Epidemiological characterization of snakebite in Colombia, 2017–2023

Abstract

Background

Snakebite is a neglected public health problem worldwide, particularly in tropical countries such as Colombia. Objective: This study aims to describe the epidemiological characteristics of snakebite cases in Colombia between 2017 and 2023.

Methods

A cross-sectional observational study was conducted using data from the National Public Health Surveillance System (SIVIGILA). A total of 37,204 snakebite cases were analyzed after exclusions. Sociodemographic, geographic, and healthcare variables were examined using descriptive statistics for the 2017–2023 period. Municipal-level associations were assessed using regression models based on 2018 census data.

Results

Most cases occurred in young and adult men living in rural areas, with low socioeconomic status and enrolled in the subsidized health system. The highest incidence rates were observed in Amazonian departments such as Vaupés, Guaviare, and Amazonas. Although most patients received care within 24 h, delays were still identified in some cases. Municipal-level regression analysis (based on 2018 census data) showed that higher snakebite incidence at the municipal level was significantly associated with structural and demographic factors. For every 10-percentage-point increase in the proportion of floors made of earth, wood, or cement, the incidence rate was 47% higher (IRR: 1.47). Similarly, a 10-percentage-point increase in households without electricity was associated with an 18% higher incidence rate (IRR: 1.18), and a 10-percentage-point increase in rural population with a 17% higher incidence rate (IRR: 1.17).

Conclusions

Strengthening surveillance systems and prioritizing prevention strategies in vulnerable rural populations are essential to reduce the burden of snakebite in Colombia.

Supplementary Information

The online version contains supplementary material available at 10.1186/s41182-026-01040-3.

Introduction

Snakebite is an injury caused by the bite of a snake. When inflicted by a venomous species, venom may be injected into the victim, triggering local tissue damage and systemic pathophysiological alterations. The clinical severity of envenoming depends on multiple factors, including the snake species, venom composition and dose, bite characteristics, and host-related factors such as age, body size, and underlying health status [1, 2]. Colombia harbors approximately 8% of the world's snake diversity (around 300 species); however, only 18% are venomous, and at least 5 species pose a clinical risk to humans [3].

Depending on the snake genus, clinical manifestations may range from local pain and edema to severe complications including neurotoxicity, coagulopathy, necrosis, cardiovascular shock, acute kidney injury, and, in severe cases, multi-organ failure [3–5]. At the global level, snakebite has been recognized by the World Health Organization (WHO) as a neglected tropical disease, with a significant impact on global health: it is estimated that between 4.5 and 5.4 million people are bitten each year, of whom between 1.8 and 2.7 million develop clinical complications and between 81,000 and 138,000 die. In the Americas, approximately 57.500 cases of snakebite envenomation are recorded annually [6, 7].

Snakebite in South America shows converging epidemiological patterns, with a high concentration in rural areas, predominance among men of working age, and a strong association with agricultural activities. For example, in Paraguay, between 2007 and 2017, 2,501 cases were reported, with an average incidence of 3.6 per 100,000 inhabitants, concentrated primarily in the Eastern region. Cases mainly affected men in rural areas engaged in agricultural work, and higher case fatality rates were observed in areas with limited access to health services and antivenoms [8].

Brazil reports a substantial burden of snakebite, with national epidemiological studies consistently identifying Bothrops species as the leading cause of envenoming. The highest incidence has been reported in the North and Central-West regions, while cases predominantly affect rural populations, particularly individuals engaged in agricultural activities. These epidemiological characteristics are similar to those reported in other Latin American countries [9].

In Colombia, between 2008 and 2016, more than 37,000 cases were registered, with annual incidence rates ranging from 7.0 to 9.7 per 100,000 inhabitants, particularly in regions such as Amazonia and Orinoquia, where the genus Bothrops accounted for 64.5% of cases. Given its tropical and biodiverse conditions, Snakebite remains a major public health problem in Colombia, primarily affecting rural and vulnerable communities [10]. Since 2004, Colombia has maintained a mandatory surveillance system, the National Public Health Surveillance System (SIVIGILA), enabling the systematic monitoring of snakebite cases since reporting began in 2007 [11].

