Unintentional Injury Predicton Among Children aged 0-10 years old in Shenzhen, China: Based on Machine Learning Models

Unintentional Injury Predicton Among Children aged 0-10 years old in Shenzhen, China: Based on Machine Learning Models | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Unintentional Injury Predicton Among Children aged 0-10 years old in Shenzhen, China: Based on Machine Learning Models Shaoyun Chen, Ying Xu, Nongping Feng, Sijing Cao, Yongqi Huo, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7260288/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract Background Unintentional injury is a leading cause of death and years of healthy life lost due to disability among children. This study aimed to develop machine learning (ML) models to predict the occurrence of unintentional injury in children aged 0–10 years in Shenzhen, China, and to identify the associated influencing factors. Methods This cross-sectional study recruited 15,796 children aged 0–10 years in Shenzhen during 2017–2018. Information on the children and guardians were collected, and unintentional injuries in the past year was examined by using two nested questions. The dataset was randomly partitioned into training and test sets at a 7:3 ratio. Class imbalance in the training set was addressed using data balancing techniques. Subsequently, seven machine learning algorithms were employed to establish a pediatric injury risk prediction model. The top five predictors for injury were ultimately identified based on the optimal performing model. Results Among 15,796 children surveyed, 915 (5.79%) experienced at least one unintentional injury within the past year. Falls constituted a leading cause of injuries, with 53.06% of such incidents occurring at home. On the training set, The Random Forest (RF) model integrated with SMOTE demonstrated optimal performance, achieving an AUC of 0.986, sensitivity of 0.955, and specificity of 0.988. On the test set, the prediction effect of the Extreme Gradient Boosting model combined with Random Under-Sampling (RUS) outperformed other algorithms, yielding an AUC of 0.593, sensitivity of 0.540, and specificity of 0.582. The children’s grade was the most important predictor of child injury. Conclusions Our research indicated that ML models, when coupled with data balancing techniques, prove to be potent instruments for predicting child injury in scenarios involving imbalanced datasets. Child injury Machine learning Prediction Risk factors Prevention Figures Figure 1 Figure 2 Figure 3 Figure 4 Background Injury is a global public health challenge that involves various age groups and different gender groups[ 1 ]. Existing research on trends in population health has demonstrated that injury is a substantial contributor to the overall loss of population health[ 2 ]. Injury has emerged as the fifth leading cause of death among people in China[ 1 ]. It is estimated that injury causes around five million fatalities annually, accounting for 9% of total deaths worldwide[ 3 , 4 ]. Similarly, child injury is a public health and social problem to be solved urgently, and more than 95% of deaths from child injuries occur in low-income and middle-income countries[ 5 ]. However, injury is an important preventable and controllable cause of premature death[ 6 ]. The current challenges in the field of child injury prevention include the diverse types of injury, the complexity of their causes, and the difficulty in identifying high-risk groups in advance. Hence, it is crucial to develop machine learning (ML) models capable of predicting child injury in order to facilitate monitoring and enhance the effectiveness of preventative measures. In the medical field, artificial intelligence (AI) has emerged as a critical tool for tapping into the profound value of diagnostic and treatment data. Machine learning (ML) constitutes a key and central process, aimed at “learning” from past data to form a model that can be applied to predict future events, particularly for data analysis of non-linear associations[ 7 , 8 ]. Compared to classical regression prediction models, ML models have relatively less strict distribution of data and can handle large-scale and complex data, especially unstructured data (images, text). Meanwhile, through techniques such as automatic feature selection, feature combination, and cross validation, machine learning models can achieve good predictive performance on unknown data. Some studies reported that machine learning algorithms were able to manage more predictors and outperform classical regression models[ 9 , 10 ]. The methods used are primarily supervised and unsupervised learning. Representative supervised learning algorithms include Logistic Regression (LR), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), Naïve Bayes (NB), K-Nearest Neighbors (KNN), and Extreme Gradient Boosting (XGBoost), while representative unsupervised learning algorithms include K-Means[ 11 ]. In recent years, supervised learning algorithms have been widely applied and extensively investigated in medicine and healthcare[ 12 ]. The study by Tiruneh1[ 13 ] showed that RF and boosting-type algorithms had the better predictive performance comparing to classical regression models in predicting pre-eclampsia. Researchers had also conducted a systematic review of ML models for neurological disorder diagnosis based on genetic and molecular pathways, emphasizing the need to