Accidental and intentional poisonings in Colombia (2015–2023): Social and territorial inequalities affecting public health

Abstract

Objective

To analyze the epidemiological and spatial differences in accidental and intentional poisonings in Colombia between 2015 and 2023, with an emphasis on social and territorial inequalities.

Methodology

An observational, retrospective, and analytical study was conducted using records from the Public Health Surveillance System (Sivigila, event 365). Cases classified as accidental and intentional (suicide) poisonings were included. Frequencies, incidence rates, and associations were estimated using chi-square tests and odds ratios (OR) with 95% CI. Additionally, spatial analysis was performed using the Global Moran’s Index and LISA to identify patterns of autocorrelation and geographic clústers. Data processing was performed in RStudio (v4.4.3).

Results

A total of 237,926 cases were analyzed, of which 70,405 were accidental and 167,521 were intentional. Accidental poisonings were primarily associated with other chemicals and pesticides, while intentional poisonings were dominated by medications, revealing distinct patterns depending on the substance group involved. Accidental poisonings were more prevalent among children and showed a heterogeneous geographic distribution, while intentional poisonings predominated among adolescents and adults. Significant associations were found with gender, age, marital status, type of health insurance, and ethnic group (p < 0.05). Men were more likely to experience accidental poisonings (OR = 2.53), while adolescents had a higher risk of intentional poisoning compared to children (OR = 137.4). Spatial analysis revealed positive autocorrelation in intentional poisonings (Moran’s coefficient = 0.20; p = 0.017), with clústers in Cauca and Tolima, while accidental poisonings did not show a significant overall spatial pattern.

Conclusion

Chemical poisonings in Colombia exhibit distinct patterns depending on intent, with clear social, demographic, and territorial disparities. These findings highlight the need to strengthen prevention strategies by regulating highly hazardous pesticides, especially in areas where agricultural activities predominate and these products are more widely available, as well as to promote the safe storage of medications and mental health interventions with a territorial and life-cycle approach.

1. Introduction

Chemical poisoning is a growing global public health problem. Each year, approximately one million people die by suicide through the ingestion of toxic substances. According to the World Health Organization, about 20% of suicides worldwide were due to pesticide poisoning [1], [2]. At the same time, alcohol and tobacco impose an even greater burden, with 3 million and 8.7 million deaths, respectively, while the use of psychoactive drugs such as opioids has risen steadily over the past decade [3].

Death by poisoning is one of the most common causes identified by forensic services, despite being potentially preventable with timely diagnosis and treatment [4], [5]. People are exposed to chemicals in their daily lives at home, at work, and in public spaces and it is estimated that chemical poisonings are four times more common in low- and middle-income countries than in high-income countries [6], [7].

In Colombia, between 2008 and 2015, more than 200,000 cases of chemical poisoning were reported to the Public Health Surveillance System (Sivigila), with intentional poisoning for suicidal purposes being the leading cause of reports, followed by accidental poisoning, which primarily affected children under 5 years of age [8], [9]. The most common clinical signs and symptoms include nausea, abdominal cramps, and liver, kidney, and cardiorespiratory damage, as well as seizures [10], [11].

In this context, analyzing the differences between accidental and intentional poisonings in Colombia not only helps assess the scale of a public health problem but also identifies social and territorial determinants to guide interventions in chemical safety, pesticide control, and suicide prevention.

2. Methodology

This study was designed using a quantitative, observational, retrospective, and analytical approach. It was based on an analysis of records of acute chemical poisonings reported by Colombia’s territorial entities between 2015 and 2023.

The source of information was a secondary database provided by Colombia National Institute of Health (INS), obtained through an information-sharing agreement between the INS and a full-time faculty member at the University of Córdoba. The database was constructed using records from the National Public Health Surveillance System (Sivigila), which collects information on events of public health interest through a standardized mandatory reporting system implemented by the country health care providers and local government entities.