Previous eco-epidemiological studies have shown that snakebite in Colombia presents marked spatial and temporal heterogeneity. The burden of disease is concentrated in rural populations and varies according to climatic, environmental, and socioeconomic conditions, with the highest incidence reported in the Amazon and Orinoquia regions. These findings highlight the need for continuous epidemiological monitoring to identify changes in disease patterns and support prevention and control strategies [11].

Events and conditions related to venomous animals affect more than one billion people worldwide, particularly vulnerable populations experiencing high levels of poverty, limited access to basic services, and inadequate education. These diseases not only reflect social inequalities in health but also represent a barrier to human and economic development [12]. Despite progress, limited medical infrastructure in isolated areas and restricted access to antivenoms as part of treatment remain ongoing challenges in reducing mortality and long-term complications or sequelae [13, 14].

Therefore, this study aimed to characterize the epidemiological profile of snakebite in Colombia between 2017 and 2023 and to explore municipal-level structural factors associated with snakebite incidence using data from the National Public Health Surveillance System (SIVIGILA).

Materials and methods

Study design

A cross-sectional observational study was conducted using annual data on snakebite cases reported in Colombia's SIVIGILA, published as open-access data, covering the period from 2017 to 2023.

Participants

All snakebite cases reported in Colombia between 2017 and 2023 were included. A total of 37,585 snakebite envenomation cases were initially identified. For the purposes of this study, 381 cases that occurred outside the country were excluded, yielding a final analytical sample of 37,204 cases occurring within Colombia. The databases were organized annually and cleaned to ensure data quality, and a variable table was developed.

Data sources and variables

Snakebite cases were obtained through SIVIGILA notifications, which compiles information from Primary Data-Generating Units (UPGDs—Unidades Primarias Generadoras de Datos), composed primarily of healthcare institutions responsible for identifying and reporting public health events of interest. Although it constitutes an official epidemiological surveillance source, the data may be subject to underreporting and variability in notification quality.

Snakebite (recorded as accidente ofídico in the Colombian surveillance system) is classified as a mandatory reporting event, notified immediately on an individual basis. Information on this event is publicly available through SIVIGILA microdata [15]. Annual databases were downloaded from the SIVIGILA open-access portal, which applies standardized data quality protocols prior to public release, including deduplication and validation of records. For the purposes of this study, 381 cases reported as occurring outside Colombia were excluded. Databases were organized annually in Microsoft Excel and variables of interest were selected for analysis. Variable names and formats were standardized across annual databases before analysis. Date variables were reviewed for consistency, and records containing implausible dates were excluded only from analyses involving time intervals. Because the publicly available SIVIGILA databases are released after institutional validation and deduplication procedures and do not contain personal identifiers, no additional duplicate-record assessment was performed.

Records with typographical errors in date fields (e.g., implausible consultation dates) were identified and excluded from time interval calculations; these are reported by year in Table 3. Missing data were minimal; the only notable instance was the absence of socioeconomic stratum information in the 2017 SIVIGILA database, which is noted in Table 1. Variables extracted included sociodemographic characteristics (age, sex, ethnicity, area of residence, and socioeconomic stratum), as well as dates of notification, consultation, hospitalization, and death. Geographic information on both the patient’s municipality of residence and the municipality where the event occurred was also included.

Table 3.

Number of days from snakebite to consultation at a health facility, by year

Number of days from the bite to the consultation	2017	2018	2019	2020	2021	2022	2023
	Number of cases (%)
Number of people with a time interval ≤ 1 day	4,092 (82.5)	4,541 (84.2)	4,706 (84.5)	4,117 (84.8)	3,951 (83.7)	4,667 (83.9)	5,124 (83.3)
Number of people with a time interval ≥ 1 day	868 (17.5)	851 (15.8)	863 (15.5)	735 (15.2)	770 (16.3)	894 (16.1)	1025 (16.7)
Median	1	1	1	1	1	1	1
SD	2.8	3.2	2.8	3.1	2.7	3.0	3.0
Interquartile range	2	1	1	1	1	1	2
Min	1	1	1	1	1	1	1
Max	20	26	27	29	20	29	29
Data with typing errors in the dates	8	3	2	4	5	7	8

Source: Authors’ own elaboration based on SIVIGILA. SD: standard deviation; Min: minimum; Max: maximum. Percentages calculated by year

Table 1.