promote existing models in real-world clinical settings[ 14 ]. ML models have been applied to predict fall occurrence in older adults, highlighting the importance of incorporating fall history, physical function, psychological factors, and home environment into prevention strategies[ 15 ]. However, the application of ML models in the field of child unintentional injury research is still uncommon. In medical datasets, cases of certain diseases are typically far fewer than those of healthy populations, leading to suboptimal performance of machine learning models in predicting these minority-class diseases[ 16 ]. Data balancing techniques aim to enhance the model’s identification performance for minority classes by adjusting the proportion of samples across different categories, thereby improving overall prediction performance[ 17 ]. Data balancing techniques, including random over-sampling (ROS), random under-sampling (RUS), and the synthetic minority over-sampling technique (SMOTE), have emerged as critical approaches for addressing imbalanced datasets in disease prediction[ 18 ]. ROS balances datasets by simply replicating minority-class instances[ 19 ]. RUS achieves data balance by randomly removing majority-class samples[ 19 ]. SMOTE is an advanced oversampling technique that balances data by synthesizing new minority-class samples[ 19 ]. Unlike simple replication, SMOTE generates novel synthetic samples through interpolation based on existing minority-class features, thereby mitigating overfitting to some extent[ 20 ]. Research demonstrates that applying data balancing techniques enhances model identification performance for individuals at risk of cardiovascular diseases and diabetic patients[ 18 , 21 ]. Through prudent selection and application of these techniques, models can achieve significant improvement in minority-class identification performance, thereby strengthening diagnostic efficacy for early-stage disease detection and therapeutic intervention. The study aimed to build a predictive model for child injury in Shenzhen using ML methods, identify the influencing factors with and formulate corresponding prevention strategies. Methods Study Design This cross-sectional study was conducted in Shenzhen, as part of the “2016–2020 Child Injury Prevention Program Child Injury Special Survey Plan” by the National Center for Chronic and Noncommunicable Disease Control and Prevention of the Chinese Center for Disease Control and Prevention in China[22]. Participants were selected through multistage cluster sampling in Bao’an, Longgang, and Longhua Districts. Ultimately, 16 schools, 8 kindergartens and 8 communities were included in the investigation sites. Children aged 6-10 years and 3-6 years in this study were recruited from selected schools and kindergartens, respectively. Children under the age of 3 were randomly selected from the chosen communities. The inclusion criteria were as follows: (1) Children who had lived in Bao’an/Longgang/Longhua for a cumulative period of more than 6 months within the past 12 months, (2) Children aged 0 to 10 years, (3) Participation of guardians in the study. Based on the formula: “N=deff*Z 2 p(1-p)/d 2 ” (deff=1.5, Z 1-α/2 =1.96, p=16.5%[23], q=1-p=83.5%, d=0.20p, and 10% non-response rate),, the minimum sample size required was 11,490. Data Collection A self-administered questionnaire (Supplementary Material 1) including demographic characteristics, knowledge-attitude-practice (KAP) of injury prevention, and injury occurrence in the past 12 months was developed by the National Center for Chronic and Noncommunicable Disease Control and Prevention. We provided a standardized training programm to ensure that the children’s guardians understood the study and the questionnaire’s content and to confirm that they had appropriate interview skills. The questionnaires were completed by the children’s guardians with the assistance of trained personnel on site. Case definition Within the past 12 months, children who have suffered from and been diagnosed with a particular injury, or who have taken a leave of absence (from school or for rest) for more than one day as a result of that injury[24]. Models and Predictors LR, SVM, DT, RF, NB, KNN, and XGBoost models were used to predict child injury with 13 predictors, including administrative district, Children’s age, Children’s gender, children’s grade, permanent registered residence, living arrangement in kindergarten/school, primary caregiver and primary caregivers’ age, primary caregivers’ educational attainment, parents’ working status, parent-child communication duration, injury prevention knowledge and attitude score, injury prevention practice score. The detailed variable assignments were shown in Table 1. Table 1 Predictor assignment situation Predictor Assignment Administrative district 1=Bao’an, 2=Longgang, 3=Longhua Children’s Age 1=0-3, 2=4-6, 3=7-10 Children’s gender 1=Boy, 2=Girl Children’s grade 1=Not enrolled in kindergarten/school, 2=Kindergarten, 3=Grade 1, 4=Grade 2, 5=Grade 3 Registered permanent residence 1=Shenzhen, 2=Other cities in Guangdong, 3=Other provinces in China, 4=Foreign nationality Living arrangement in kindergarten/school 1=Not enrolled in kindergarten/school, 2=Boarding at kindergarten/school, 3=Attending a non-boarding, 4=kindergarten/school Parents’ working status 1=Father works away from home, 2=Mother works away from home, 3=Both parents work away from home, 