Records corresponding to the event coded as 365 in Sivigila were used; this code corresponds to acute chemical poisonings. In Colombia, this event encompasses eight groups of chemical substances subject to epidemiological surveillance: medications, pesticides, psychoactive substances, other chemical substances, metals, solvents, gases, and methanol. For this study, only cases classified under codes 2 (accidental exposure) and 3 (intentional exposure for suicidal purposes) were included.

The information recorded in Sivigila is collected in accordance with the guidelines set forth in the Event 365 notification form and its corresponding surveillance protocol. The cause of the poisoning is determined by the National Institute of Health through the epidemiological surveillance process, which includes mandatory case reporting, field epidemiological investigation when required, clinical evaluation, toxicological analysis of samples when available, and classification of the causative agent according to the group of chemicals involved and the circumstances of exposure. This process enables the standardization and validation of the information used for national surveillance of acute chemical poisonings.

The analytical framework took into account the classic epidemiological dimensions of time, place, and person. The analysis examined sociodemographic variables (age, gender, marital status, health insurance status, and ethnicity), characteristics of the chemical agent (substance group), and details of exposure (type).

Incidence rates were calculated using the number of cases recorded each year as the numerator and the official population projections from the National Administrative Department of Statistics (DANE) for the period 2015–2023 as the denominator.

The statistical analysis was conducted in three stages. First, descriptive methods were used, including absolute and relative frequency distributions for qualitative variables. Next, independence tests (chi-square) were applied to identify significant associations between the type of poisoning (accidental vs. intentional), the substance group, and sociodemographic variables, with a 95% confidence level.

In addition, odds ratios with their respective 95% confidence intervals were estimated to quantify the magnitude of the associations identified in selected comparisons, such as gender, health insurance status, and age group (childhood vs. adolescence and old age).

Finally, an inferential spatial analysis was conducted by calculating the Global Moran’s Index and the LISA (Local Indicators of Spatial Association) statistic to identify patterns of autocorrelation and spatial clusters in poisoning rates by department. This analysis allowed for a comparison of the territorial distribution of accidental and intentional events and the identification of significant spatial differences.

All data processing and statistical analyses were performed using Microsoft Excel® and RStudio (version 4.4.3), utilizing the tidyverse, epitools, spdep, sf, and tmap packages.

3. Results

The database of chemical poisonings in Colombia for the period 2015–2023 was analyzed, comprising a total of 237,926 cases, of which 70,405 were accidental exposures and 167,521 were intentional exposures.

When breaking down the distribution of cases by type of substance involved, differences were observed between accidental and intentional exposures. In accidental poisonings, “other chemicals” and pesticides were the predominant causes, followed by medications, indicating a greater variety of agents involved in this type of event.

In contrast, intentional exposures were concentrated primarily on medications, followed by pesticides; together, these groups accounted for the largest proportion of cases recorded during the study period. The other categories, meanwhile, accounted for a considerably smaller proportion of cases.

In general terms, intentional poisonings were heavily concentrated in a small number of substance groups, particularly medications. In contrast, accidental poisonings showed a more heterogeneous distribution, with a more balanced representation of different chemical agents. This pattern was confirmed by the association analysis, which revealed a statistically significant relationship between the substance group and the type of poisoning (χ² = 61,125; df = 7; p < 0.001), with a moderate magnitude of association (Cramer’s V = 0.51). In accidental poisonings, exposure to other chemicals and pesticides predominated, in addition to solvents, gases, metals, methanol, and psychoactive substances, whereas intentional poisonings were primarily associated with medications, reflecting a pattern of differential exposure depending on the intent of the event (See Table 1).

Table 1.

Number of poisoning cases by substance group for accidental and intentional exposure, Colombia, 2015–2023.

Substance Group	Accidental Poisonings	Intencional Poisonings
Medications	14772	110282
Pesticides	17228	38458
Other chemicals	22274	14589
Psychoactive substances	3422	1802
Solvents	6587	1637
Gases	4474	362
Methanol	602	333
Metals	1046	58
Total	70405	167521

When examining the distribution of cases by life stage, childhood accounted for the highest frequency of accidental poisonings, representing more than half of all cases, while intentional exposures were rare in this group.