Sociodemographic characteristics by study year

Characteristic	Description	2017	2018	2019	2020	2021	2022	2023	Total
									N (%)
Age	Early childhood (0–5 years)	181	222	220	170	185	191	230	1399 (3.8)
	Childhood (6–11 years)	413	409	439	371	342	369	463	2806 (7.5)
	Adolescence (12–17 years)	600	664	605	596	530	566	604	4165 (11.2)
	Youth (18–28 years)	1101	1135	1204	1129	1064	1258	1271	8162 (21.9)
	Adulthood (29–59 years)	2104	2312	2429	2017	2000	2447	2716	16,024 (43.1)
	Old age (≥ 60 years)	561	650	672	569	600	730	865	4648 (12.5)
Sex	Men	3665	3893	3966	3565	3439	4115	4403	27,046 (72.7)
	Women	1295	1499	1603	1287	1282	1446	1746	10,158 (27.3)
Area of residence	Urban	1179	1314	1435	1223	1326	1697	1781	9955 (26.8)
	Rural area	3781	4078	4134	3629	3395	3864	4368	27,249 (73.2)
Health insurance system*	Subsidized	3825	4148	4229	3434	3623	4366	4841	28,466 (76.5)
	Contributory	838	797	833	816	613	717	741	5355 (14.4)
	Uninsured	226	353	410	455	358	339	286	2427 (6.5)
	Other	71	94	97	147	127	139	281	956 (2.6)
Ethnicity	Indigenous	528	484	507	441	413	467	551	3391 (9.1)
	Rom, gypsy	10	13	24	12	12	9	4	84 (0.2)
	Raizal	3	7	5	5	5	1	2	28 (0.1)
	Palenquero	2	3	–	1	2	3	2	13 (0.1)
	Black or mixed-race (negro o mulato)	351	299	355	320	273	294	319	2211 (5.9)
	None	4066	4586	4678	4073	4016	4787	5271	31,477 (84.6)
Socioeconomic stratum**	1	**	3320	4299	3738	3696	4468	5057	24,578 (83.1)
	2	**	559	748	691	790	869	907	4564 (15.4)
	3	**	37	57	41	54	44	66	299 (1.2)
	4	**	5	2	4	6	8	5	30 (0.1)
	5	**	6	3	2	2	–	3	16 (0.1)
	6	**	25	13	10	14	11	5	78 (0.3)

*Social insurance "other" includes the special, undetermined, and exception regimes (as defined by the Colombian General Social Security Health System)

**Missing socioeconomic stratum data due to unavailability of information in the 2017 SIVIGILA database. Percentages calculated over the six-year period. For clarification, in Colombia, stratum 1—designated "Low-low"—classifies dwellings with the greatest infrastructure and environmental deficiencies, corresponding to households with the fewest economic resources

This study incorporated sociodemographic variables relevant to the Colombian context, including the type of enrollment in the General Social Security Health System (SGSSS—Sistema General de Seguridad Social en Salud), which classifies the population into insurance regimes based on payment capacity: the contributory regime (for individuals with formal employment or stable income) and the subsidized regime (for individuals without payment capacity, state funded). Additional variables included socioeconomic stratum (an official classification from 1 to 6 based on housing conditions and income), area of residence (urban or rural), and ethnic affiliation (including indigenous, Afro-descendant, and Raizal populations, among others).

