Older Age and Time to Medical Assistance Are Associated with Severity and Mortality of Snakebites in the Brazilian Amazon: A Case-Control Study

The Amazon region reports the highest incidence of snakebite envenomings in Brazil. We aimed to describe the epidemiology of snakebites in the state of Amazonas and to investigate factors associated with disease severity and lethality. We used a nested case-control study, in order to identify factors associated with snakebite severity and mortality using official Brazilian reporting systems, from 2007 to 2012. Patients evolving to severity or death were considered cases and those with non-severe bites were included in the control group. During the study period, 9,191 snakebites were recorded, resulting in an incidence rate of 52.8 cases per 100,000 person/years. Snakebites mostly occurred in males (79.0%) and in rural areas (70.2%). The most affected age group was between 16 and 45 years old (54.6%). Fifty five percent of the snakebites were related to work activities. Age ≤15 years [OR=1.26 (95% CI=1.03-1.52); (p=0.018)], age ≥65 years [OR=1.53 (95% CI=1.09-2.13); (p=0.012)], work related bites [OR=1.39 (95% CI=1.17-1.63); (p<0.001)] and time to medical assistance >6 hours [OR=1.73 (95% CI=1.45-2.07); (p<0.001)] were independently associated with the risk of severity. Age ≥65 years [OR=3.19 (95% CI=1.40-7.25); (p=0.006)] and time to medical assistance >6 hours [OR=2.01 (95% CI=1.15-3.50); (p=0.013)] were independently associated with the risk of death. Snakebites represent an occupational health problem for rural populations in the Brazilian Amazon with a wide distribution. These results highlight the need for public health strategies aiming to reduce occupational injuries. Most cases of severe disease occurred in the extremes of age, in those with delays in medical attention and those caused by Micrurus bites. These features of victims of snakebite demand adequate management according to well-defined protocols, including prompt referral to tertiary centres when necessary, as well as an effective response from surveillance systems and policy makers for these vulnerable groups.


Introduction
Snakebites represent a treatable health problem affecting predominantly tropical low-and middle-income countries [1,2]. The annual number of snakebite cases worldwide is largely unknown and probably underreported with estimates that could reach 5.5 million snakebites annually, with 94,000 estimated deaths [1]. Rural areas in developing countries are the most affected, with varying seasonality usually characterized by peak incidences seen in the rainy and harvesting seasons [3]. The burden of human suffering caused by snakebite has not received the attention it deserves from the public health community, development agencies and governments, and may therefore be appropriately categorized as a neglected tropical disease [1,4]. Most snakebite reporting systems are fragile and underestimate the actual injury burden [5]. Better epidemiological surveillance is necessary to assess the extent of this important public health problem to improve prevention and treatment measures. Most deaths and sequelae from snakebites are preventable by interventions such as making antivenom widely available [6].
In 2013, the Brazilian Ministry of Health reported 27,181 snakebite cases [7]. The highest incidence was observed in the Brazilian Amazon (52.6 cases/100,000 inhabitants) followed by the Midwest (16.4/100,000). These values are expected to be higher in remote areas of the Brazilian Amazon [8] due to considerable underreporting. In this region, adult males living in rural areas [3,[9][10][11][12][13] and/or workers exerting farming, hunting and forestry activities [14,15] are the most affected groups, strongly suggesting an occupational risk. A community survey conducted with forest-dwelling Indians and rubber tappers (seringueiros) revealed that 13% of the surveyed population had been bitten during their lifetime [14]. Snakebite incidence correlates with the period of higher rainfalls as increased river levels force snakes to seek new shelters closer to human settlements [3,9,11,12], highlighting the vulnerability of Amazonian communities living in riverbank areas.
Some studies indicate that snakes of the genus Bothrops account for nearly 80% of the snakebites in the Brazilian Amazon region [16][17][18]. Bothrops atrox and Lachesis muta share the same popular name 'surucucu' in certain Amazonian areas, bringing a confounding factor in snake identification by the local population [12]. In the Amazon, clinical data from Bothrops atrox bites show pain as the most frequent local manifestation, followed by swelling, warmth on palpation and necrosis [18,19]. Around 25% of the patients present systemic manifestations, including spontaneous systemic bleeding, ranging from 16-18% [16,18,19] and acute renal failure registered in 10.9% in a case series [19]. A total of 39.0% of patients developed secondary complications, such as cellulitis and abscesses [18]. The clinical picture caused by Lachesis bites is very similar to Bothrops', except for vagal symptoms that may be present especially in severe Lachesis cases [7,8]. In the Brazilian Amazon, relatively few accidents are caused by coral snakes (Micrurus sp.) [20] or rattlesnakes (Crotalus sp.) [21]. Surveillance systems shows an overall case-fatality rate ranging from 0.4 to 3.9% in the Brazilian Amazon [9,11,12,14,15].
Snakebite clinical severity varies by snake species and patient characteristics, but there are few studies identifying the more vulnerable subgroups developing poor outcomes. In India, patient-related factors such as child age, previous health condition and time elapsed until medical assistance can act as risk factors for severity, sequelae and death [22][23][24][25][26]. In the same country, patients admitted with symptoms of neurotoxicity and vomiting [22,24], coagulopathy and leukocytosis [23], hypertension, albuminuria, changes in bleeding and prothrombin time [25] were more likely to develop severe outcomes. The delay in antivenom administration was also a risk factor for severity in Nigeria [27]. Capillary leak syndrome, bleeding and respiratory paralysis were risk factors for mortality in patients who received snake antivenom in Nigeria and Korea [28,29].
A better knowledge of severity and mortality due to snakebite would lead to improved management, with expected reduction of sequelae and lethality rates in remote localities in the Brazilian Amazon. The aim of this study is to analyze the profile of snakebites reported in the state of Amazonas, in the Western Brazilian Amazon, and to investigate potential risk factors for severity and lethality.

