In our study, the incidence of IUIs was 33.42%, higher than the results in other studies [17, 29], which may be the result that our subjects were children aged 0 ~ 6, who occupied a higher injury risk than other age groups. In addition, children aged 0 ~ 3 in this study were investigated by face-to-face interview, which made the data of younger children more reliable and had a certain influence on the incidence rate. Meanwhile, this finding confirmed the necessity of special attention and targeting preventions towards IUIs among Chinese urban children, which need to be carried out with adequate evidence bases. Thus, based on Bronfenbrenner’s ecosystem theory and our previous efforts, we systematically explored the causes of children’s IUIs in Chinese urban areas and revealed significant relationships among risky behaviors, parental supervision, home environmental risks and IUIs through SEM.
The CFA exhibited that all latent variables had good reliability and discriminant validity through the results of composite reliabilities and the AVEs of each construct. GFI, CFI and RMSEA were all acceptable, so we considered that the measurement tools used in this study met the conditions for the structural model. Results of SEM of the general sample revealed that the total value of effect from risky behavior was much higher when compared with parental supervision and home environmental risks, though behavior was influenced by the latter two to some extent. Home environmental risks played direct and indirect roles in children’s IUIs, whereas parental supervision could only influence special groups, such as children aged 4 ~ 6 years and girls.
Our investigation tool for risky behaviors involved animal bites, burns, falls, suffocation and many other injury types, all of which corresponded to the relevant child IUIs, for example, the behavior of a child poking a power socket was in corresponding to the unintentional burns. SEM of subgroups showed the contribution of risky behaviors with aging, because the better motor control and coordination of older children, the higher possibility for hazard situations , which was similar to the fact that children who climb frequently were vulnerable to falls. Likewise, the behavioral effect was greater for boys than for girls, which may be due to the gender differences in early developmental trajectories for emotional expression. Boys are more inclined to express anger-related emotions and have greater tendency to engage in rough and tumble play than girls , which leads to more frequent conduct problems and more serious IUIs. Similarly, multimethod strategies study  revealed that most of the boy’s injuries followed misbehaviour (i.e., mothers reported that 60% of boys suffered significantly more injuries when they engaged in inappropriate behavior). Also, boys express more optimistic bias and tend to attribute injuries to bad luck, while girls express more considerations about the consequences of their actions in the same situation . Thus, our results served as reminders for IUIs prevention programs, in which appropriate efforts according to age and gender characteristics may be needed.
Despite risky behaviors occupied the strongest effect on children’s IUIs, reducing risks caused by risky behaviors required more reasonable parental supervision and better home environment. The direct effects of environmental risks and supervision were validated across the entire sample, similar to previous studies [29, 34]. More importantly, we proved the indirect role of the both on IUIs. The underlying mechanism may be that parents who provided direct attention, with closest proximity on a continual basis were able to moderate their children’s injury-prone tendencies and ensure child safety in the home setting , and home environments may influence the emergence of externalized behavior problems in children and thus create favorable conditions for being exposed to IUIs, in line with other studies [36, 37].
The indirect effects of the home environmental risks implied that the hazards of IUIs were on the rise when children with high behavioral risks are in unsafe households, particularly 4-6-year-old children and boys. The reason may be that high-risk settings would stimulate children to show more externalized behavioral problems and fostered interactions with hazards. Similarly, Schwebel  revealed that in a complicated housing environment, children with higher behavioral risk might judge environmental hazards with a rushed, impulsive manner, which might lead to misestimation of the risk involved in a particular activity, and in turn lead to injury. Therefore, the co-existence of direct and indirect paths highlighted the dual value of addressing risk factors in house. However, contrary to previous results, the direct impact of home risk varied more in girls than boys. A possibility involved greater perceptual sensitivity in girls. If girls are more attuned to the fine details of their environment, they will experience more stressors and curiosity, which can trigger their susceptibility to more injury events around them . Also, this finding suggested that injury risk for girls arose more from poor conditions that already existed at home than motivation of child’s behaviors from home environment, while the opposite was true for boys. Hence, reducing IUIs risk in girls necessitates more complete disengagement from dangerous situations, and for boys, timely home inspection and regulating children’s behavioral tendencies may be more important.
Results suggested that parental supervision was not significant to deter younger children’s injury risk, either directly or indirectly. We speculated that this may be associated with differences in child care. Morrongiello  reported that parents of younger children only focused on increasing the frequency and proximity of observations and routinely overestimate desired effect of their supervision strategies. By contrast, parents of older children exercised safety education and fear-inducing strategy, in which parents would stress the serious consequences if children did not follow their guidance. In addition, the early IUI experiences of 4-6-year-old children reminded their parents of the relative necessity for parental supervision, which made the roles of supervision reported by parents more prominent. We did not find any significant effects on parental supervision of boys, but did of girls, which was beyond our expectations. One possible explanation was that specific societal expectations for males and females. Boys are always expected to protect their families and to overcome dangers that interfere with their ability to provide for their families, whereas girls are allowed to show more emotions to take on the traditional role of caregivers . Thus, more encouragements to participate in risky games are given to boys, while more caution about safety to girls, which relatively weakened the roles of supervision.
Whether in the general population or each subgroup, risky behavior was involved as an intermediary, and meanwhile, the effects of risky behavior and home environment were greater than that of parental supervision, which made we acknowledge the relative materiality of all three in preventing IUIs. Although parental supervision has been advocated as a protective method in the past, we found that it separately cannot completely offset the crisis caused by environmental risks and behaviors with limited power. Consequently, the best strategy to moderate risk for IUIs is an effective combination of behavioral regulations and environmental interventions, and measures targeting parental supervision are considered as promising remedies. For younger children, this means keeping the consistency of space as much as possible to timely take behavioral modification and reasonable planning of indoor activity space. For older children with increased cognitive awareness, this signified necessity of regular health education and improvement of risk perception. Besides, specific efforts, such as caregivers’ safety education and enhanced safety equipment (e.g., child-resistant container, cabinet latches, stair gates), will be better able to contain crises from environment and supervision. In terms of gender, girls are highly sensitive to environmental risks, so it is critical to carry out a thorough screening of the risk of home from the perspective of children. At the same time, moderate behavioral constraints and verbal reminders may yield better results, because girls are more compliant with their parents . For boys, efforts should be directed towards positive regulation of impulsive behavioral traits. Traditional didactic education should also be changed and more participatory activities can be adopted, such as training tasks in hazard perception, sound decision-making, and safety rules, which will be better targeted to ameliorate boys’ potential risks.
To decrease child IUIs more efficiently, more efforts need to be done besides the above measures. Current data of Chinese children’s IUIs stemmed mainly from the China’s Disease Surveillance Points system and National Injury Surveillance System consisted of 127 sentinel hospitals ; but due to the relatively low proportion of deaths and hospital visits caused by injury events, IUIs disposed of by family members were often missed. It is a call to attach importance to the collection, analysis and utilization of basic IUIs information. Measures to expand the monitoring sites to primary units (e.g., schools, kindergartens, and communities), standardize the information collection process, and formulate multi-sector strategies involving the whole society will be good complement to Chinese’s non-medical treatment IUIs. Moreover, parents in Chinese urban areas generally lacked sufficient recognition of the IUIs’ controllability, and especially parents who belong to floating families were basically in the blind spot of child IUIs prevention . Therefore, further strengthening investment in home safety publicity, focusing on children with weaker parental supervision and larger needs of injury prevention, such as migrant children who often be left at home alone, urging parents to enhance literacy of IUIs may be the key to future injury control program.