Data on housing characteristics and population used in the statistical models were drawn from updated population projections and the 2018 National Population and Housing Census (Censo Nacional de Población y Vivienda) produced by the National Administrative Department of Statistics (DANE—Departamento Administrativo Nacional de Estadística) for each year. The dependent variable was the number of snakebite cases per municipality (for the year 2018). Independent variables included housing characteristics (specifically, the proportion of dwellings with floor materials such as earth, cement, or wood, and the proportion of dwellings without access to electricity) derived from DANE's updated population projections and the National Census, as well as selected sociodemographic variables available at the municipal level in Colombia. The use of 2018 census data for the regression analysis was justified by the fact that Colombia’s most recent National Population and Housing Census was conducted in 2018, making this the only year for which municipality-level structural variables are available. It is acknowledged that these data represent a cross-sectional snapshot and may not fully reflect conditions across the entire 2017–2023 descriptive period; this limitation is addressed in the Limitations section. The regression model should therefore be interpreted as describing associations between municipal structural conditions measured in 2018 and snakebite incidence in that year, not as explaining variation across the full study period (supplementary material).

Statistical analysis

A descriptive statistical analysis was performed, including absolute and relative frequencies for sociodemographic variables, as well as calculations of annual incidence and average incidence of snakebite. Measures of central tendency and dispersion were estimated for the time elapsed between the bite and healthcare-seeking behavior. The time interval was calculated by subtracting the date of the event from the date of consultation at a health facility, expressed in days. DANE population projections for each year of the 2017–2023 period were used as denominators for calculating incidence and case fatality rates.

For spatial analysis, snakebite cases were aggregated by municipality of residence for the period 2017–2023. The average annual municipal incidence was calculated by dividing the mean annual number of cases by the mean municipal population estimated from DANE population projections for the same period and multiplying by 100,000 inhabitants. Municipal incidence estimates were linked to the official municipal administrative boundaries of Colombia using DANE municipal codes and represented cartographically to visualize the geographic distribution of snakebite incidence.

A bivariate analysis was conducted using Pearson correlation coefficients to explore the relationship between average snakebite incidence and sociodemographic and housing condition variables at the municipal level. Statistical significance was set at p < 0.05. The association between municipal-level factors and the number of cases for 2018 was assessed using generalized linear models for count data with a logarithmic link function. Population size was incorporated as an offset (log of population) to model incidence rates.

Independent variables included the proportion of dwellings with floor materials classified by DANE as earth, wood or cement, the proportion of dwellings without access to electricity, and the proportion of the population residing in rural areas. To facilitate interpretation, all proportional variables were scaled to represent increments of 10 percentage points. Model coefficients were subsequently exponentiated to obtain incidence rate ratios (IRRs), allowing results to be interpreted as relative changes in the event rate.

Regression coefficients were exponentiated to obtain IRRs with their corresponding 95% confidence intervals. Statistical significance was set at p < 0.05.

A Poisson regression model was initially fitted. Overdispersion was subsequently assessed using the likelihood ratio test for the dispersion parameter (α). As significant overdispersion was detected, a negative binomial regression model was selected as the final specification. Multicollinearity among explanatory variables was assessed using variance inflation factors (VIF). All VIF values were below 2, indicating no evidence of problematic collinearity. Municipalities with zero registered snakebite cases (n = 414; 36.9% of all municipalities) were retained in the analysis, as they represent valid observations within the distribution of snakebite occurrence across Colombia. Excluding these municipalities could bias model estimates. The negative binomial distribution appropriately accommodates overdispersed count data, including zero counts.

Statistical analyses were performed using Stata version 18.0 (licensed through Universidad Rey Juan Carlos), and the average incidence map was generated using QGIS v3.42 (open-source software).

Ethical considerations

The information presented in this study corresponds to the notification of public health events of interest, publicly accessible through SIVIGILA. Accordingly, the data are considered to pose no risk in accordance with Resolution 08430 of 1993 issued by the Colombian Ministry of Health. Although this study was conducted using secondary, publicly available data sources, it received a favorable response from the International Ethics Committee of Universidad Rey Juan Carlos on April 24, 2025.