Study area
The State of Amazonas is located in the western Brazilian Amazon, comprising an area of 1,570,946.8 km 2 , with 62 municipalities. The estimated population of the state was 3,807,921 inhabitants in 2010, with 74.2% living in urban zones and 25.8% in rural areas. Approximately 45% of the population lives in the state capital, Manaus. The state has a reduced coverage of highways and roads, with much of the displacement happening via river transportation. The state is densely covered by an evergreen rain forest, standing out the upland forests (terra firme forest), floodplains (várzeas) and flooded areas (igapós).

Data source
Snakebites are compulsorily recorded by the Brazilian Notifiable Diseases Surveillance System [Sistema de Informação de Agravos de Notificação (SINAN)] based on data report forms used in the investigation and follow-up of cases of venomous animals. The SINAN is a national electronic surveillance system that contains a variety of diseases in an integrated database that also includes snakebites [15]. We included all snakebite cases in the state of Amazonas reported to SINAN from January 1, 2007 to December 31, 2012. Snakebite treatment is provided free of charge only by the Brazilian Government and is not available for purchase in the private sector.

Exposure and outcome definitions
Reporting on snakebites' severity grading and outcomes (discharge or death) is required by SINAN and it is entered by healthcare providers at the time of case notification (which often happens after discharge). Clinical severity of snakebites in this work was classified according the Brazilian Health Ministry guidelines [17]: i) mild cases: local pain, local swelling and bruising for Bothrops and Lachesis bites or ptosis, slight blurred vision and mild or absent myalgia for Crotalus bites (rarely found in the study area) [21]; ii) moderate cases: local manifestations without necrosis, minor systemic signs (coagulopathy and bleeding, no shock) in Bothrops and Lachesis bites; myalgia and dark urine in may occur in Crotalus bites; iii) severe cases: lifethreatening snakebite, with severe bleeding, hypotension/shock and/or acute renal failure for Bothrops and Lachesis bites. In severe Lachesis bites, vagal symptoms may be present in about 14% of cases [30]. Severe Crotalus are those presenting with generalized myalgia, dark urine, oliguria or anuria. It is recommended to consider all confirmed Micrurus cases, although scarcely seen in the Amazonas, as potentially severe due to the risk of respiratory failure. For more details concerning the severity grading and treatment of snakebites used in Brazil see S1 File.
The variables analyzed were sex, age (in years), anatomical region of the bite, area of occurrence (rural or urban), work-related injury (yes or no), time elapsed between the bite and medical assistance (in hours), perpetrating species (Bothrops, Lachesis, Crotalus, Micrurus or nonvenomous snakes), severity grading (mild, moderate or severe), outcome (discharge or death), clinical features and antivenom administration data. All variables were checked by two independent researchers before analysis and further investigated for a possible association with severity and lethality as dependent variables.
Due to the potential severity of Lachesis confirmed bites, these few cases are generally considered clinically moderate or severe. However, in areas where there is overlap in the geographical distribution of Bothrops and Lachesis, such as in the Brazilian Amazon, the differential routine diagnosis when the snake has not been captured can be misleading and be associated with misclassification of the causing species.
In order to identify factors associated with snakebite severity and mortality, a nested case control study was used wherein patients evolving to severity or death were classified as cases and those with mild and moderate bites were included as controls.

