Acute pesticide poisoning (APP) has been previously noted as a serious problem in Tanzania both for children and adults [
Other adverse health effects arising from pesticide exposure in children have been reported. In a study conducted in Egypt, children aged 9–15 years, who work seasonally in cotton farms to apply pesticides, reported significantly more neurological symptoms and reduced cholinesterase levels compared to controls [
Currently, Tanzania has no surveillance system for acute pesticide poisoning amongst children. Poisoning is reported in the health management information system within Tanzanian hospitals with minimum information on the causative agent. Previous studies have reported the lack of specificity of reporting of poisoning in hospital information systems [
Children, particularly those under 5 years of age, can be exposed to pesticides directly through ingestion involving swallowing pesticides or their concentrates, consumption of contaminated food, or hand-to-mouth behaviors typical of early childhood development. A study from Peru reported that 24 children died after drinking powdered milk substitute contaminated with the organophosphate methyl parathion [
Dermal absorption can occur with exposure to contaminated clothing or to dust or residues on floors and other surfaces or objects. Use of agricultural pesticides to control domestic pests and vermin has resulted in accidental poisoning of young children in urban homes in South Africa [
Other possible sources for exposure in this age category include pesticide drift [
These indirect risks of childhood pesticide are documented in various studies worldwide. A Tanzanian study in 2005 reported that mothers who retail pesticides may take their children with them to their shops where pesticides are stocked and sold [
Children are more vulnerable than adults to the adverse health effects of pesticides [
Inexperience, lack of maturity, illiteracy, and an inability to assess risk make accidental ingestion of pesticides amongst children more likely. Children have higher metabolic rates and their bodies are less able to detoxify and expel harmful chemicals. In short, children are absorbing a higher load of pesticides at a time when their bodies are still developing and are thus least equipped to protect themselves. Therefore, data on the extent of risk in children from pesticide exposure are needed to make preventive interventions to protect children from harmful effects of chemicals [
Because the magnitude of APP in children in Tanzania is unknown, this study was undertaken with the aim of estimating the burden of APP amongst children and characterizing the patterns of APP affecting children reported in healthcare facilities in Tanzania. Because Tanzania currently lacks a comprehensive surveillance system for acute pesticide poisoning, such a study would provide recommendations to address this gap.
Data on childhood APP were obtained as part of a hospital-based review in 4 selected districts of Tanzania where intensive coffee and vegetable production was associated with intense pesticide use. The study methods have been previously described [
These 10 facilities were selected as having recorded the majority of APP cases in the retrospective study in 2005 and included regional (
This study focuses on data reported for children in the prospective study in 2006. Since children are legally defined in Tanzania (and in many other countries) as persons under 18 years, we used 18 years as the cut-off to define children in this study. The prospective study involved intensive training of facility staff to record all APP cases with the aim of improving quality of data collection. Comparison of retrospective review to prospective data collection suggested that the training intervention was successful in reducing missing information on the circumstances, outcomes, and agents responsible for APP by about 50%, 50%, and 20%, respectively [
The data collected were, firstly, retrieved from the register book at the facility and included patient registration number, date of consultation, location, gender, circumstances of poisoning, and outcome. Secondly, the patient registration number was used to locate the patient folder from which further information was extracted including agents responsible, circumstances, and treatment of poisoning. Comparisons were also made to data found in the poisoning register. The data were collected using a standardized data collection sheet by specifically trained medical data recorders.
A case of APP was defined in this study as a diagnosis of APP made by the clinician and recorded in either the register or patient folder or both. In general, clinician diagnosis was based on a history of exposure (from the patient, relative, or accompanying person) to one or more pesticides and clinical manifestations of poisoning or specific laboratory test results compatible with APP, within 14 days of exposure. Children were defined as persons aged 17 years or younger.
Description of childhood poisoning by age was done using 5-year intervals. The group of children under 6 is a common category used internationally and the grouping of children 6 to 17 enables comparison to similar studies conducted elsewhere. Poisoning circumstances were classified as (a) suicide, (b) accidental, (c) occupational, (d) homicide, or (e) unknown, based on information recorded in the diagnostic sections of the patient folders or directly documented in the patient register books. In cases where the two sources disagreed, information from the patient folder was used, because information completed by a medical professional was judged to be more likely precise as compared to the register book which was completed by more junior staff.