4=Neither parent works away from home Primary caregiver 1=Parents, 2=Stepparents, 3=Grandparents, 4=Other Primary caregiver’s age 1=50 Primary caregiver’s educational attainment 1=Primary school or below, 2=Junior high school, 3=High school/technical school/technical secondary school, 4=Junior college, 5=Bachelor degree or above Parent-child communication duration 1=Never, 2=Within 10 minutes, 3=10-30 minutes, 4=30-60 minutes, 5=More than 60 minutes Knowledge and attitude score 1=<15, 2=≥15 Practice score 1=<50, 2=≥50 Model Derivation and Validation The data were randomly divided into a training set for model derivation and a test set for model validation in a 7:3 ratio. On the training set, ROS, RUS and SMOTE were used to address the imbalanced data between injury and non-injury participants(approximate 1:16). A 10-fold cross-validation strategy was applied for ML model derivation and validation. All participants were randomly divided into 10 subgroups with similar injury reporting incidence, 7 of which were combined as the training set, and the remaining part as the test set. This process was repeated 10 times, and finally, the means of the results were estimated to evaluate the prediction effect ( Fig.1 ). Statistical Analysis All data were statistically analyzed using R package version 4.2.2. Continuous variables were represented by mean±SD. Categorical variables were described using frequency and percentage. Student’s t -test, chi-square test, or Fisher exact test were used to compare participants between injury and non-injury groups. A two-sided test with a p-value of less than 0.05 was considered statistically significant. For variables with missing values<5%, mode/mean imputation was used. The area under the receiver operating characteristic curve (ROC)and the area under the curve (AUC)were used to compare the model’s ability to predict injury. At the same time, the performance of the model derivation was measured by its accuracy, sensitivity, specificity. The VarImp function was employed to evaluate the importance of all predictors. Higher scores indicate more influential predictors. Results Characteristics of Participants This study conducted a comparative analysis of demographic characteristics and family factors between 915 injured children and 14,881 non-injured children, revealing statistically significant differences ( P <0.05) across multiple indicators (Table 2). Geographical variations were observed across administrative districts ( c 2 =8.804, P =0.012), with Longhua District showing the highest injury incidence rate (6.51%). Boys had a significantly higher incidence of injury than girls (6.36% vs 5.10%; c 2 =9.671, P =0.002). Injury rates varied substantially by educational stage ( c 2 =28.939, P <0.001), peaking in kindergarten attendees (7.08%). Children with foreign nationality registration demonstrated elevated injury risk (7.41%, c 2 =8.516, P =0.036), while boarding school arrangements were associated with a higher incidence of injury (7.26%, c 2 =10.009, P =0.007). Parental employment patterns revealed particular vulnerability when mothers worked remotely (9.50%, c 2 =18.683, P <0.001). A higher incidence of injury was associated with caregivers who had primary education or less (8.63%, c 2 =21.936, P <0.001). Inverse relationships emerged between injury and parent-child communication duration (13.33%, c 2 =32.088, P <0.001). Low knowledge/attitude scores (8.87%, c 2 =27.829, P <0.001) and poor practice scores (7.12%, c 2 =23.423, P <0.001) were correlated with a higher incidence of injury. Table 2 Characteristics of the study participants. Characteristics Injury (n=915) Non-injury (n=14,881) P -value Administrative district, n(%) 8.804 0.012 Bao’an 182(4.95%) 3498(95.05%) Longgang 460(5.81%) 7463(94.19%) Longhua 273(6.51%) 3920(93.49%) Children’s Age, n(%) 1.374 0.503 0-3 192(5.60%) 3239(94.40%) 4-6 435(5.68%) 7227(94.32%) 7-10 288(6.12%) 4415(93.88%) Children’s Gender, n(%) 9.671 0.002 Boy 554(6.36%) 8157(93.64%) Girl 361(5.10%) 6724(94.90%) Children’s Grade, n(%) 28.939 <0.001 Not enrolled in kindergarten/school 42(3.47%) 1170(96.53%) Kindergarten 303(7.08%) 3976(92.92%) Grade 1 176(5.15%) 3240(94.85%) Grade 2 192(5.41%) 3355(94.59%) Grade 3 202(6.04%) 3140(93.96%) Registered permanent residence, n(%) 8.516 0.036 Shenzhen 373(5.40%) 6531(94.60%) Other cities in Guangdong 199(5.46%) 3446(94.54%) Other provinces in China 317(6.47%) 4579(93.53%) Foreign nationality 26(7.41%) 325(92.59%) Living arrangement in kindergarten/school, n(%) 10.009 0.007 Not enrolled in kindergarten/school 70(4.35%) 1540(95.65%) Boarding at kindergarten/school 69(7.26%) 882(92.74%) Attending a non-boarding kindergarten/school 776(5.86%) 12459(94.14%) Parents’ working status, n(%) 18.683 <0.001 Father works away from home 116(7.48%) 1435(92.52%) Mother works away from home 17(9.50%) 162(90.50%) Both parents work away from home 130(6.62%) 1834(93.38%) Neither parent works away from home 652(5.39%) 11450(94.61%) Primary caregiver, n(%) — 0.002 Parents 768(5.64%) 12846(94.36%) Stepparents 4(25.00%) 12(75.00%) Grandparents 121(6.23%) 1821(93.77%) Other 22(9.82%) 202(90.18%) Primary caregiver’s age, n(%) 3.324 0.344 50 121(6.54%) 1729(93.46%) Primary caregiver’s educational attainment, n(%) 21.936 <0.001 Primary school or below 97(8.63%) 1027(91.37%) Junior high school 253(6.00%) 3962(94.00%) High school/technical school/technical secondary school 250(5.79%) 4067(94.21%) Junior