Results

Between 2017 and 2023, a total of 37,204 snakebite cases were registered in Colombia. Regarding sociodemographic characteristics, most cases occurred in adults (43.1%) and young people (21.9%), with a mean age of 34.4 years (SD: 19.4; min: 1; max: 97). In terms of sex, men accounted for the largest proportion of cases (72.7%). With respect to area of residence, most cases occurred in dispersed rural areas (73.2%). Regarding enrollment in the SGSSS, many individuals were affiliated with the subsidized regime (76.5%), while a smaller proportion belonged to the contributory regime (14.4%) or were uninsured (6.5%) (Table 1 and Fig. 1).

Fig. 1.

Distribution of snakebite cases during the study period (2017–2023) by life cycle stage, sex, and area of residence.

Source: Authors’ own elaboration based on SIVIGILA and DANE data. Description of age groups: Early childhood (0–5 years), Childhood (6–11 years), Adolescence (12–17 years), Youth (18–28 years), Adulthood (29–59 years), Old age (≥ 60 years)

In terms of ethnic self-identification, most cases (84.6%) did not report belonging to any ethnic group. Among those who did self-identify with an ethnic group, the indigenous population was the most frequently represented (9.1%), followed by individuals identifying as Black or mixed-race (negro o mulato) (5.9%). Regarding socioeconomic stratum, many cases corresponded to individuals in stratum 1 (83.1%) (Table 1).

The highest incidence rates were concentrated in departments such as Vaupés, Guaviare, and Amazonas, with Vaupés recording figures above 120 per 100,000 inhabitants in most years, reaching a maximum of 163.3 in 2017. Guaviare showed notable peaks in 2020 (111.3) and 2021 (146.6), while Amazonas reached its highest value in 2022 (62.2) (Table 2).

Table 2.

Snakebite incidence by department and study year (2017–2023)

Source: Authors’ own elaboration based on SIVIGILA and DANE data. Incidence per 100,000 inhabitants, calculated using DANE population data by year (based on figures updated as of August 8, 2025). Source: SIVIGILA Portal

Departments such as Cundinamarca, Quindío, and Valle del Cauca presented incidence rates between 1.8 and 3.6, falling below the national average. In contrast, departments such as San Andrés and Bogotá showed very low or negligible incidence rates throughout the period. At the national level, incidence remained relatively stable between 2017 and 2019 (10.5–11.3 per 100,000), declined markedly during 2020–2021 across most departments (reaching a minimum of 9.2 in 2021), and subsequently increased broadly during 2022–2023, surpassing pre-pandemic levels and reaching the highest value of the study period (11.8) in 2023 (Table 2).

Figure 2 illustrates marked spatial heterogeneity in the average municipal incidence of snakebite in Colombia during the 2017–2023 study period. The highest incidence rates were concentrated in municipalities located primarily within the Amazon, Orinoquía, Pacific, and parts of the Caribbean regions, whereas the lowest incidence rates, as well as municipalities reporting no cases, were predominantly observed across large areas of the Andean region. The clustering of municipalities with high incidence rates indicates a well-defined geographic pattern rather than a random distribution of the event. This pattern may reflect, among other factors, regional differences in climatic conditions and the geographic distribution of the country's principal medically important venomous snake species.

Fig. 2.

Average snakebite incidence during the study period (2017–2023) at the municipal level.

Source: Own elaboration using QGIS. Average incidence during the study period per 100.000 inhabitants. Calculated using DANE population data by municipality (average total snakebite cases per municipality 2017–2023 / average population per municipality 2017–2023 × 100.000 inhabitants)

Figure 3 presents the annual snakebite incidence and case fatality rate in Colombia between 2017 and 2023. Regarding incidence, moderate variations were observed, with a minimum value of 9.2 per 100,000 inhabitants in 2021 and a maximum of 11.8 in 2023. Despite these fluctuations, levels remained relatively stable. Regarding case fatality, the highest value was recorded in 2020 at 0.89%, while the lowest was in 2019 at 0.45%.

Fig. 3.