The outcomes of APP were classified as (a) recovery, (b) absconded, (c) referred, (d) residual disability after discharge, (f) death, or (g) unknown. A second pair of analyses of outcomes further reduced classification as (a) fatal versus nonfatal and (b) known versus unknown outcomes. Agents responsible for APP were classified as (a) specific (active ingredient was identified), (b) nonspecific (active ingredient was known by general category), or (c) unknown (active ingredient was not known to the clinician and/or not recorded). Comparisons involving circumstances and outcome of poisoning, agents, gender, and age were conducted using
Data on APP cases were used as numerator data to calculate morbidity rates stratified for gender, geographical area, and age. To calculate denominators for rates, population census data was obtained from the Tanzania Bureau of Statistics based on a national census conducted in 2002 [
Because the study involved record review and no data were collected directly from individuals, there was no consent required. To ensure confidentiality, patient names were replaced by codes which were used as identifiers in data analysis. Ethical approval was secured from the TPRI, the National Institute for Medical Research (NIMR) in Tanzania (Ref. NIMR/HQ/Vol XI/371), and the University of Cape Town (Ref. 328/2004).
Of the 10 facilities followed up for 12 months during the study, 9 facilities reported 53 cases of childhood APP. The facility which did not report children poisoning was a small health center near Arusha. The highest number of cases (
The age category with highest proportion of poisoned children was 16-17 years (30.2%) (Table
Age category and circumstances of poisoning for APP amongst children in Tanzania.
Age category | Known | Unknown | Accidental | Occupational | Suicide | Homicide | Total |
---|---|---|---|---|---|---|---|
1–5 | 11 | 0 | 11 | 0 | 0 | 0 | 11 |
6–10 | 12 | 0 | 12 | 0 | 0 | 0 | 12 |
11–15 | 11 | 3 | 2 | 2 | 6 | 1 | 14 |
16-17 | 13 | 3 | 1 | 2 | 10 | 0 | 16 |
Male | 21 | 2 | 15 | 1 | 4 | 1 | 23 |
Female | 26 | 4 | 11 | 3 | 12 | 9 | 30 |
Total | 47 | 6 | 26 | 4 | 16 | 1 | 53 |
The two most commonly recorded circumstances of poisoning were accidents (
The proportion of circumstances due to suicide was significantly higher in females compared to males (46.2% versus 19.0%; PRR females/males = 1.7; 95% CI = 1.0–2.7). Suicide was concentrated entirely in children older than 10 years, comprising just over half of all cases (16 out of 30 cases) in this age group but was not present at all in younger children (
Only two children were reported to have died (3.8%). However, the outcome was unknown for 10 children (18.9%).
The agents responsible for poisoning are reported in Table
Classification of agents responsible for APP in children in Tanzania.
Product | Chemical group | WHO Class | Frequency | Percentage by category | Percentage of all agents |
---|---|---|---|---|---|
|
|||||
Zinc Phosphide | IN | 1b | 2 | 18.2% | 3.7% |
OP | OP | II | 7 | 63.6% | 13.2% |
Sulphur | IN | IV | 1 | 9.1% | 1.9% |
Endosulfan | OC | II | 1 | 9.1% | 1.9% |
|
|
|
|
||
|
|||||
Livestock dip | UN | UN | 1 | 6.67% | 1.9% |
Food poisoning | UN | UN | 11 | 73.33% | 20.7% |
Rat poison | UN | UN | 3 | 20.00% | 5.6% |
|
|
|
|
||
|
|||||
UN | 27 | 100.00% | 50.9% | ||
|
|
|
IN: inorganic; OP: organophosphate; OC: organochlorine; UN: unknown.
The majority of the specific agents reported (
The proportion of circumstances due to suicide was higher in cases with unknown agents compared to cases with known agents but the difference was not statistically significant (40.0% versus 29.6%; PRR unknown/known = 1.3; 95% CI = 0.7–2.5). The proportion of known agents was significantly higher in younger (age: 0–5 years) compared to older (age: >5 years) children (62.2 versus 25.0%; PRR younger/older = 1.6; 95% CI = 1.1–2.3). The proportion of cases for which the outcome was known was higher for females compared to males (PRR 1.4; 95% CI = 0.8–2.1), for older compared to younger children (PRR = 1.4; 95% CI = 0.6–3.5) and for suicidal circumstances compared to other circumstances (PRR = 1.6; 95% CI = 0.8–2.7), but none of these differences were statistically significant.
The annual IR for APP was 1.61/100,000 with higher rates reported for the Arusha region. The age group 16-17 years reported the highest IR (6.17/100,000) (Table
Incidence rate, mortality rate, and case fatality rate for APP for children in Tanzania.