Spatial distribution
A map was created with the software ArcMap 10.1 in ArcGIS 10.1 (ESRI, USA) using estimates of the mean incidence by municipality. The incidence rate was calculated dividing cases by the population of each municipality [31] by 100 thousand inhabitants. Spatial interpolation of snakebites' incidence was mapped using data from 62 municipalities by the Inverse Distance Weighting (IDW) method.

Statistical analysis
Only variables presenting completeness higher than 85% were analyzed. Duplicity was solved before data analysis. Data were analyzed using SPSS version 21.0 for Windows (SPSS Inc. Chicago, IL, USA). Proportions of severe cases and deaths were compared by Chi-square test (corrected by Fisher' test if necessary); differences were considered statistically significant for p<0.05. The crude Odds Ratio (OR) with its respective 95% Confidence Interval (95% CI) was determined considering severity and death as the dependent variables. Logistic regression was used for the multivariable analyses and the adjusted Odds Ratios with 95% CI were also calculated. All variables associated with the outcomes at a significance level of p<0.20 in the univariable analysis were included in the multivariable analysis. Statistical significance was considered if p<0.05 in the Hosmer-Lemeshow goodness-of-fit test.

Ethical clearance
This study was approved by the Ethics Review Board (ERB) of the Fundação de Medicina Tropical Dr. Heitor Vieira Dourado (approval number 872.520/2014), as well as by the ERB of the Health Surveillance Foundation of the Amazonas State. All data analyzed were anonymous. Since data were obtained exclusively from surveillance databases, the ERB gave a waiver of informed consent.

Study population
According to the official reporting systems, 9,191 snakebites were recorded in the Amazonas State from 2007 to 2012, resulting in an incidence rate of 52.8 cases per 100,000 person/year. There was a slight variation in the annual incidence rates during the study period (Fig 1).
Incidence was higher in 2009 (1,697 cases; 58.5 per 100,000 inhabitants) and 2012 (1,558 cases; 53.7 per 100,000) and lower in 2011 (1,436 cases; 49.5 per 100,000). All the variables retrieved from the original dataset presented completeness higher than 85% ( Table 1). Most of the snakebites occurred in males (7,257 cases; 79.0%). Regarding the area of occurrence, 70.2% were reported in rural areas. The most affected age group was between 16 and 45 years old (5,019 cases; 54.6%). A proportion of 55.0% of the snakebites were related to work activities. Regarding time elapsed from the bite until medical assistance, 68.4% of the cases received treatment within the first six hours after the snakebite, 13.7% within 6-12 hours and 17.9% with more than 12 hours after bite. Most of the snakebites occurred in the feet (58.8%) and legs (27.4%).
Most cases were mild (46.1%), followed by moderate (45.8%) and severe (8.1%) cases. There were 55 deaths due to snakebites in the study period, resulting in a 0.6% lethality rate. Lethality rates were 0.7% for Bothrops snakebites and 0.6% for Lachesis snakebites. There were no records of deaths due to Crotalus and Micrurus snakebites in the study period. According to age groups, lethality rate was 0.5% for the 15 years group, 0.5% for the 15 to 60 years group and 2.0% for the 60 years group.
Time elapsed from bite to antivenom administration at a health facility was 6.2 (SD±7.2) hours for mild cases, 7 (SD±7.1) hours for moderate cases, 9 (SD±8.1) hours for severe cases and 15.5 hours (SD±10.2) for cases evolving to death (Fig 2).
The  evolving to death did not receive antivenom. One per cent of accidents by Bothrops considered mild received less than the recommended dosage of specific antivenom, 3% of moderate cases received less than recommended and 52% of severe cases received less than expected. A total of 50% of accidents by Lachesis considered moderate received less than the recommended dosage and 81% of severe cases were underdosed. All accidents by Micrurus are reported as severe and 35% of cases received less than the recommended dose of antivenom. Lastly, for Crotalus injuries, 37.8% received less than the recommended dose of antivenom. Incidence rates were unevenly distributed across the Amazonas State, although there were snakebites cases reported from all the 62 municipalities. Mapping showed a large area with high incidence rates extending from the Northeast to the Central region of the state, including three hotspots located in São Gabriel da Cachoeira (in the Colombian border), in the Uarini/Alvarães municipalities area and in Novo Airão, where incidence rates were higher than 150 cases per 100,000 inhabitants/year. Smaller areas with high snakebite incidences were located in Rio Preto da Eva, in the vicinity of the capital city, and in Borba, in the upper Southeast region of the state (Fig 3). Incidence rates by municipality are presented in S1 Table. Risk factors for severity and mortality From the 9,191 snakebite cases reported in the Amazonas State in the study period, we excluded 102 cases of non-venomous snakebites and 422 cases without severity classification. We used the remaining 8,667 eligible cases to evaluate risk factors for severity and mortality. A total of 708 severe cases and 52 deaths were included as cases and 7,959 snakebites were included as controls (Fig 4).