In contrast, starting in adolescence, a shift in the epidemiological pattern was observed, with a predominance of intentional poisonings—a trend that continued into young adulthood and adulthood. A high proportion of the total cases occurred in this latter stage, with both types of exposure occurring simultaneously, although intentional cases were more common.

Finally, in old age, a more balanced distribution was observed between accidental and intentional poisonings, suggesting a smaller distinction between the types of exposure compared to younger age groups (See Fig. 1).

Fig. 1.

Distribution of accidental and intentional poisoning cases by life stage in Colombia, 2015–2023.

When studying the distribution of cases according to sociodemographic characteristics, differences were identified between accidental and intentional poisonings.

With regard to gender, a higher proportion of women was identified in intentional cases, while the male population predominated in accidental poisonings. Regarding marital status, single individuals predominated in both groups; however, this proportion was more pronounced in accidental poisonings. In contrast, people living in common-law relationships (who lived together without being formally married) accounted for a relatively higher proportion of intentional poisonings, while the other marital categories (married, divorced, and widowed) accounted for a smaller and stable proportion.

With regard to the type of health insurance, both types of exposure showed a higher incidence in the subsidized and contributory systems. In Colombia, the subsidized plan is intended for people with lower incomes and is funded by the government, while the contributory plan is for people who make contributions to the health system through premiums. The remaining categories included uninsured individuals, those enrolled in exceptional plans (such as certain members of the Armed Forces, the National Police, and the teaching profession), special plans, and cases with undetermined information.

The distribution by ethnicity was largely homogeneous, with a higher proportion of the “other” group in both types of poisoning. The other ethnic categories accounted for low numbers, with no substantial variations between accidental and intentional exposures.

It is worth noting that the chi-square test revealed a statistically significant association between the type of poisoning (accidental and intentional) and all the sociodemographic variables analyzed, including gender, marital status, type of health insurance, and ethnicity (p < 0.05 in all cases), indicating that the distribution of these characteristics differs according to the type of exposure. (accidental or intentional). (See Table 2).

Table 2.

Distribution of accidental and intentional poisonings by sociodemographic variables. Colombia, 2015–2023.

Variable	Category	Accidental Poisonings		Intentional Poisoning		P value
		Cases	%	Cases	%
Gender	Female	32444	46,08	114054	68,08	< 0.05
	Male	37961	53,92	53467	31,92
Civil Status	Single	53970	76,66	114987	68,64	< 0.05
	common-law marriage	8182	11,62	33680	20,10
	Married	7127	10,12	14209	8,48
	Divorced	438	0,62	2503	1,49
	Widower	688	0,98	2142	1,28
Type of insurance	Subsidized	36057	51,21	78328	46,76	< 0.05
	Contributory	28043	39,83	75314	44,96
	Uninsured	3292	4,68	6792	4,05
	Exception	1339	1,90	3920	2,34
	Undetermined	843	1,20	1850	1,10
	Special	831	1,18	1317	0,79
Ethnic group	Other	66285	94,15	158891	94,85	< 0.05
	Afro-colombian	2566	3,64	4748	2,83
	Indígenous	1246	1,77	3263	1,95
	Rom, Romani	173	0,25	394	0,24
	Raizal	94	0,13	183	0,11
	Palenquero	41	0,06	42	0,03

Additionally, the strength of association was quantified using odds ratios (OR), based on selected comparisons of epidemiological relevance. These comparisons allowed us to estimate how likely a specific group was to experience intentional poisoning compared to another group, beyond the proportional differences already observed in the overall distribution.

The results indicated that men were 2.5 times more likely to experience intentional poisoning than women (OR = 2.53; 95% CI: 2.49–2.58; p < 0.001).

