Article Text
Abstract
Background Unintentional injuries in and around the home are important causes of preventable death and disability among young children globally. In Nepal, there is a lack of data regarding home injuries and home hazards to guide the development of effective interventions for preventing childhood home injuries. This study aimed to determine the burden of unintentional home injuries in children <5 years in rural Nepal and quantify the injury hazards in their homes.
Methods A survey was conducted in 740 households in rural areas of the Makwanpur district during February and March 2015. The primary carer reported home injuries which occurred in the previous 3 months and data collector observation identified the injury hazards. Injury incidence, mechanism and the proportion of households with different hazards were described. Multivariable logistic regression explored associations between the number and type of home hazards and injuries.
Results Injuries severe enough to need treatment, or resulting in non-participation in usual activities for at least a day, were reported in 242/1042 (23.2%) children <5 years. The mean number of injury hazards per household was 14.98 (SD=4.48), range of 3–31. Regression analysis found an estimated increase of 31% in the odds of injury occurrence associated with each additional injury hazard found in the home (adjusted OR 1.31; 95% CI 1.20 to 1.42).
Conclusions A high proportion of young children in rural Nepal sustained injuries severe enough to miss a day of usual activities. Increased frequency of hazards was associated with an increased injury risk.
- surveys
- multiple injury
- child
- home
Statistics from Altmetric.com
Introduction
Globally, unintentional home injuries are a major cause of premature death and serious disability among children under 5 years.1 These deaths impose a large economic burden on families and society,2 especially in low-income and middle-income countries (LMICs). Rates of mortality related to injury among children in LMICs are four to six times higher than those in high-income countries (HICs).3 Compared with adults, young children are a high-risk group for unintentional injuries.2 Most injuries to children under 5 years occur in the home environment in both LMICs4 5 and HICs.1 6 Low-income communities live in environments with a greater numbers of hazards.3 The exposure of children to those hazards is likely to increase their risk of injury.
Some observational studies from LMICs have reported the frequency of hazards in the home that could potentially result in injury7–12 but there has been little research reporting the injuries arising from such hazards. These studies revealed that there is a significant burden of home injury hazards in LMICs, representing an important opportunity for injury prevention. Environmental change in the home may limit a child's injury risk, either by eliminating the hazards or by using safety equipment and practices to restrict exposure to the hazards.13 However, accurate information about the injury and the associated risk factors are essential in the design and implementation of effective home environmental change interventions.14 15
Nepal is one of the ‘least developed’ countries of the world.16 In the absence of robust death registration and injury surveillance systems, the burden of injury in Nepal has not been measured comprehensively. Household surveys and hospital data suggest that home injuries (eg, falls, burns, poisoning and animal related injuries) are common among young children in Nepal.17 18 No studies have quantified the injury hazards in their homes or assessed the association between home hazards and injuries in children under 5 years. To address this knowledge gap, this study aimed to describe the epidemiology of unintentional home injuries in children under 5 years, explore the prevalence of home environmental hazards and investigate the relationship between home environmental hazards and unintentional home injuries among children living in rural areas of Makwanpur district, Nepal.
Methods
Study design
The study, conducted in Makwanpur district, Nepal, used a community-based, cross-sectional design. This district was selected because of its geographical location (a predominantly rural district, 4 hours’ drive from Kathmandu), the demographics of its population (similar to Nepal as a whole) and the fact that it has three distinct types of landscapes (high hills, mid hills and lowland/plains) which are similar to most other districts of Nepal17 outside of the mountains.19
Sample size
The sample size of 740 households was calculated, based on UN guidelines for household surveys.20 Previously published research showed that 72% of households in a community survey in Pakistan had six or more injury hazards,11 and that having six or more hazards increases the risk of child injury.21 Therefore, the proportion of homes with the indicator of interest in the sample size estimate was 72%. The Makwanpur district (area 2426 km2) has a population estimated at 420 477 people living in 86 127 households. The proportion of the population aged 0–59 months (9%) and the average household size (4.88) was ascertained from the 2011 population census.22 The design effect was 2.0, based on published guidelines for household surveys using cluster sampling.20 The anticipated non-response rate was 10% based on the study conducted in Nepal17 and described in the WHO guidelines.23
Sampling method
Multistage cluster sampling, with probability proportional to size (PPS) methodology, was applied as a sampling method.20 A household was defined as a group of individuals living together and sharing the same kitchen. A household was eligible to be included in this study if there was at least one child aged 0–59 months and the family had been residing in that house for three or more months. Simple random sampling was used to select the primary cluster (ie, village development committees) and secondary cluster (ie, households) (figure 1).