Discussion
Few previous attempts have been made to quantify the burden of envenoming resulting from snakebites in the Brazilian Amazon. Our study confirms previous hospital based reports of a wide geographical distribution of snakebites in the state of Amazonas [9]. We provide a more accurate, up-to-date estimate of the scale of the problem in the Brazilian Amazon by using official surveillance data to estimate the disease burden due to snakebites and to identify independent risk factors for severity and mortality. In this study, systemic manifestations were present in 19.4% of patients, being the most common: hemorrhage, myolysis/hemolysis, paralysis and renal injury, in agreement with previous reports from the Amazon Region, where Bothrops atrox is the main causal agent [18,19,30].
The mean incidence rate in the study area was 52.8 cases per 100,000 person/year, reaching more than 150 cases per 100,000 person/year in several municipalities, comparable to   projections for sub-Saharan Africa and Oceania [1]. In the Brazilian Amazon, reaching health centers is difficult for riverine and indigenous populations as a result of distance and underdevelopment of means of transportation. If we account for a possible underreporting to official surveillance systems [32], the actual incidence of snakebites in this region could be among the highest in the world. We observed a large area with high incidence rates extending from the Northeast to the Central region of the state, where incidence rates were above 150 cases per 100,000 inhabitants/year. This area coincides with municipalities where Amerindian populations predominate, highlighting the vulnerability of these groups. A previous study found and association between snakebite incidence and deforestation [33]. Indeed, the hotspots identified in Novo Airão and Rio Preto da Eva, both in the Manaus region, and in Borba, correspond to areas with intense deforestation, and likely reflect the emergence of economical activities in this municipalities, namely farming and cattle raising. Most snakebites occurred in males, between 16 and 45 years, living in rural areas and mainly related to work activities. This profile has been reported in other regions in Brazil [13,16,34,35] and in other countries [36,37] and reaffirms the strong relationship between snakebites and rural economical activities in poor regions [38], thus, negatively affecting family economies in the area. Currently, there are no interventions for primary prevention of snakebites as an occupational hazards or to speed access to health services.
More than 30% of patients took more than six hours to receive medical assistance and such delay was an independent risk factor for severity and mortality. A delay to medical care was also identified as a risk factor for poor outcomes in India [23,39], Nepal [36,40] and Nigeria [27,28]. Patients from Amazonian communities depend mainly on river transportation to reach hospitals in urban seats, causing a delay in treatment. Moreover, many victims seek medical care after a considerable delay or fail to reach a hospital in time as a result of first seeking treatment from traditional healers [9]. The lethality rate of 0.6% observed in this work is higher than some previous national reports [7,35]. We hypothesize that implementing a rapid transport system for snakebite patients would significantly reduced lethality rates as has been reported in other settings [41]. Interestingly, occupational accidents were more likely to develop severe disease. This relationship between severity and occupational exposure was also observed in India, where rural workers bitten by snakes were more prone to death [37]. We can not precisely establish a direct association between snakebite severity and occupational exposure, but the distance to be walked from the workplace in farming areas to medical care could explain this risk association. In India, patients who walked more than 1 km after the snakebite, had a delay in care and a higher risk of death [22]. Reassurance and immobilization of the affected limb with prompt transfer to a medical facility are the cornerstones of the immediate care of snakebites [29,42].
In this work, age less than 15 years old was also an independent risk factor for severity, probably as a result of smaller body volumes [2]. This risk factor was also observed in India [22] and Turkey [43]. One half of deaths from snakebites in California affected children under 5 years of age [44].
