this finding without more information on the actual incidence and outcomes of snake bite in the study area. This study was not without limitations, some of which may affect the extrapolation and interpretation of the data collected. First, victims who may have sought treatment outside the formal healthcare system and did not report bites to our study sites were not captured. This also applies to victims who may have sought care outside the catchment area of our study. Thus, the incidence we have reported may most likely be an underestimate of the true incidence of snake bites in the areas. Secondly, the retrospective nature of our study did not allow for the identification of the offending snakes. Finally, clinical signs of envenomation, adverse effects of antivenom, traditional therapies advanced by herbalists in snake bite management and inventory of antivenom were not reported. The findings of this research can, therefore, be generalized only within these limitations. However, we hold the belief that our facility based study has provided useful insights on factors associated with snake bites in selected areas of Kenya.

Conclusion
These findings suggest that there is low reporting of snake bites in  There is no antivenom that is pre-clinically effective against all the snakes of medical importance in Kenya.