Spatial distribution of snakebites in the State of Ceará, Brazil (2008-2018)

Snakebites represent an important, though neglected, public health problem especially in the Northeast of Brazil. There is still a large number of underreported cases, the collection of epidemiological data is deficient, and the ecological and epidemiological knowledge of this envenomation remains limited due to the lack of updated regional and local information. This study describes the epidemiological profile of snakebites recorded by the Notifiable Diseases Information System from 2008 to 2018 in the state of Ceará. The data were organized according to the municipalities of the health macro-regions, considering the variables: sex, snake, gravity and deaths. A total of 8,233 cases was found, with an annual incidence of 8.6/100,000 inhabitants, and a higher prevalence of accidents caused by Bothrops snakes (64%), in men (87.4%) in the Sobral region (25.5%). However, the highest incidence rate was found in the Sertão Central (299.7/100,000). Most cases were considered mild (62%), but the mortality rate was 0.07/100,000 and the case-fatality rate was 0.8%. Among the most affected municipalities, cities like Tauá (1.3%), Sobral (8.1%) and Fortaleza (8.5%) are concerning, considering that the case-fatality rates were above the national average (1%). Snakebites in Ceará follow a pattern similar to that of other states in the Northeast, with a high frequency of incidents where the causative snake is not identified (16%). Considering the gravity of snakebites, studies that evaluate the anthropic and environmental influence in the distribution of cases may help to define priority risk areas, to improve health surveillance and patient care.


Introduction
Every day, an average of 7400 people are bitten by snakes, representing between 81,000 and 138,000 deaths, annually, worldwide, with about 400,000 disabilities such as amputations and post-traumatic stress disorder (Minghui et al., 2019, Gutiérrez et al., 2017. Most of these accidents are distributed in tropical and subtropical regions of the world, where there is a great diversity of snakes coexisting with the human peasant population with limited access to health and rural technology (Chippaux, 2012, Mise et al., 2016. Although in 2009 the World Health Organization (WHO) included snake bite on the list of Neglected Tropical Diseases (NTDs), it left this list in 2013, returning in 2017 as a global strategy to have the number of deaths and disabilities caused by this envenomation by 2030 (WHO, 2019). However, some studies argue that snakebites are not neglected (Simpson & Norris 2009), but underestimated and fundamentally poorly understood (Fry, 2018).
In Central and South American countries, most of the available data on snakebites is fragmentary and not very representative (Chippaux, 2017). Among the South American countries, Brazil is the one that reports the highest number of cases of snakebites per year (Gutiérrez et al. 2017), registering 265,701 incidents in 2019(Sinan, 2019. In Brazil, of the 405 species of snakes distributed throughout the territory, only 76 of them (19%) are considered of medical importance (WHO, 2019). However, most research is focused in bioprospecting biologically active molecules from the venom of these animals and developing antivenoms. As a result, the ecological and epidemiological knowledge of snakebites remains limited mainly in the states of the Northeast region, since there is still a lack of updated regional and local information, compatible with medical relevance, which hinders efforts to accurately map and mitigate these incidents (Murray et al., 2020).
Despite the existence of mandatory reporting systems in Brazil containing information on the epidemiological profile of snakebites in the State of Ceará, updated local and/or regional epidemiological information compatible with the medical importance of this envenomation is lacking. In addition, in the Northeast region there is still a large number of underreported cases and the collection of epidemiological data is deficient (Tavares et al., 2020). There are only three epidemiological studies in the literature that address snakebites in the state of Ceará (Feitosa et al., 1997, Albuquerque et al., 2013, Belmino, 2015 and in none of them the incidents are distributed by health macro-regions. Considering that snakebite rates follow considerable regional variation, the scarcity and outdated information on this type of notification, associated with the high number of victims, the importance of regionalized epidemiological mapping is evident to update knowledge and help to improve preventive measures and case management. Thus, this study aims to map the distribution of snakebites for health macroregions in the State of Ceará between 2008 and 2018.

Methodology
This is an epidemiological descriptive, retrospective, cross-sectional study, of quantitative approach of snakebite cases recorded between January 2008 and December 2018 in the state of Ceará, Brazil.

