Our study is possibly the first study that investigated non-fatal unintentional injury risk among children with disabilities in China based on the biopsychosocial model of the International Classification of Functioning, Disability and Health – ICF. In the present study, we found a strong association between disability status and injury in this sample of Chinese pediatric population. Compared with children without disabilities, children with disabilities were over 4 times more likely to be injured in the previous 12 months. We also found that above half of injuries among children with disabilities occurred inside the home, and nearly 50% injuries were caused by falls.
Xiang and colleagues found that children with developmental disabilities or chronic medical conditions were at a higher risk for injury than were children without those conditions 
. Sherrard et al. conducted a longitudinal study in a group of intellectually challenged individuals in Australia and found the rate of injury hospitalizations in these individuals to be twice that of the general population.14
Other studies also indicated that children with disabilities were over two times more likely to experience an injury that needs medical attention compared with their healthy controls 
. Findings from this study are consistent with these studies using parent-reported or caregiver-reported data on the association between injury and disability.
Consistent with results from previous studies, we found that a higher percentage of injuries among children with disabilities was caused by falls and occurred inside the home. In addition, a higher proportion of children with disabilities were injured when they are sleeping or resting 
. These findings suggested that the strategies to minimize the risk of falls and to prevent injuries occurred in residential areas among children with disabilities are of great importance. Children with disabilities suffered residential injuries more often than children without a disability probably because they are less likely to participate in activities outside the home compared with their healthy peers due to their activity limitation and participation restrictions caused by their disability.
Extraordinary numbers of children worldwide suffer from preventable injuries that occur within their home 
. It is estimated that approximately 12 million non-fatal, medically treated injuries occur in and around the home annually 
. In this study, we found that above 50% pediatric injuries happened inside the home in this sample of Chinese children. There are evidence that falls, burns, unintentional cuts, poisonings and animal bites are especially common for young children in low- and middle income countries, and that these injuries often happen at home 
. Home environment is one of the most significant contextual factors of injury risk for young children 
. Hazards such as open fires, owning cats/dogs, fireworks stored unlocked, and slippery floor are very common in children's homes in China, and all these contribute to young children's high levels of unintentional injury risk. However, in the present study, we only found that child whose family had cat/dog(s) was over 3 times more likely to be injured, compared with child whose family did not have any cat/dog during the last 12 months after adjustment for sociodemographic factors and disability status. The rest of 7 physical home environmental risks did not have significant association with pediatric injuries. Future work should explore these risks in larger samples, and identify whether and how they might translate into actual injury incidence and examine which disabilities are associated with the greater risk of injury. Although some studies suggest that children from larger families are at higher risk of injury 
, our findings suggested a reversed association which was consistent with a previous research that children with less family members are at higher risk of injury 
. This is probably because there are strong family ties and a phenomenon that older children and other family members would watch over the younger child in Chinese family.
Like all research, our study has some limitations. First, the definition of disability in our study was based on the new ICF which is quite broad, and includes a very broad range of children with a very broad range of types of disabilities. Future study might consider specific types of disabilities individually (e.g., mental retardation, wheelchair-users, etc). Second, our analysis results are limited by a small sample size because of the comparatively low prevalence of injury among children with and without disabilities (10.2% and 4.4%, respectively). The injury definition in our study is different from the previous studies in China. Injuries defined in the previous studies included: (1) medically attended injury; or (2) injury leading children off school more than half a day; or (3) first-aid by family members or teachers or friends 
. Previous studies reported that the prevalence of injury among healthy children aged 1 to 14 was 5.3~11.3% 
. In our study, the prevalence of injury (4.4%) among nondisabled children is lower than those reported in previous studies. However, if we counted injuries treated by family members or teachers or friends, the prevalence of injury among children with and without disabilities is 17.5% and 8.6%, respectively. In order to compare our findings with previous studies conducted around the world, we choose to report only medically treated injuries in our study. Another reason of low prevalence of injury both among children with and without disabilities is that caregivers' medical care-seeking behavior when their children injured is extremely low in China, especially in rural areas where medical service is not well developed. Unfortunately, a larger percent of children with disabilities are in rural counties in China so that the majority of respondents in our study are from those underdeveloped areas. Third, there was no information about validity of self-report of time per day that children were supervised, and some of the other measures used in the study. This limitation is inherent in the self-reported surveys. And also, there were no data on self-inflicted injuries. Injury were reported from primary caregiver and were not validated by medical records is another limitation of our study. Previous studies indicated that underestimation of injury episodes were likely to happen when using a 12-month recall period 
. Information bias is also possible if parents of children with disabilities systematically reported injuries differently than parents of normal children. Parents of children with disabilities may tend to request medical attention more often than other parents when an injury happens to their child. Further research is needed to identify how disability status influences recall bias in injury reporting by caregivers. Finally, one would expect to see an association between age and injury risk because children have different level of injury risk at different developmental stages. Our study did not find an association between developmental age and injury rate. This could be due to small sample size in age subgroups. In addition, our results indicated non-significant differences with respect to certain subgroups of age, education, and income between children with and without disability, which may due to the small numbers for each of these. A larger study would probably find statistical differences.
Despite these limitations, our findings have implications for injury prevention targeting children with disabilities. Parents and other caretakers of children with disabilities should receive additional education on home safety and injury prevention. Injury prevention needs to be emphasized for the sake of reducing the excess incidence of injury among children with disabilities. In addition, nurses, pediatricians and other healthcare professionals should be aware of the significantly high risk for injury in children with disabilities. They could plan an important role in educating parents and caregivers about safety and injury prevention when they provide medical services to children with disabilities. Furthermore, in view of a higher proportion of residential injuries, intervention strategies need to be initiated to improve the safety of children's home environments. Although every child is susceptible to unintentional injury, children with disabilities not only have a specific vulnerability, but they are more likely to stay at home and therefore are at significant higher risk of residential injuries than children without disabilities. Interventions to prevent residential injuries are an important public health priority in children with disabilities in the People's Republic of China.