college 163(5.21%) 2964(94.79%) Bachelor degree or above 152(5.04%) 2861(94.96%) Parent-child communication duration, n(%) 32.088 <0.001 Never 14(13.33%) 91(86.67%) Within 10 minutes 54(8.02%) 619(91.98%) 10-30 minutes 215(7.02%) 2849(92.98%) 30-60 minutes 213(5.60%) 3592(94.40%) More than 60 minutes 419(5.14%) 7730(94.86%) Knowledge and attitude score, n(%) 27.829 <0.001 <15 129(8.87%) 1325(91.13%) ≥15 786(5.48%) 13556(94.52%) Practice score, n(%) 23.423 <0.001 <50 354(7.12%) 4618(92.88%) ≥50 561(5.18%) 10263(94.82%) Epidemiological Characteristics of Injury Out of 15,796 participants, 915 experienced injuries in the past 12 months, accounting for a total of 1,076 injury episodes. The incidence of injury in children aged 0-10 years old in Shenzhen was 5.79%, while the cumulative injury incidence rate is 6.81%. Out of 790 participants experienced one injury, 102 participants experienced two injuries, and 23 participants experienced three or more injuries. Of these injuries, 45.72% and 53.07% were found in the afternoon (12:01–18:00), and at home, respectively. The top three causes of injury were falls (60.22%), blunt force trauma (11.06%), and knife or sharp force injuries (6.97%). Injuries to the head, upper limbs, and lower limbs accounted for 25.65%, 21.93%, and 29.93% of cases, respectively. Contusions and bruises were the most common types of injury, comprising 42.01% of cases. Of all injuries, 85.87% were successfully treated, while the remainder were either still undergoing treatment or resulted in conditions such as disability (Table 3). Table 3 Characteristics of the 1076 unintentional injuries cases Variables Number Proportion(%) Time of injury 00:01-06:00 64 5.95 06:01-12:00 260 24.16 12:01-18:00 492 45.73 18:01-24:00 260 24.16 Cause of injury Fall 648 60.22 Blunt force trauma 119 11.06 Knife or sharp force injury 75 6.97 Others 234 21.75 Location of injury Home 571 53.06 School/Kindergarten 224 20.82 Outside home/school/Kindergarten 281 26.12 Nature of injury Sprain/strain 146 13.57 Open injury 163 15.15 Contusion/abrasion 452 42.01 Others 315 29.27 Area of injury Head 276 25.65 Upper limb 236 21.93 Lower limb 322 29.93 Torso 48 4.46 Multi-area 29 2.70 Others 165 15.32 Result of injury During the treatment period 89 8.28 Recover 924 85.87 Physical disability 4 0.37 Others 59 5.48 Performance of ML Models Without applying any data balancing methods, all models achieved identical training and test set accuracy (0.942); however, their sensitivity was zero or near zero indicating a complete failure to detect the minority class despite perfect specificity. The overall performance of the machine learning methods improved, particularly in the classification accuracy of positive samples, after applying data balancing methods. The RF model consistently demonstrated the best training performance on the training set (AUC: 0.962-0.986). However, on the test set, the LR model with ROS, the XGBoost model with RUS, and the XGBoost model with SMOTE achieved the best predictive effects for injuryoccurrence (with AUCs of 0.586, 0.593, and 0.569 respectively). However, with SMOTE, the sensitivity and specificity of XGBoost model were 0.900 and 0.182, respectively (Table 4). The AUCs of the training set and test set of each model under three data balancing methods were shown in Fig.2 and Fig.3 . Table 4 Performance of ML models in different data sets. Model Balance method Training set Test set Accuracy Sensitivity Specificity AUC Accuracy Sensitivity Specificity AUC LR — 0.942 0.000 1.000 0.625 0.942 0.000 1.000 0.583 SVM — 0.942 0.000 1.000 0.500 0.942 0.000 1.000 0.500 DT — 0.942 0.000 1.000 0.500 0.942 0.000 1.000 0.500 RF — 0.942 0.000 1.000 0.758 0.942 0.000 1.000 0.576 NB — 0.942 0.000 1.000 0.602 0.942 0.000 1.000 0.579 KNN — 0.942 0.003 1.000 0.792 0.942 0.004 1.000 0.519 XGBoost — 0.942 0.000 1.000 0.619 0.942 0.000 1.000 0.572 LR ROS 0.597 0.599 0.596 0.629 0.595 0.511 0.600 0.586 SVM ROS 0.591 0.624 0.559 0.591 0.584 0.536 0.587 0.562 DT ROS 0.559 0.460 0.658 0.564 0.438 0.672 0.424 0.565 RF ROS 0.933 0.972 0.893 0.962 0.818 0.131 0.860 0.542 NB ROS 0.546 0.255 0.837 0.587 0.770 0.292 0.800 0.582 KNN ROS 0.817 0.972 0.662 0.954 0.598 0.394 0.610 0.496 XGBoost ROS 0.727 0.743 0.710 0.809 0.675 0.383 0.692 0.560 LR RUS 0.609 0.593 0.626 0.647 0.565 0.544 0.566 0.588 SVM RUS 0.612 0.577 0.647 0.612 0.603 0.496 0.609 0.553 DT RUS 0.562 0.457 0.666 0.564 0.575 0.511 0.579 0.539 RF RUS 0.945 0.930 0.961 0.982 0.630 0.460 0.640 0.569 NB RUS 0.500 1.000 0.000 0.614 0.084 0.978 0.029 0.567 KNN RUS 0.657 0.633 0.680 0.724 0.591 0.471 0.598 0.544 XGBoost RUS 0.612 0.604 0.619 0.637 0.579 0.540 0.582 0.593 LR SMOTE 0.718 0.667 0.765 0.787 0.742 0.310 0.768 0.554 SVM SMOTE 0.720 0.661 0.776 0.718 0.762 0.288 0.791 0.539 DT SMOTE 0.653 0.655 0.651 0.680 0.636 0.445 0.648 0.550 RF SMOTE 0.972 0.955 0.988 0.986 0.752 0.307 0.780 0.565 NB SMOTE 0.660 0.475 0.833 0.757 0.814 0.245 0.849 0.560 KNN SMOTE 0.789 0.610 0.957 0.915 0.881 0.117 0.928 0.550 XGBoost SMOTE 0.861 0.793 0.925 0.923 0.859 0.182 0.900 0.569 AUC, area under the curve; DT, decision tree; KNN, k-nearest neighbors; LR, logistic regression; NB, naïve bayes; RF, random forest; ROS, random over-sampling; RUS, random under-sampling; SMOTE, synthetic minority over-sampling technique; SVM, support vector machine; XGBoost, extreme gradient boosting. Important Predictors for Injury According to the XGBoost model with RUS, the most