Differences were also observed based on the type of health insurance: Those enrolled in the subsidized plan, which is intended for the lower-income population, were 26% more likely to experience intentional poisoning compared to those enrolled in the contributory plan, which consists of individuals who contribute to the health care system (OR = 1.26; 95% CI: 1.24–1.28; p < 0.001).

With regard to the life course, the highest rate was observed among adolescents compared to children, with an odds ratio of 137.4 (95% CI: 127.3–148.3; p < 0.001), indicating a distinct pattern by age group. Similarly, older adults were 22 times more likely to experience intentional poisoning compared to children (OR = 21.97; 95% CI: 20.5–23.5; p < 0.001), reinforcing the significant differences by life stage (See Table 3)

Table 3.

Association between sociodemographic characteristics and type of poisoning (OR and 95% CI).

Variable	Comparison	OR	IC 95%	P-value
Gender	Male vs Female	2.53	2.49 – 2.58	< 0.001
Type of insurance	Subsidized vs Contributory	1.26	1.24 – 1.28	< 0.001
Age	Adolescence vs Chilhood	137.4	127.3 – 148.3	< 0.001
Age	Old age vs Chilhood	21.97	20.51 – 23.53	< 0.001

The spatial autocorrelation analysis revealed clear differences between the types of exposure (See Fig. 2).

Fig. 2.

LISA spatial analysis of accidental and intentional poisonings (2015–2023).

For accidental poisonings, the Global Moran’s I was –0.13 (p = 0.866), indicating an absence of global spatial autocorrelation. Accordingly, the LISA analysis identified a single significant high-high (HH) spatial cluster in the department of Caquetá, indicating that this area had a high rate of poisoning and was surrounded by departments with similarly high rates. No significant spatial clusters were identified in other parts of the country.

In contrast, intentional poisonings had a Global Moran’s index of 0.20 (p = 0.017), reflecting significant positive spatial autocorrelation. The LISA map identified two significant HH clusters: in the departments of Cauca and Tolima, suggesting a concentration of events in these southern and central regions of the country.

4. Discussion

The results of this study showed that chemical poisonings in Colombia represent a public health problem with distinct social and gender patterns. In cases of accidental poisoning, “other chemicals” and pesticides were the main substances involved. According to the Sivigila classification, the “other chemicals” category corresponds to a residual group that includes various chemicals used in domestic, industrial, and commercial settings that cannot be classified into the system’s other specific categories. This finding is consistent with observations in other countries in Asia and Latin America, where these products continue to pose a high risk to both human health and the environment [12], [13]. Research has also indicated that, despite legal restrictions, their availability remains high and poorly regulated in rural areas [14], [15].

In Colombian regions such as the Pacific Coast and Putumayo, a significant burden of acute poisonings associated with pesticides and medications has been reported, predominantly among young people with suicidal intent [16]. These regions are characterized by a high proportion of rural population, higher levels of multidimensional poverty, and significant dependence on agricultural activities—conditions that facilitate the availability of and access to pesticides in both domestic and occupational settings, thereby increasing the risk of intentional and accidental poisonings [17] Similarly, paraquat—a herbicide widely used in agricultural activities and considered a highly hazardous pesticide due to its high acute toxicity and high lethality—has been identified as a critical problem in Antioquia, with high associated morbidity and mortality rates [18]. Similarly, other highly hazardous pesticides, such as carbofuran (widely marketed as Furadan®) and aldicarb, continue to pose a significant risk to public health due to their high toxicity and the high mortality rate associated with ingestion, particularly in cases of intentional poisoning. Its widespread availability in agricultural areas and the narrow margin between the toxic and lethal doses increase the likelihood of fatal outcomes, which is why various international organizations have recommended progressively restricting its use and promoting less hazardous alternatives [19], [20]. These findings reaffirm that highly hazardous pesticides must be more strictly regulated, as recommended by international health organizations.