Data collection tools
Data were collected using a structured two-stage questionnaire. In the first stage, data were collected on the household residents and hazards in the home. The second stage was completed only for those households who reported an injury in a child under 5 years. A structured questionnaire, adapted from the WHO Guidelines,23 and a hazard checklist, adapted from a similar study conducted in Pakistan,11 were developed as data collection tools. The checklist included 32 possible hazards that could result in eight common injuries. The questionnaires and checklist were piloted before being used for data collection in the field. Data collectors were trained to complete the structured questionnaire and to assess the presence or absence of only those hazards included on the checklist.
Definitions used
Injury: The definition of injury used in this study was ‘any unintentional injury that occurred within the home environment and was severe enough to need treatment or resulted in non-participation in usual activities for at least a day’. Unintentional injury is defined as any injury that occurred without any intent of self-harm, homicide or suicide. A 3-month recall period was applied to achieve sufficient detailed information about non-fatal injuries and to minimise recall bias.24 All the children aged under 5 years in the sampled households were included in the study. If a child had sustained more than one injury in the last 3 months, the respondent was asked to provide information about the one injury that they considered to be the most severe. Home environment: the kitchen, bathroom, bedroom or sleeping area, the courtyard, the rooftop and the immediate vicinity of the house. Home hazard: Anything that represents a physical or structural hazard that has the potential to cause injury.
Data collection process
A door-to-door survey was conducted by six trained data collectors during February and March 2015. Where possible, information was collected from the main caregiver of any child in the household. In the absence of a main caregiver, another member of the household providing care to the child/children provided information. If no adult members were at home during the first visit, these households were visited twice. If no one was at home for the second visit, an alternative household was selected for the survey. A household nearest to the sampled household was used as an alternative household. Verbal consent was obtained prior to the interview. With permission of the parent/carer, the data collectors who were trained to undertake home hazard assessments also visited each area of the house to identify and document the hazards for injury using the checklist. Unused areas of each household that had been locked continuously for the previous 6 months, or places designated for worship, were not observed.
Data analysis
IBM SPSS Statistics for Windows V.22.025 was used for data analysis. Injury incidence and the proportion of households with hazards were analysed. Continuous data were assessed for normality distribution using the Shapiro-Wilk test in SPSS. The data were normally distributed as the statistic values were greater than 0.05 and therefore transformation was not required. Rates and proportions for child injuries were calculated and reported by age groups, gender and types of injury mechanisms. Multivariable logistic regression analysis was used to explore the association between having any home hazards and children sustaining any injury, controlled for family and home variables. Secondary analyses explored the association between the presence of specific hazards and children sustaining hazard-specific injuries. A value of p less than 0.05 was considered as statistically significant.
Results
The 740 households surveyed housed 4967 residents, 1042 (21%) of these were children aged less than 5 years, with an average number of 1.40 children per survey household. For data collection purposes, due to the adult household member not being available after two visits, 12 (1.6%) households were substituted with other households. The characteristics of the surveyed population are summarised in table 1.
Injury incidence
Overall, 31.4% (n=232/740) of households reported at least one child injury. The overall injury rate among children aged under 5 years was 232.2 per 1000 children (95% CI 206.9 to 259.1) (table 2). Injury rates were highest among children aged 36–47 months and lowest in children aged <12 months. The injury rate increased as age increased up to 47 months but at 48 months and older the rates fell. Overall, injury rates in male children were marginally higher than in female children but this difference was not statistically significant.
The largest proportion of injury events occurred due to a fall (n=89, 37%), followed by fire, burns, scalds (n=67, 27%), cuts or crushes (n=53, 22%) and then animal-related injuries (eg, bite, sting or hit by domestic or wild animal) (n=24, 10%). Only the proportion of fire-related injury, burns or scalds was slightly higher in female children (table 3).
Of the total 242 injury events, 43.8% (n=106) occurred inside the home and 56.2% (n=136) occurred outdoors but within the immediate vicinity of the home. Fires, burns or scalds were found to be the most frequent (58.5%) injury event among all 106 injury events that occurred inside the home whereas falls were found to be the most frequent (47.8%) among all 136 injury events that occurred outside the home but within the immediate vicinity of the home.