important factors influencing injury among children aged 0-10 in Shenzhen were children’s grade (with a score of 100.00), practice score (69.87), children’s age (56.64), administrative district (39.66), and registered permanent residence (37.94), respectively (Fig.4) . Discussion Injury in childhood contribute substantially to global mortality and morbidity rates, with children residing in low- and middle-income countries being particularly susceptible[ 25 ]. According to epidemiological survey data, the incidence of child injury in Shenzhen (5.79%) was lower than that in Shanghai (9.3%) and Suzhou (7.7%)[ 26 , 27 ]. Compared to specific high-risk groups, Shenzhen’s rate is significantly lower than Guangzhou’s preschool-aged children (34.24%), and left-behind children in rural poverty-alleviation regions (15.6%)[ 28 , 39 ]. In cross-national comparisons, child injury incidence in Shenzhen exceeded that of Nepal’s Makwanpur District (2.46%), Thailand (1.5%), and India (3.42%)[ 30 , 31 ]. Regional disparities arose from differences in survey sample sources, survey periods, and research scopes. Falling was the primary cause. These findings were similar to other studies[ 32 , 33 ]. Another significant fact was fall as the primary contributor to traumatic brain injury (TBI) in children, comprising 90% of TBI cases in children aged 0 to 4 years and 42% of those in children between 5 and 14 years old[ 34 ]. Child injury constitutes a critical public health issue due to its substantial disease burden and economic impact, necessitating effective prevention and control measures.. In our study, the injury reporting incidence was low (imbalance). The distribution of classes in datasets was highly unequal. It is a common challenge in machine learning, particularly in classification problems. Resampling techniques and synthetic data generation were used in our analysis to handle these data[ 35 , 36 ]. Among them, ROS was performed by randomly duplicating samples in a minority class to match the number of samples in a majority class[ 35 ]. SMOTE is easy to synthesize “noisy sample”and“boundary sample”[ 37 ]. Findings emphasized the importance of balancing data in ML applications. Our comprehensive evaluation of the results-based on accuracy, AUC, sensitivity, specificity, indicated that the RF model with SMOTE outperformed the other ML models (AUC = 0.986, sensitivity = 0.955, specificity = 0.988) on training set. RF is one of the state-of-the-art ML methods for developing predictive models. It is a non-parametric methodology capable of handling diverse response types, encompassing categorical, quantitative outcomes, as well as survival times[ 38 ]. RF excels in machine learning due to its ensemble learning framework, which aggregates multiple decision trees through bagging and random feature selection, effectively reducing overfitting while maintaining high predictive accuracy[ 39 ]. It demonstrates robustness to high-dimensional data and noise by automatically evaluating feature importance, enabling efficient handling of complex datasets without requiring prior feature scaling[ 39 ]. For instance, in a study comparing supervised ML algorithms, a total of 48 articles were included for disease prediction, among which 15 applied the RF, and 9 of these achieved the highest accuracy[ 40 ]. On the test set, the XGBoost model combined with RUS (AUC = 0.593, sensitivity = 0.540, specificity = 0.582) and the LR model combined with ROS (AUC = 0.586, sensitivity = 0.511, specificity = 0.600) were superior to other ML models. The XGBoost model also demonstrated excellent performance in the risk prediction of coronary heart disease[ 41 ]. The discrepancy between the training set and test set results may arise from data balancing applied to the training set while the test set retains its original imbalanced distribution. Additionally, the suboptimal predictive performance could be attributed to the limited number of variables included in the model, particularly the absence of critical factors such as household environment characteristics and external environment variables. The analysis of predictor importance revealed that the children’s grade was the most significant factor influencing child injury, And the other important predictors included practice score, children’s age, administrative district, and registered permanent residence. These findings emphasized the need for comprehensive injury prevention programs that target not only children but also their caregivers, focusing on increasing the time spent together, improving communication, and addressing the unique challenges faced by families across diverse socioeconomic and demographic backgrounds. Modern injury prevention and control aims to prevent, limit, or control injuries through the 4 Es of injury prevention: engineering, enforcement, education, and economic[ 42 ]. The results of this study had important implications for the development of effective injury prevention strategies in Shenzhen. First, targeted interventions aimed at improving working status, the educational level and safety knowledge of primary caregivers, particularly in low-income and under-resourced communities, could help reduce the incidence of child injury. Second, promoting regular and meaningful communication between parents and children, especially regarding safety issues, could also contribute to injury prevention. Third, the appropriate use of ML models facilitated