Study area
The state of Ceará is located in the north of the northeast region of Brazil and is bordered by the Atlantic Ocean (North), Rio Grande do Norte and Paraíba (East), Pernambuco (South), and Piauí (West). Ceará has a total area of 148,920.4 km², and an estimated population of 9,132,078 inhabitants (IBGE, 2019). The climate is predominantly hot semi-arid, characterized by scarce and irregular rainfalls (IBGE, 2019).

Data collection
The data were obtained through the records of the Information System for Notifiable Diseases (SINAN), nationwide, and grouped according to the health macro-regions (Fortaleza, Sobral, Cariri, Sertão Central and Litoral Leste/Jaguaribe), delimited by the Health Department of the state of Ceará (SESA). The location of the accidents (macro-health region and municipality) was considered for the analysis, associated with variables such as age, sex, snake identification, severity of the accident and death. Population data were obtained from the Brazilian Institute of Geography and Statistics (IBGE), based on the National Population Census (2010), and the estimated population for the inter-census years.

Ethics Statement
The present study followed the rules set out in the Resolution from the National Health Council No. 466 from December 12th, 2012, which states that research involving only secondary data in the public domain, without nominal identification of the research participants, does not require appreciation by a research ethics committee involving human beings.

Statistical analysis
The data were compiled, tabulated and analyzed using descriptive statistics (BiostatTGV software) to determine simple (n) and relative (%) frequencies and the results were presented in the form of contingency tables using Microsoft Excel 2019 software. The incidence, mortality and lethality rates were also calculated and it was used the estimated population for the state of Ceará between 2008 and 2018 as the standard. The chi-square test (χ 2 ) was used to test the association among variables to the significance level p < 0.05. Microsoft Excel 2019 was used to create tables and graphics. To create the distribution maps, the QGIS desktop 2.18.18 with GRASS 7.4.0 program was used.
Overall, most incidents affected men (n = 7184, 87.4%) and there was a significant difference (χ2 = 2914.6, p <0.05) compared to the number recorded for women (n = 2008, 12.6%). Most incidents were classified as mild (n = 4,665, 62%), followed by moderate (n = 1,918, 25%) and severe (n = 304, 4%) (χ2 = 4346.6, p < 0.05). The disproportion between men and women in cases of snakebites is probably related to the difference in the risks of human activities and/or distribution of snakes (Chippaux, 2015). In this way, snakebites could be the result of agricultural activities, assuming that they also involve more men than women. According to Da Silva et al. (2015), the main victims of snakebites in Brazil are men of working age (20 to 59 years old). These data agree with the results of the present study for the state of Ceará. Snakebite mortality in Brazil is associated with agricultural work, advanced age, delayed treatment time and inadequate antivenom therapy, characterizing this envenomation as an agricultural occupational health problem (Mise et al., 2019, Mise et al., 2016. The macro-regions of Sobral (n = 2,097, 25.5%), Sertão Central (n = 1,891, 23.0%) and Cariri (n = 1,871, 22.7%) recorded the highest number of cases (χ2 = 509.06, p < 0.05) (Table 1). However, in the historical series, the incidence was higher in the macro-regions of Sertão Central (299.7) and Litoral Leste / Jaguaribe (183.9) (Figure 1). Chaves et al. (2015) suggest that snakebites are influenced by climatic factors such as El-Niño, and are more frequent at high temperatures and with poverty indicators, but decrease with altitude, factors that correspond to the characteristics of Ceará's hotspots.
The increase in cases of snakebites occurs in the rainiest and hottest periods, coinciding with the period of greatest human activity in the field, and the reproductive period of snakes (Fry, 2018). It is suggested that climate changes that have  Bothrops envenomation was the most frequent in Ceará (5,276 cases, 64%, χ2 = 11391, p <0.05), with an annual average of 479.6 ± 125. Incidents by Crotalus were responsible for 713 cases (9%), followed by Micrurus (169 cases, 2%) and Lachesis (236 cases, 0.33%) (Figure 2). Incidents with non-venomous snakes (n = 660) represented 8% of the total recorded in the studied period, and the number of incidents with unidentified snakes (n = 1,324) represented a high percentage of the total records (16%), showing a growth trend of these accidents from 2015 ( Figure 2). These results correspond to the data of SINAN (2019) also for the Northeast region. Our study revealed a similar pattern for Ceará, but the percentage of incidents in which the snake was not identified is worrying, considering that this is an important for antivenom therapy. Lira-da-  suggests that the high frequency of bothropic incidents in the Northeast is justified by the fact that 50% of Brazilian Bothrops species occur in this region. was found in Fortaleza and Viçosa do Ceará. It is known that in the state of Ceará most municipalities dispose of their waste in landfills (Bastos et al., 2017), so the high rate of bothropic incidents can be explained by the enormous adaptive plasticity of these snakes, allowing them to also inhabit forested or deforested areas occupied by humans (Bochner et al., 2014).