Snakebite incidence and case fatality rate in Colombia between 2017 and 2023.

Source: Authors’ own elaboration based on SIVIGILA and DANE data. Incidence per 100,000 inhabitants and case fatality rate, calculated using DANE population data by year. The figure uses two vertical axes to represent variables measured on different scales. The left axis (blue line) corresponds to annual snakebite incidence per 100,000 inhabitants; the right axis (orange line) corresponds to the annual case fatality rate (%)

Table 3 shows the time in days elapsed between the snakebite and consultation at a health facility. The majority of individuals sought care within the first 24 h, with a median of 1 day across all years analyzed (IQR: 1–2 days). However, the proportion of individuals taking more than one day to seek medical attention fluctuated moderately across the period (15.2%–17.5%), with a modest increase toward the end of the study period (2023). The maximum time from bite to consultation indicates that some individuals sought care up to 29 days after being bitten, recorded for the years 2020, 2022, and 2023 (Table 3).

The bivariate analysis identified significant associations between snakebite occurrence and several sociodemographic and housing variables (p < 0.001). A Poisson regression model was initially fitted; however, significant overdispersion was detected, leading to the use of a negative binomial (NB) regression model. Model fit was substantially better for the NB model (AIC = 4986.24; BIC = 5011.36) than for the Poisson model (AIC = 8299.51; BIC = 8319.60). The final NB model was statistically significant (LR χ² = 342.76; p < 0.001). Multicollinearity assessment showed no evidence of problematic correlations among the explanatory variables included in the final model (all VIF values < 2). Municipalities reporting zero snakebite cases were retained in the analysis.

In the negative binomial regression model, adjusted for municipal population size through an offset term (log_pop), several housing and territorial characteristics were significantly associated with municipal snakebite rates. For every 10-percentage-point increase in the proportion of dwellings with wood, earth, or cement floors, the municipal snakebite rate was 47% higher (IRR = 1.47; 95% CI 1.40–1.55). Similarly, 10-percentage-point increases in the proportion of dwellings without electricity and in the rural population were associated with 18% (IRR = 1.18; 95% CI 1.12–1.24) and 17% (IRR = 1.17; 95% CI 1.12–1.22) higher snakebite rates, respectively. The model was globally significant (LR χ² = 342.76; p < 0.001), indicating that these municipal-level characteristics were significantly associated with variation in snakebite rates across Colombia (Table 4).

Table 4.

Structural factors associated with the municipal snakebite rate in Colombia, 2018: negative binomial regression model

Variable	IRR	95% CI	p value
Proportion of homes with flooring materials such as wood, earth, cement	1.47	1.40–1.55	< 0.001
Proportion of homes without electricity	1.18	1.12–1.24	< 0.001
Proportion living in rural areas	1.17	1.12–1.22	< 0.001

LR χ² = 342.76; p < 0.001 Pseudo R² = 0.064 α = 1.15

All proportions were scaled in increments of 10 percentage points to facilitate interpretation

Discussion

Several studies recognize that snakebite constitutes a global public health problem, particularly severe in tropical and subtropical regions, yet rarely prioritized by health systems. The true burden of this condition is underestimated due to factors such as the lack of systematized data, underreporting, limitations in data collection, and poor accessibility to health services, especially in rural and highly vulnerable social contexts [14, 16, 17, 23]. The burden of snakebite is particularly high among men living in rural areas, where the exposed population is predominantly engaged in agricultural activities and belongs to lower socioeconomic groups [16–21]. The results of the present study confirm this pattern in Colombia, where the largest proportion of cases is concentrated among young and adult men residing in dispersed rural areas, predominantly enrolled in the subsidized health insurance regime and belonging to low socioeconomic strata.

International evidence reinforces these findings. In Sri Lanka, the population-based study by Jayawardana et al. demonstrated a high prevalence of snakebites among adult men engaged in agriculture, highlighting occupational context and living conditions as key vulnerability factors [18]. Similarly, another study in India reported incidence and mortality figures exceeding official estimates, as well as a notable economic and social impact on victims and their families, underscoring the need to improve the accessibility and affordability of antivenom treatment [19].