Age 65 years was a risk factor for severity and mortality. This is relevant as literature scarcely reports elderly as being at a higher risk. In Bothrops and Crotalus snakebites reported in São Paulo, Brazil, patients 50 years-old had a higher mortality when compared to younger patients [45]. Taking into account that most snakebite complications, especially by Bothrops, are due to acute renal failure [46][47][48][49][50], the greater severity associated with the elderly may be related to higher prevalences of chronic comorbidities (i.e. hypertension and diabetes) that can lead to a predisposition to evolve to necrosis and acute kidney injury after a bite [51,52]. One especulates that comorbidities leading to vascular disease, such as hypertension and diabetes, commonly found in the elderly, may precipitate life-threatening complications after snakebite envenomings. In a case series from Brazil, eight of 309 patients (2.6%) developed a cerebrovascular event after Bothrops envenoming; five of the eight patients died despite therapy, and the remaining three had irreversible sequelae [53]. Cases of stroke related to Bothops envenomings have been reported in Brazil in the elderly, including in the Amazon region [54,55]. Population aging will likely affect snakebite epidemiology and its public health impact thus requiring preparation by local health services with a wider distribution of antivenom and life-sustaining supportive therapies besides health care workers' training in the management of envenoming in this neglected age group.
Our study reports significant underdosing of antivenom in the Brazilian Amazon as has been previously observed [32]. This has several possible explanations: we used an epidemiologic surveillance system to obtain our data where clinical and serum administration dosing information is collected hours or even days after the event, possibly reflecting a difficulty predicting the amount of antivenom needed upon admission; another possible explanation is the lack of appropriate training in antivenom dosing in medical personnel in remote settings; or possibly, unavailability of serum in rural areas were most of accidents occur. Underdosing may play a role in morbidity and mortality associated with snakebites in the region, exemplified by the fact that more than half of deaths lacked antivenom administration; this represents a missed opportunity to improve snakebite outcomes in the region.
Record keeping may have been influenced by the nature of the surveillance system. Some patients with mild bites in inaccessible areas may not be reported to hospitals and those evolving to severity may die on the way before reaching medical attention. In Brazil, antivenoms are prescribed exclusively by physicians in hospital settings. However, only a few patients bring the snakes for identification to the health unit, therefore the clinical-epidemiological classification is subject to misclassification, as is mostly based on clinical grounds. For instance, in the absence of the Lachesis or Bothrops-Lachesis antivenoms or in case of snake misclassification, health services often use Bothrops antivenom in the treatment of laquetic accident. However, previous studies in the same region have shown that Bothrops antivenom was ineffective to neutralize the coagulant activity of L. muta venom in humans, hence, contributing to severity [56]. In an uncertain proportion, completion of severity grading may occur retrospectively, according to the number of antivenom vials administered. Since patients are commonly seen initially at the emergency room, case notification mostly occurs after antivenom therapy and sometimes even after patient discharge. As a result, severity classification of snake envenoming by healthcare providers depends on the quality of this retrospective information. As all Micrurus envenomings are considered potentially severe by the Brazilian Ministry of Health guidelines, it was expected for this type of envenoming to be associated with severity. In fact, it was an expected result resulting from an biased association, rather than a real risk factor for poor outcome.
In conclusion: i) the incidence of snakebites in the Western Brazilian Amazon is among the highest in the world; ii) snakebites show a wide distribution in this region; iii) snakebites represent an occupational health problem for rural populations and public heath measures are needed to prevent occupational snakebites; iv) age 15 years and 65 years, work related bites and delayed medical treatment >6 hours were independently associated with the risk of developing severity and mortality; age 65 years and time to medical assistance >6 hours were independently associated with mortality. These features highlight the need for public health interventions aiming to decrease time to medical attention in vulnerable groups. In the Amazon, a major concern relates to the failure in antivenom distribution and storage. The lack of an adequate cold chain impairs antivenom distribution to rural areas and may result in loss of material. Moreover, there is a lack of systematic professional training on diagnosis, specific therapy and clinical management of complications [32]. Improvements in antivenom distribution and time to medical attention could have a significant impact in patient outcomes.