the identification of high-risk groups and informed the allocation of resources and interventions. This study had several limitations. First, The generalization of findings was limited due to the study being conducted in a developed megacity. Multi-center research in the future should be considered to validate the results in diverse populations. Second, Fewer variables could lead to lower predictive performance in our study. More variables should be designed and included.. Finally, while the ML models demonstrated promising predictive performance, further validation in an independent dataset is necessary to confirm their robustness and generalizability. Conclusions Although the injury reporting incidence in Shenzhen was relatively low, it remains a challenge in the field of public health. By identifying key predictors and developing effective ML models, the factors influencing child injury became clearer, enabling the development of targeted prevention strategies. Abbreviations AI Artificial intelligence AUC, area under the curve DT Decision tree KNN K-nearest neighbors LR Logistic regression ML Machine learning NB Naïve bayes RF Random forest ROC Receiver operating characteristic curve ROS Random over-sampling RUS Random under-sampling SMOTE Synthetic minority over-sampling technique SVM Support vector machine XGBoost Extreme gradient boosting Declarations Ethics approval and consent to participate Ethical approval was granted by the Ethics Committee of National Center for Chronic and Noncommunicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention (No. 201713). The study adhered to the Declaration of Helsinki, with written informed consent obtained from all participants and/or their legal guardians. Consent for publication Not applicable. Data availability The datasets used and/or analyzed during the current study are available from the corresponding authors on reasonable request. Competing interests The authors declare that they have no competing interests. Funding This study was supported by United Nations International Children’s Emergency Fund (Grant No. 2022XQLH115) and the District-Level Scientific Research Project of Shenzhen Longhua District Medical Institutions (Grant No. 2023048). Authors’ contributions XF and CN designed the study. SY, XY and NP analyzed the data and wrote the manuscript. YQ, CN, JR and MX participated in on-site investigations and collected and organized data. SJ and JD helped perform the analysis with constructive discussions. LL revised the manuscript and edited the English language. All authors read and approved the final manuscript. Acknowledgements Gratitude is expressed to the United Nations International Children’s Emergency Fund (UNICEF), and National Center for Chronic and Noncommunicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention for financial support, and to all child and their guardians with their involvement. Author details 1 Longhua Center for Chronic Disease Control, Shenzhen, Guangdong 518110, China 2 Baoan Chronic Diseases Prevent and Cure Hospital, Shenzhen, Guangdong 518102, China 3 Longgang Center for Chronic Disease Control, Shenzhen，Guangdong 518100，China 4 Shenzhen Center for Chronic Disease Control, Shenzhen 518020, China Supplementary Information The questionnaire used to collect data in this study was compiled by the National Center for Chronic and Noncommunicable Disease Control and Prevention. Sections of this questionnaire were utilized in the research by Xie et al.[43], and the full questionnaire is detailed in Supplementary Material 1 . References Duan LL, Ye PP, Haagsma JA, Jin Y, Wang Y, Er YL, Deng X, Gao X, Ji CR, Linhong Wang LH, et al. The burden of injury in China, 1990-2017: findings from the Global Burden of Disease Study 2017. Lancet Public Health. 2019; 4:e449-61. James SL, Castle CD, Dingels ZV, Fox JT, Hamilton EB, Liu Z, NL SR, Sylte DO, Henry NJ, LeGrand KE et al. Global injury morbidity and mortality from 1990 to 2017: results from the Global Burden of Disease Study 2017. Inj Prev. 2020; 26:i96-114. World Health Organization. Injuries and violence: the facts 2014. Geneva: World Health Organization. 2014. Tang EHM, Bedford LE, Yu EYT, Tse ETY, Dong W, Wu T, Cheung BMY, Wong CKH, Lam CLK. Unintentional Injury Burden in Hong Kong: Results from a Representative Population-Based Survey. Int J Environ Res Public Health. 2021; 18. Peden M, Oyegbite K, Ozanne-Smith J, Hyder AA, Branche C, Rahman A, Rivara F, Bartolomeos K. WHO Guidelines Approved by the Guidelines Review Committee. World Report on Child Injury Prevention.2008. Jullien S. Prevention of unintentional injuries in children under five years. BMC Pediatr. 2021; 21:311. Huang S, Cai N, Pacheco PP, Narrandes S, Wang Y, Xu W. Applications of Support Vector Machine (SVM) Learning in Cancer Genomics. Cancer Genomics Proteomics. 