Poisoning cases per year | Population | Annual IR (per 100,000) | |
---|---|---|---|
Regions | |||
Arusha | 15 | 738,618 | 2.03 |
Mwanza | 25 | 1,764,238 | 1.42 |
Kilimanjaro | 13 | 782,442 | 1.66 |
|
|
|
|
|
|||
Age groups | |||
1 to 5 | 11 | 1,258,717 | 0.87 |
6 to 10 | 12 | 976,393 | 1.23 |
11 to 15 | 14 | 790,902 | 1.77 |
16 to 17 | 16 | 259,286 | 6.17 |
|
|
|
|
|
|||
Gender | |||
Males | 23 | 1,649,495 | 1.39 |
Females | 30 | 1,635,803 | 1.83 |
|
|
|
|
This study identified 53 acute pesticide poisoning cases in one year involving children in 9 of the 10 facilities in this study. The fact that all but one facility reported a childhood poisoning case indicates that childhood pesticide poisoning is ubiquitous in the study area. With intensive surveillance and improvements to the reporting system, a larger number of cases are likely to be captured in the health management information system.
The regional hospital reported the highest number of cases (
The age distribution of APP in children suggests the highest proportion of APP is located in the age group 16-17 years and lowest in the age category 1–5 years. This may be explained by the fact that younger children under five receive more parental attention than adolescents whereas older children 16 to 17 years may be at higher risk of suicide (given psychological challenges faced by young people approaching adulthood) and face occupational hazards under conditions of child labour.
Children under 5 years comprised 21% of all cases of poisoning of children in the study. This proportion was lower than reported in South Africa (34.4%) [
In terms of gender, this study found that the majority of childhood poisoning cases were female. These findings are consistent with studies in India [
Suicide was concentrated exclusive amongst older children, particularly females. This gender predilection for suicide is consistent with findings in the literature from Sri Lanka, China, Chile, and Korea [
Accidents were the single highest category of circumstance (49.1%) and were more common in males (65.2%). The findings in this study are similar to studies conducted in South Africa, which reported that the main circumstance of children exposure to pesticides was accidental ingestion of organophosphates (OPs), either as residues on unwashed fruit or from poorly marked storage containers, or from dermal and respiratory absorption following OP application for pest control in and around homes [
Similarly, a study conducted in Zambia reported that the major circumstances of poisoning for children poisoning were accidental [
Occupational circumstance was uncommon, accounting for only 7.5% of APP cases. Nevertheless, the presence of four APP cases in working children is an indication that children are involved in pesticides application, which is against the Tanzania Law of the Child Act of 2009 [
This situation is similar to findings of a Philippine study which found that children start working in vegetable farms with pesticides as early as 6 to 9 years old [
There were only 2 fatalities (3.8%) reported amongst the children cases, which is about 40% lower than the case fatality rate (CFR) found in adults [
However, the outcomes for 10 cases were unknown and these might have included unrecorded fatal outcomes. Where there were fatal outcomes, delays in accessing timely care may have contributed to the deaths.
The poor percentage of cases where poisoning agents could be identified (about half of the cases were of unknown agents) could reflect the fact that children have little capacity to identify or remember agents in cases of poisoning and that caregivers are not present when the poisoning occurs so they cannot identify the agent.
Organophosphates (OPs) emerged as the single most important group of agents responsible for poisoning (
The study reported cases of poisoning involving endosulfan and zinc phosphide. Endosulfan is a toxic pesticide earmarked for elimination under the Stockholm Convention on Persistent Organic Pollutants (POPs) and is also listed under the Rotterdam Convention, although its inclusion was stalled for many years because of the seeming absence of data on severe incidents involving endosulfan [
Zinc phosphide is an extremely toxic agent that is not registered in Tanzania. Illegal distribution of this product for rat control in households is a potential reason for the association with childhood poisoning. The product is thought to be imported illegally from neighboring countries and its presence in Tanzania as a cause of APP in children points to the need for strengthening border controls by the government to discourage this practice.
The study estimated the first population-based incidence rates for pesticide poisoning for children in northern Tanzania. The childhood IR for APP in this study (1.61/100,000) was lower than rates found in studies of APP in children in Korea (3.6/100000) [
The study indicates that acute pesticide poisoning is common among children in northern Tanzania with estimated IR of 1.61/100,000. The most common known agents were WHO Class I and II pesticides including a number of organophosphates. The most common circumstances of APP in children were accidents for children 10 years or younger and suicide for children over 10. To reduce APP related to accidents and suicide, attention should be paid to safer storage, improved hygiene measures, and control of access to toxic pesticides. Occupational APP, though uncommon, signals the need for attention to the eradication of worst forms of child labour. Multifaceted intervention efforts are needed to reduce pesticide poisoning among children in Tanzania.
The authors declare that there are no conflicts of interest regarding the publication of this manuscript.
The authors would like to acknowledge the National Research Foundation of South Africa, the Fogarty International Center of the National Institutes of Health (Health, Environment, and Economic Development program), and the Work and Health in Southern Africa (WAHSA) program funded by Swedish International Development Aid (Sida) for their great support of this research work.