These international precedents support the relevance of targeting preventive and care interventions toward productive rural populations, as suggested by the findings observed in Colombia.

Other studies have reported that agricultural workers, livestock farmers, and rural residents, particularly men, are more frequently exposed to snakebite and that housing conditions, walking barefoot, and inadequate lighting have been associated with increased snakebite risk [5, 19, 22]. Consistently, the negative binomial regression model applied in this study identified statistically significant associations between municipal snakebite rates and the proportion of dwellings with earth, wood, or cement floors, the absence of electricity, and a higher proportion of rural population. One possible contextual interpretation is that these structural conditions may reflect greater proximity to natural ecosystems and higher likelihood of occupational or domestic exposure to snakes, though the ecological and cross-sectional nature of this analysis precludes causal inference. Furthermore, these findings are consistent with those of a survey administered to 116 individuals in the departments of Antioquia and Chocó, in which the majority of participants reported living in dwellings with floors made of cement, brick, wood, or similar materials, and nearly half reported structural or environmental conditions that increase the risk of snake encounters, such as cracks, openings, lack of windows, or the presence of tall grass [22].

These findings suggest that intersectoral approaches addressing structural and environmental conditions in rural municipalities may contribute to snakebite prevention in areas with persistently high incidence. However, these results should be interpreted within the ecological framework of the present study. The municipal-level associations identified are more likely to represent contextual socioeconomic and territorial vulnerability than direct causal effects of any individual structural characteristic [17, 20].

In Nicaragua, a mapping study of snakebite epidemiology showed that incidence varies significantly according to environmental, demographic, and socioeconomic factors. Geographic regions exert a notable influence, with the Eastern Coast presenting the highest rates and the Western Highland Coast the lowest. Additionally, a higher proportion of rural population and male predominance were positively associated with incidence, while a greater proportion of individuals under 15 years of age reduced the risk in the adjusted model [20].

The marked concentration of snakebite incidence observed in municipalities of the Amazon and Orinoquia regions is consistent with previous eco-epidemiological analyses conducted in Colombia. These regions are characterized by extensive forest ecosystems, high diversity of medically important venomous snakes, dispersed rural settlements, and productive activities such as agriculture, livestock farming, fishing, and extractive activities, all of which increase the likelihood of human–snake contact. In addition, geographic isolation and limited access to health services may contribute to delayed treatment and underreporting, reinforcing the disproportionate burden observed in these territories [11, 17].

These results reinforce the utility of spatial and municipal-level analyses for identifying priority territories, which in the Colombian context would allow for more efficient targeting of prevention and surveillance strategies in higher-risk areas.

The global analysis by Longbottom et al. provides a relevant framework by integrating ecological, health, and species distribution data to identify regions with the greatest risk and least access to health services, concluding that hundreds of millions of people live in areas with venomous snake presence and inadequate treatment coverage [17]. In this regard, the findings of the present study support the need to use information generated by surveillance systems such as SIVIGILA to improve territorial planning and resource allocation, prioritizing rural municipalities with high incidence rates and unfavorable structural conditions.

Regarding the time intervals between snakebite and healthcare-seeking, international evidence consistently shows that survival and the reduction of severe sequelae depend directly on the time elapsed between the bite and the initiation of treatment. Each hour of delay increases the probability of irreversible damage, such as tissue necrosis, renal failure, or respiratory paralysis [14, 23]. Despite this limitation, although the median consultation time remained at one day and the majority of cases sought care within the first 24 h, a proportion of individuals presented with delays exceeding one day, ranging from 15.2% to 17.5%. Because the available variable measured the interval between snakebite and consultation only in days rather than hours, these findings should be interpreted as an indirect indicator of healthcare access. Although representing a minority, delayed consultation remains relevant from a public health perspective and may indicate opportunities to strengthen community education strategies and improve geographic and functional access to health services in rural areas, with the aim of reducing avoidable delays in care.