The presence of Bothrops snakes in precarious urban areas in Brazilian municipalities can occur considering that they feed mainly on small rodents, abundant in these places, but also in wealthier neighborhoods that border natural preservation areas, offering shelter for snakes, and creating the conditions for the so-called "urbanization of snakebites" (França & França, 2019, Bochner et al., 2014. Interestingly Mise et al. (2007) showed that, although Bothrops are nocturnal snakes, most cases registered in the state of Bahia occurred during the day, showing the greater contact of man with these animals in periods of foraging and thermoregulation. In Brazil, the genus Crotalus is widely distributed and has a diet characterized by an extreme specialization in endothermic preys, such as rodents, because of its abundance throughout the year (Hoyos & Almeida-Santos, 2016). Our study revealed the wide distribution of Crotalus in the state of Ceará and the literature has already described that these snakes tend to occupy areas disturbed by human activities, which include landscape mosaics as original habitats, new environments and urban and agricultural areas.
In the Brazilian semi-arid region, where water is unpredictable, snakes may have morphological, physiological, behavioral and ecological adaptations to survive in this scenario (Da Costa et al., 2019). A recent study revealed that the diet is more important than the habitat to explain the partitioning of ecological niches by snakes, and thus, the coexistence of different species would be possible due to a high overlap in a specific niche dimension (microhabitat) would necessarily be possible due to a high overlap in a specific niche dimension (diet), which would avoid exclusion by competition (Di Pietro et al., 2020).
This hypothesis could perhaps explain the high number of incidents caused by Bothrops in extremely arid macro regions such as the Sertão Central, an environment predominantly preferred by snakes of the genus Crotalus.
Incidents involving Micrurus ( Figure 3C) showed great dispersion, but with an incidence and geographical concentration comparable to Lachesis, whose records were concentrated in the Sertão Central ( Figure 3D). Fortaleza and Viçosa do Ceará were the main hotspots of incidents by Micrurus (n= 10-12) and the incidents by Lachesis were concentrated in the municipalities of Caridade, Guaraciaba do Norte, Itatira, Ipueiras, Milhã, Quixeramobim, São Paulo. Benedito, Ubajara and Viçosa do Ceará (n= 1-10). Despite the great dispersion of accidents caused by Micrurus, due to its less aggressive behavior, the number of recorded incidents was low. Similar results in Paraíba and the Brazilian Amazon suggest that Micrurus appears to have a high degree of environmental plasticity, living well in both natural and anthropogenic environments (França & França, 2019 Figure 4 shows that of the 184 municipalities in the state of Ceará, only ten represent 21.1% (n = 1738) of the total records of snakebites for the period, highlighting Tauá (n = 305), Viçosa do Ceará (n = 291) and Tianguá (n = 233). The macro-region of Fortaleza had a higher case-fatality rate (n = 1396, 1.7%) in relation to Sobral (n = 2097, 0.7%) and Sertão Central (n = 1891, 0.7%) together. The municipality of Tauá had the highest number of snakebites notifications (n = 305), and the case-fatality rate (1.3%) was higher than the average for the state of Ceará (0.8%). Therefore, the case-fatality rate of municipalities such as Tauá (1.3%), Fortaleza (8.5%) and Sobral (8.1%) are worrying, considering that they were also higher than the national average (1.0%) (SINAN, 2019).
The different snakebite patterns presented by the various health macro-regions of Ceará suggest that snakebite studies should incorporate socioeconomic and environmental peculiarities in the analysis of this serious health problem. According to the demographic indicators in the last census (IBGE, 2010), only the macro-region of Fortaleza presented urbanization rates above the average in the state of Ceará, while the other geographic areas had a significant portion of the population residing in the rural area (30%), mainly in the macro-regions of Sertão Central and Litoral Leste/Jaguaribe, areas of high incidence of bothropic incidents. In addition, areas of the Caatinga biome show less socioeconomic development and are characterized by intensified agricultural activities during the rainy seasons, which may increase the chances of encounters with snakes (Mise et al., 2016).