Injury hazards in the home environment
Across all surveyed households (n=740), the mean number of injury hazards was 14.98 (SD=4.48, range=3–31). The prevalence of hazards in the home varied across different types of injury (table 4). It was recorded that 98.1% households did not have protective railings on stairs or ladders. The windows in 83.6% households and the balconies in 50% households lacked a protective barrier. Cooking stoves were within the reach of children in 98.4% households and 42.2% household did not have a barrier or door between the sleeping and cooking areas. About 80%–83% households had sharp or hard protruding components and breakable objects within the reach of young children. Common poisoning hazards, including a child’s ability to access alcoholic beverages, were found in 91.5% households, agricultural chemicals or fertilisers in 61.5% households and fuels in 44.4% households. Plastic bags were within the reach of children in 52.2% households. Outside the home, cattle sheds were reported to be fenced in less than 10% of households and more than 95% households had accessible ponds, lakes and streams. A photograph of a typical wooden Nepali home is presented in the online supplementary figure S1.
Supplemental material
Association between home hazard and child injury
The regression analysis to explore the association between home hazards and reported child injury included 233 injury cases and 800 non-injury cases. Nine (4%) injuries categorised as ‘other injuries’ were excluded due to the small numbers of specific injury types. The potential confounding variables identified from the univariable logistic-regression analysis are presented in the online supplementary tables 1-4. A positive association between the number of home hazards and number of children with an injury was found for both any hazards and specific hazards. There was an estimated increase of 31% in the odds of a child sustaining an injury with each additional injury hazard found in the home (adjusted OR (AOR) 1.31; 95% CI 1.20 to 1.42). Similar associations, but with increased ORs, were obtained for falls, fires, burn or scald injuries, and for cut or crush injuries (table 5).
Discussion
This is the first study in Nepal to explore the injury risk for children under 5 years old of a range of different hazards found in homes in rural communities. In summary, this study found that 23.2% of young children in rural Nepal sustained injuries severe enough to miss a day of usual activities. An increased frequency of injury hazards in the home was found to be associated with an increased risk of child injury.
Injury incidence
While the incidence of injury in children less than 5 years reported in this study (23.2%) is broadly similar to some studies conducted in other LMICs,26 27 other studies have found both higher9 28 29 and lower rates.11 30 The most likely reasons for such differences may be due to the different ages of the children studied, different recall periods, study areas (eg, rural vs urban), living circumstances, cultural practices and different socioeconomic conditions of the sampled households. Studies with longer recall periods tended to have lower rates of injury; older events are more likely to be forgotten, leading to an underestimation of the true rate.
In the current study, injury incidence in male children (24.1%) was only marginally higher than in female children (22.2%), which is similar to findings from Iran,30 Egypt31 and Turkey.32 In contrast, several studies in other LMICs have demonstrated differences by gender greater than could have occurred by chance.28 33 The rates of injuries for different age groups were similar in other studies compared with this study30 34 and the peak in injury rate at 47 months was also supported by earlier studies.31 32 Similar to the findings of this study, in other studies, falls and burns were reported as the most frequent injury mechanisms.28 30 31
Injury hazards in the home environment
This study found a substantial number of hazards in most of the surveyed households (mean of 14.98 hazards, range=3–31 hazards). A community based study in China, which investigated home injury hazards among toddlers (24–47 months), reported similar findings with mean home hazards of 12.29 (SD=6.39) and a range of 0–36 hazards.12 However, the prevalence of home hazards reported in the study was based on parent-reported data, not objective observation. Parents may have been reluctant to report high numbers of hazards for fear their home would be considered unsafe. This may have resulted in an underestimate of the true number of hazards.
This study observed that 98% of households had unprotected stairs or ladders. Similar findings were reported in one Indian study10 but in other studies, the proportion of households with either unsafe or no railings was much lower, between 23% and 25%.9 29 In the current study, 84% of households had unprotected window rails and 50% had unprotected balconies. Similar findings were reported from Karachi, Pakistan,11 but the study from Egypt found these hazards in only 6%–8% households.29 This is likely to be due to differences in housing construction between the countries: homes in Pakistan are more likely to be similar to those in Nepal and those in Egypt more likely to be different.
Most cooking in the rural areas of Nepal is on open fires using firewood, often at ground level. In this study, 98% households had cooking stoves within the reach of the child. A recent systematic review also highlighted that the use of open fires for cooking was the most common hazard leading to burn injuries in the Nepalese population.35 This is different to most other studies conducted in similar settings, where only about half, or less, of the households had open fires that children could reach.9–11
This study found that sharp or hard protruding components (eg, big stones or pieces of wood, woodpiles, old machinery, etc) were within the reach of children in most surveyed households (83%). These objects were not reported as potential hazards for cut or crush injuries in any study from LMICs.