Finally, while recent literature has broadly documented the long-term physical and psychological sequelae associated with snakebite [22, 24], the present study lacked detailed clinical information to analyze these outcomes. Nevertheless, the identification of territories and populations with high incidence constitutes a fundamental input for guiding future research and strengthening epidemiological surveillance, through the incorporation of clinical and treatment variables that would allow for a comprehensive assessment of the overall burden of snakebite.

Taken together, the study findings support the implementation of comprehensive strategies centered on territorial targeting, improvement of structural and environmental conditions in rural settings, strengthening of surveillance and data quality, and the design of preventive interventions directed at the most exposed population groups, as key steps toward reducing the burden of this neglected tropical disease in Colombia.

Conclusions

Snakebite constitutes a significant public health problem in Colombia, primarily affecting rural and vulnerable populations in tropical regions. Its incidence was associated with socioeconomic, environmental, and geographic characteristics, with young and adult men residing in rural areas being the most affected group. Between 2017 and 2023, incidence remained within a relatively narrow range at the national level, with a temporary decline during 2020–2021 followed by a broad recovery in 2022–2023, and persistently high values in departments of the Amazonia and Orinoquia regions, particularly in low-density municipalities.

The results showed significant associations between municipal snakebite incidence and factors such as rurality, absence of electricity, and housing conditions, underscoring the importance of addressing this condition from a territorial and intersectoral perspective. Although most cases accessed health services within the first 24 h, a relevant proportion presented delays in consultation, highlighting the need to strengthen timely access to care in rural areas and to investigate the factors associated with such delays.

Taken together, the study findings provide useful evidence to guide preventive interventions, strengthen epidemiological surveillance, and inform public health planning directed at the territories and populations at greatest risk, contributing to the design of more equitable and efficient strategies for reducing the impact of snakebite in Colombia.

Limitations

The main limitations of this study include the potential underreporting of cases, delays in snakebite notification, and differences in reporting quality across regions. The results depend largely on the accuracy with which Primary Data-Generating Units report information. Additionally, given the ecological and cross-sectional nature of the municipal-level regression analysis, the associations identified should be interpreted as contextual rather than causal.

Additionally, other key variables—such as the offending species, antivenom use, and clinical conditions—could not be analyzed, as they are not included in the data publicly available through SIVIGILA.

The regression analysis relied on municipal structural variables obtained from the 2018 National Census, which represent a single cross-sectional measurement and may not fully capture changes occurring throughout the entire 2017–2023 study period.

Finally, a further limitation is that time intervals could not be calculated in hours, as the available information was recorded in days only, which precluded a more granular level of analysis.

Author contributions

MCRP carried out conceptualization, formal analysis, data curation, visualization, and writing of the original draft. MDRG, JAL, and MJO were involved in manuscript review and editing and critical revision of the results and contributed key insights to the Discussion section. Juan Antonio López also contributed to the review of the statistical models. LG performed geospatial analysis and map generation (QGIS), provided support in data analysis, and contributed key elements to the Discussion section. All authors read and approved the final manuscript.

Funding

This study did not receive any funding.

Data availability

The data used in this study are publicly available from open-access sources. Epidemiological data on snakebite were obtained from the National Public Health Surveillance System (SIVIGILA), managed by the Instituto Nacional de Salud (INS), and are available at https://portalsivigila.ins.gov.co/Paginas/Buscador.aspx. Population data were obtained from the National Administrative Department of Statistics (DANE) and are available at https://www.dane.gov.co/index.php/estadisticas-por-tema/demografia-y-poblacion/proyecciones-de-poblacion. No restrictions apply to the availability of these data.

Declarations

Ethics approval and consent to participate

This study was conducted using open-access, secondary data from the National Public Health Surveillance System (SIVIGILA) of Colombia. Although the use of secondary, anonymized, and publicly available data does not require ethical oversight under standard regulations, this study was reviewed and approved by the Ethics Committee of the Universidad Rey Juan Carlos (URJC). No individual informed consent was required, as no identifiable personal data were collected or processed.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.