Although in this study medications constituted the main group of substances involved in intentional poisonings, while pesticides predominated among accidental poisonings, it is important to distinguish between the frequency of occurrence and the severity of their consequences. Medications represented the most frequently involved group of substances; however, highly hazardous pesticides are characterized by a considerably higher case-fatality rate and a greater risk of fatal outcomes, even with lower-level exposures [21].

With regard to medications, various studies have documented that medications represent one of the main causes of self-poisoning, with psychotropic drugs (antidepressants, benzodiazepines, and analgesics) being the most frequently implicated agents, having a significant impact on morbidity and mortality due to their use in overdoses. [22], [23]. The available evidence also supports the view that this type of poisoning is closely linked to suicidal behavior, given that these medications are relatively easy to access in the home setting [24], [25].

With regard to sociodemographic differences, the higher prevalence of accidental poisonings among males could be related to men’s greater involvement in occupational activities that involve handling chemicals, especially in the agricultural and industrial sectors, where exposure to pesticides and other toxic agents is more common. This association has been described in various studies and is attributed to greater occupational exposure to potentially hazardous substances [26].

Furthermore, it was observed that intentional poisonings are more common among adolescent and adult women, which is consistent with European and Latin American studies reporting a higher proportion of women among cases of intentional poisoning [27], [28]. Meanwhile, research on older adults in Sweden has shown that intentional poisonings are a significant cause of morbidity in this age group [29], while in Turkey and Poland, clear differences in poisoning profiles have been documented among men, women, and children [30], [31].

Accidental poisonings in childhood were primarily associated with other chemicals and pesticides. The category of “other chemicals” includes various household and industrial products, such as chlorine-based disinfectants, acids, alkalis, and other chemical compounds commonly used in the home. In contrast, a study conducted in Brazil identified medications as the leading cause of accidental childhood poisonings, attributing this pattern to the availability of medications in the home and improper storage practices, especially among children under five years of age [32], [33], [34].

Beyond these findings, this study provides evidence on the territorial dimension of the phenomenon. The identification of significant clusters in the departments of Cauca and Tolima suggests that intentional poisonings tend to be concentrated in areas with specific socioeconomic and cultural characteristics. In the case of Cauca, the literature describes a high degree of rurality, a significant presence of indigenous and Afro-descendant communities, persistent social inequalities, and limited access to health services—especially mental health services—as well as the historical impact of the armed conflict [35]. For its part, Tolima has been identified as a department with a high suicide burden in Colombia, primarily in rural municipalities, where factors such as economic hardship, access to pesticides, and barriers to mental health care could contribute to the spatial concentration of these events [36]. This finding suggests the need to incorporate a territorial approach into prevention strategies—one that considers not only individual characteristics but also the structural and regional conditions that contribute to the occurrence of these events [37].

5. Conclusion

Acute chemical poisonings in Colombia are not only a public health problem because of their scale, but also because of the social and regional inequalities they reveal. Accidental poisonings, primarily associated with other chemicals and pesticides, disproportionately affect children and rural areas; meanwhile, intentional poisonings, which are predominant among adolescent and adult women, clustered in departments with high social vulnerability. This calls for differentiated and sustained interventions that integrate pesticide regulation, household chemical safety strategies, and suicide prevention programs with a gender- and territorially-focused approach. The integration of these measures will enable progress toward a sustainable reduction in the impact of chemical poisonings on the Colombian population.

Funding

This research did not receive any specific grants from public, commercial, or nonprofit funding agencies.

CRediT authorship contribution statement

Ernesto E. Fuentes-Fabra: Conceptualization, Investigation, Data curation, Writing – original draft, Project administration. Melissa Saenz-Ruiz: Methodology, Formal analysis, Data curation, Visualization, Writing – original draft, Writing – review & editing. Justiniano J. Arrázola Díaz: Conceptualization, Investigation, Supervision, Writing – review & editing, Project administration.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Prof. L.H. Lash

Data availability

Data will be made available on request.