2018; 15:41-51. Clift AK, Dodwell D, Lord S, Petrou S, Brady M, Collins GS, Hippisley-Cox J. Development and internal-external validation of statistical and machine learning models for breast cancer prognostication: cohort study. Bmj. 2023; 381:e073800. Goldstein BA, Navar AM, Carter RE. Moving beyond regression techniques in cardiovascular risk prediction: applying machine learning to address analytic challenges. Eur Heart J. 2017; 38:1805-14. Beam AL, Kohane IS. Big Data and Machine Learning in Health Care. Jama. 2018; 319:1317-18. Fu P, Lu J. Research progress of artificial intelligence in ischemic stroke imaging. Chin Geriatr Heart Brain Vessel Dis. 2020;22:1110-12. Krishnan R, Rajpurkar P, Topol EJ. Self-supervised learning in medicine and healthcare. Nat Biomed Eng. 2022; 6:1346-52. Tiruneh SA, Vu TTT, Rolnik DL, Teede HJ, Enticott J. Machine Learning Algorithms Versus Classical Regression Models in Pre-Eclampsia Prediction: A Systematic Review. Curr Hypertens Rep. 2024; 26:309-23. Aljarallah NA, Dutta AK, Sait ARW. A Systematic Review of Genetics- and Molecular-Pathway-Based Machine Learning Models for Neurological Disorder Diagnosis. Int J Mol Sci. 2024; 25. Chen X, He L, Shi K, Wu Y, Lin S, Fang Y. Interpretable Machine Learning for Fall Prediction Among Older Adults in China. Am J Prev Med. 2023; 65:579-86. Gentili E, Franchini G, Zese R, Alberti M, Ferrara M, Domenicano I, Grassi L. Machine learning from real data: A mental health registry case study. CMPB Update. 2024; 5:100132. Jadhav A, M. Mostafa SM, Elmannai H, Karim FK. An Empirical Assessment of Performance of Data Balancing Techniques in Classification Task. Appl Sci. 2022; 12:3928. Hasanah U, Soleh AM, Sadik K. Effect of Random Under sampling, Oversampling, and SMOTE on the Performance of Cardiovascular Disease Prediction Models. JMSK. 2024; 21:88-102. Khan AA, Chaudhari O, Chandra R. A review of ensemble learning and data augmentation models for class imbalanced problems: Combination, implementation and evaluation. Expert Syst Appl. 2024; 244:122778. Zhou H, Tong J, Liu Y, Zheng K, Cao C: An oversampling FCM-KSMOTE algorithm for imbalanced data classification. J King Saud Univ-Com. 2024; 36:102248. Zhang L, Wang RY, Yang H, Zhu SL. Application of resampling technique in the classification of imbalanced diabetes data in middle-aged and elderly residents. Mod Prev Med. 2023; 50:1339-44. Geng X. 2016-2020 Childhood Injury Prevention Program launched (in Chinese). Disease Surveillance. 2017; 32:485. Duan LL,Yang Y, Zhang R, Wu F. Analysis of childhood unintentional injury situation in the three municipalities in China. Chin J Health Educ. 2006; 23:248-50. Duan LL, Wang LH. Theory and methods for injury and violence prevention and control. People's Medical Publishing House. 2020. Tupetz A, Friedman K, Zhao D, Liao H, Isenburg MV, Keating EM, Vissoci JRN, Staton CA. Prevention of childhood unintentional injuries in low- and middle-income countries: A systematic review. PLoS One. 2020; 15:e0243464. Peng JJ, Gao N, Yu Y, Zhou DD, Su HJ, Xu NT, Shi Y, Zhong WJ. Epidemiological survey on non-fatal injuries among students in Shanghai.Shanghai J Prev Med. 2018; 30:723-9. Dai NB, Wang J, Gong T, Huang QL. Analysis of injury epidemiological characteristics in children aged 0-14 years in Suzhou. Chin J Dis Control Prev. 2019; 23:299-303. Wei YH, Wang DH, Jing QL, Lu JY, Hu HH, Chang HM, Liu WJ. Epidemiological analysis of primary students unintentional injury in Guangzhou area. J Med Pest Control. 2014; 30:711-13. He JZ, Chen YL, Liu YF, Du WJ. Analysis on the influencing factors of injuries among left-behind children in the first grade of junior middle. Chin J Health Educ. 2022; 38:438-46. Pant PR, Towner E, Ellis M, Manandhar D, Pilkington P, Mytton J. Epidemiology of Unintentional Child Injuries in the Makwanpur District of Nepal: A Household Survey. Int J Environ Res Public Health. 2015; 12:15118-28. Pant PR, Towner E, Pilkington P, Ellis M. Epidemiology of unintentional child injuries in the South-East Asia Region: a systematic review. Int J Inj Contr Saf Promot. 2015; 22:24-32. Bao W, Li H, Li N, Sun JX, Zhang XL, An Hl. Analysis of children injuries in Hebei Province from 2011 to 2015. Chin J Sch Health. 2018; 39:709-12+15. Centers for Disease Control and Prevention. Injury prevention and control: WISQARS injury statistics. 2019. Taylor CA, Bell JM, Breiding MJ, Xu L. Traumatic Brain Injury-Related Emergency Department Visits, Hospitalizations, and Deaths - United States, 2007 and 2013. MMWR Surveill Summ. 2017; 66:1-16. He H, Garcia EA. Learning from Imbalanced Data. IEEE Trans. 2009; 21:1263-84. Chawla NV, Bowyer KW, Hall LO, et al. SMOTE: synthetic minority oversampling technique. J Artif Intell Res. 2002; 16:321-57. Han H, Wang WY, Mao BH. Borderline-SMOTE: a new over-sampling method in imbalanced datasets learning.International conference on intelligent computing, Springer, Berlin, Heidelberg. 2005. Hu J, Szymczak S. A review on longitudinal data analysis with random forest. Brief Bioinform. 2023; 24. Breiman L, Cutler A. Random Forests. Berkeley: University of California. 2004. Uddin S, Khan A, Hossain ME, Moni MA. Comparing different supervised machine learning algorithms for disease prediction. BMC Med Inform Decis Mak. 2019; 19:281. Yue HT, He CC, Cheng YY, Zhang SC, Wu Y, Ma J. Coronary heart disease risk prediction model based on machine learning. Chin Gen Pract. 2025; 28:499-509. Mace SE, Gerardi MJ, Dietrich AM, Knazik SR, Mulligan-Smith D, Sweeney RL, Warden CR. Injury prevention and control in children. Ann Emerg Med. 2001; 38:405-14. Xie Y, Yu XY, Wu XY, Zhang WY, Feng ZL, Xiao F, Deng X, Dai WJ, Zhao SJ. Association between the guardians'educational levels and unintentional injuries in children aged 6-18 in Shenzhen, China. BMC Public Health. 2024; 24:2344. Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial1.