In Nepal, there was a greater proportion (52%) of bodies of water that were accessible to children that were near to the home, and were therefore potential drowning hazards than in a neighbouring country such as India (32%–36%).8 9 This study found that open containers of water (or other liquids) were within the reach of a child in 84% of households. In contrast, only 18%–48% households in Pakistan had this drowning hazard in their home environment.11
Alcoholic beverages were not reported as a potential poisoning risk in any other study from LMICs, possibly because they were not perceived as risks, or the households sampled did not consume alcohol. This study found that agricultural chemicals or fertilisers were within the reach of the child in 62% households. This may be due to the lack of lockable cupboards for storage or poor safety practices in Nepalese households, as also seen in other LMICs.8 10 11 Similar to the study in Pakistan,36 there was a high risk of paraffin poisoning in Nepalese children as 44% households had such fuels within the reach of the child. The proportion of households with suffocation and choking hazards found in the current study was similar to the studies conducted in China,12 India9 and Pakistan.11
Association between home hazard and child injury
In line with the result of this study, an Indian study reported a positive relationship between the number of home injury hazards and child injury incidence.9 That study showed that the odds of having an injury increased by 55% with each additional injury hazard found in the home (AOR 1.55; 95% CI 1.3 to 1.8), adjusted for confounding variables. Positive associations between increasing numbers of home hazards and increasing numbers of injuries have also been found in HICs including New Zealand21 and Canada.37
In contrast, studies from Australia,38 Egypt31 and the UK39 did not show positive relationships between numbers of hazards and numbers of injuries, once confounding factors were adjusted for. The reasons might be that the children in these countries were less likely to be exposed to the hazards or the potential impact of hazards to the children were moderated or reduced. It is important to note that a household may have many hazards but whether or not the child interacts with that hazard determines if they are at risk of injury.
Strengths and limitations
The findings of this study increase our understanding of the impact of home injury hazards on child injuries in rural households of Nepal. One limitation was the length of the recall period for child injury. While 3-month recall periods are useful in the collection of detailed information regarding non-fatal injury, a longer recall period would have allowed inclusion of seasonal injuries such as those that occur during the monsoon. However, longer recall periods risk caregivers forgetting some injuries.40 The hazard assessment was carried out at the time of the survey, so it is possible that the household could have contained a different number and types of hazards at the time any injury event occurred.
In many studies, poisoning, drowning, suffocation and choking have been reported as causes of mortality and morbidity in preschool children. In this study, no injury events were reported due to poisoning and drowning, and one event of suffocation or choking was reported. The reported incidences of poisoning, drowning, and suffocation or choking may be true, and the incidence estimate valid for this sample. Alternatively, there may have been under-reporting of these injury types, possibly by the parents not wanting to share information of these events if they were uncomfortable reporting them, or that because these injuries did not leave visible marks, they were less well remembered or not considered as injuries. The lack of drowning cases might have been due to the fact that the data were collected during the dry season and the recall period used.
Conclusions
This study found a significant burden of injuries that occurred to young children in the home: falls, fire, burn or scald, cut or crush injuries were the most common types of injuries reported. This study also revealed that, as the number of home hazards increased, the odds of children with injury also increased, even after adjustment for confounding factors. These findings suggest that addressing the number of injury hazards in rural Nepalese homes may be effective in reducing home injuries in children. Injury prevention initiatives should consider the development and evaluation of interventions to reduce the number of hazards in the home.
What is already known on the subject
Unintentional injuries in young children commonly occur in or around the home.
Few publications have reported both injury hazards and injury incidence in low-income countries.
The burden of unintentional home injuries among children under 5 years is higher than other age groups of children in Nepal.
What this study adds
Falls, burns and scalds and cut or crush injuries were the most commonly occurring injuries in and around the home in children under 5 years in Nepal.
Increased frequency of injury hazards in the home is associated with an increased risk of child injury.
Interventions to reduce injury hazards in rural Nepalese homes may reduce child injuries and should be evaluated.
Acknowledgments
This research was presented as a poster at the World Injury Prevention and Safety Promotion Conference (Safety 2018), Thailand, 6 November 2018. The authors thank Mother and Infant Research Activities (MIRA), an organisation in Nepal for support in the data collection of this study. The authors also thank the interviewers and participants who took part in this study.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @bhatta111, @TDeave
Contributors TD and JAM were the supervision team for this PhD research. SB conducted the fieldwork and data analysis, and prepared the first draft of the manuscript. TD and JAM provided comments and feedback on the first and consequent versions. All authors read and approved the final manuscript.
Funding This study was funded by the University of the West of England (UWE Bristol), United Kingdom as part of a PhD Scholarship.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval Study approval was obtained from the Ethical Review Board (ERB) of the Nepal Health Research Council (NHRC), Nepal and the Faculty Research Ethics Committee (FREC) of the University of the West of England (UWE Bristol), United Kingdom.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information.