In Ceará, the Atlantic Forest is restricted to brejos-de-altitude, which are usually residual remains with altitudes above 600 m, covered by forest type vegetation (Melo-Santos et al., 2006). The regional threats that occur in the brejos-de-altitude occur mainly because they are inserted in a region of great scenic beauty and pleasant climate. These characteristics make them historically exploited areas, attractive for ecotourism, but also suitable regions for planting subsistence crops that absorb the majority of the economically active population. An economy based on primary activities causes problems related to deforestation and the introduction of exotic species (Borges-Nojosa et al., 2019). These impacts, mainly caused by anthropic action, have caused the alteration or complete elimination of microhabitats, being the loss of habitat probably the greatest threat to biological diversity.
In many areas of Ceará, environmental degradation has already reached practically irreversible conditions, showing clear marks of desertification due to the strong impact of population growth pressure on the potential vulnerability of Caatinga's natural resources (Albuquerque et al., 2014). The State has contrasting humid and sub-humid areas due to geographical factors, such as altitude and proximity to the ocean, which alter the temperature, humidity and rainfall distribution, resulting in a landscape mosaic that dictate the distribution of snakes (Guedes et al., 2019). The study carried by Mateus et al. (2015) in Ceará found a decreasing trend in rainfall in urban and populated regions (Fortaleza), and in drier regions with less vegetated area (Sertão Central). However, in the regions of Cariri, Sobral and Litoral Leste/ Jaguaribe there was a tendency for increased rainfall, which may be affecting the distribution of snakes, especially Bothrops.
The growth observed in the number of snakebite incidents in Ceará can be attributed not only to a higher occurrence, but also to improvements in the Brazilian health information system. On the other hand, difficulties in accessing health services can contribute to the underreporting of accidents due to snakebites (Bochner et al., 2014). In addition, the incidence of snakebites may be underestimated due also to the use of traditional medicine in unreported cases. Thus, the main limitation of our study could be the reliability and precision of the notifications obtained by SINAN. A study developed by Benício et al. (2019) in Piauí found several flaws in the medical records of patients injured by snakebites, including the lack of identification of the causative species, time to care and poor clinical outcomes. However, considering the similarity with previous studies for Ceará (Feitosa, 1997, Albuquerque et al., 2013, Belmino, 2015, the data presented here seem to be consistent.
Regarding the underreporting and incomplete filling of forms, our data on the epidemiology of snakebites in the state of Ceará corroborate previous studies for the Northeast region (Saraiva et al., 2012, Fiszon & Lima et al., 2009. These notification failures are aggravated in small municipalities in Northeastern Brazil, where qualified health professionals are usually scarce to properly collect information on envenomation cases by venomous animals (Mise et al., 2018). This failure in official records indicates the need for better training of health professionals, and the establishment of more adequate protocols for recording epidemiological information on snakebites, seeking to improve serotherapeutic care and antivenoms distribution logistics (Bochner & De Sousa, 2019). Among the challenges for effective monitoring of snakebites in Brazil, its continental dimension and demographic, socioeconomic and environmental diversity stand out, resulting in difficulties in managing the distribution of antivenom at a local operational level (Bochner et al., 2014, Chippaux, 2015. Thus, it is essential to collect epidemiological data that assist in the seasonal anticipation of the number and location of cases.

Conclusion
Mandatory snakebite cases reporting appears to be a useful tool for improving antivenom management. However, local studies must be continued to consider the variability of the circumstances of the accident and to refine the necessary measures for its management. The pattern of snakebites in Ceará is similar to the one observed in other northeastern states, characterizing this envenomation as a regional public health problem, especially for the semi-arid municipalities, which indicates that future studies evaluating anthropic influence and environmental factors can define priority risk areas for this envenomation.