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 26 Sep, 2025 Reviewers invited by journal 25 Sep, 2025 Editor assigned by journal 14 Aug, 2025 Editor invited by journal 11 Aug, 2025 Submission checks completed at journal 11 Aug, 2025 First submitted to journal 11 Aug, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7260288","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":525561926,"identity":"39596ee9-0b69-452d-bffd-4c1ff0a0f7dc","order_by":0,"name":"Shaoyun Chen","email":"","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Shaoyun","middleName":"","lastName":"Chen","suffix":""},{"id":525561927,"identity":"d21520fd-0ce4-4c70-912c-0a90211b4e6d","order_by":1,"name":"Ying Xu","email":"","orcid":"","institution":"Baoan Chronic Diseases Prevent and Cure Hospital","correspondingAuthor":false,"prefix":"","firstName":"Ying","middleName":"","lastName":"Xu","suffix":""},{"id":525561928,"identity":"1e5962f0-213f-44b2-836a-6f5746d4ccf1","order_by":2,"name":"Nongping Feng","email":"","orcid":"","institution":"Longgang Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Nongping","middleName":"","lastName":"Feng","suffix":""},{"id":525561929,"identity":"d2cd89be-9578-4eaf-9c66-951bad294cc2","order_by":3,"name":"Sijing Cao","email":"","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Sijing","middleName":"","lastName":"Cao","suffix":""},{"id":525561930,"identity":"0f65741f-73c1-4352-8e66-e9693ebb188b","order_by":4,"name":"Yongqi Huo","email":"","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Yongqi","middleName":"","lastName":"Huo","suffix":""},{"id":525561931,"identity":"f933e4dc-d41f-42ac-b2cc-e3c605338d9e","order_by":5,"name":"Jiandong Chen","email":"","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Jiandong","middleName":"","lastName":"Chen","suffix":""},{"id":525561932,"identity":"b8578eaf-ef48-4538-8048-ceb8df456459","order_by":6,"name":"Chaonan Gu","email":"","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Chaonan","middleName":"","lastName":"Gu","suffix":""},{"id":525561933,"identity":"d0c05e73-d00e-4310-8021-d14961e874df","order_by":7,"name":"Jianrong Li","email":"","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Jianrong","middleName":"","lastName":"Li","suffix":""},{"id":525561934,"identity":"83f05ec3-de0f-442b-ad81-ec2b922b4aa1","order_by":8,"name":"Mingxia Sun","email":"","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Mingxia","middleName":"","lastName":"Sun","suffix":""},{"id":525561935,"identity":"48432edf-f3ff-46c7-b65f-fc9f9c20477f","order_by":9,"name":"Lin Lei","email":"","orcid":"","institution":"Shenzhen Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Lin","middleName":"","lastName":"Lei","suffix":""},{"id":525561936,"identity":"e18fec5b-cade-4437-b6b2-36e35eb0d243","order_by":10,"name":"Chuanning Yu","email":"","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":false,"prefix":"","firstName":"Chuanning","middleName":"","lastName":"Yu","suffix":""},{"id":525561937,"identity":"1efd8fc0-91e3-4ffb-9d9b-bc79b55c2535","order_by":11,"name":"Xinfeng Yan","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwklEQVRIiWNgGAWjYBACfmbm4z8+/qth5idai2R7W4LkDLZj7JINxGoxOHNGQZqHjZnf4ADR1tzIYTDg4WGTNj6evIHhR8U2wjoYZ+QeSJCQkDE2O/OsgLHnzG3CWpgl8hIOGBiwJZvdyDFgZmwjQgubRI5hQ0ICc/3mGcRq4eE5Y8xw4AAzs4EEsVok2NvSGBsbjjFLAP1ykCi/2B9mPsb8twEYle3JGx/8qCBCCxJIICFq4FpI1TEKRsEoGAUjBAAArh860PKGwC0AAAAASUVORK5CYII=","orcid":"","institution":"Longhua Center for Chronic Disease Control","correspondingAuthor":true,"prefix":"","firstName":"Xinfeng","middleName":"","lastName":"Yan","suffix":""}],"badges":[],"createdAt":"2025-07-31 09:08:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7260288/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7260288/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":93062532,"identity":"2ff1a132-7e8b-4fa5-a569-6945be3f8c3d","added_by":"auto","created_at":"2025-10-08 16:37:59","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":695769,"visible":true,"origin":"","legend":"","description":"","filename":"RevisedUnintentionalInjuryPredictonAmongChildrenaged010yearsoldinShenzhenChinaBasedonMachineLearningModels.docx","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/ffee09756ff95ce929569fd8.docx"},{"id":93062846,"identity":"8e9f8000-7406-4ed3-a04c-9ea7562a0458","added_by":"auto","created_at":"2025-10-08 16:38:04","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12443,"visible":true,"origin":"","legend":"","description":"","filename":"faaa7f3ab2da49b5819934c7b6aa7b33.json","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/2de470ae5dfd62e79e16d184.json"},{"id":93062528,"identity":"7bad1f2e-643c-4bbd-a7d8-ab6f5d812d6c","added_by":"auto","created_at":"2025-10-08 16:37:59","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":45102,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial1.docx","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/72e759525fa2423f2a4c61cb.docx"},{"id":93062614,"identity":"b2033691-3b7a-4423-93cc-13658e884fa2","added_by":"auto","created_at":"2025-10-08 16:38:00","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":157035,"visible":true,"origin":"","legend":"","description":"","filename":"faaa7f3ab2da49b5819934c7b6aa7b331enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/4e9c6319b8d76b52a6a3e5f7.xml"},{"id":93062800,"identity":"cd0a92ef-42cc-47a0-bf69-938f11bca497","added_by":"auto","created_at":"2025-10-08 16:38:03","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":155622,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/a3bb1c5293582314828be6c4.jpeg"},{"id":93062640,"identity":"7e68c942-a8c9-4faa-a9e7-5bf971720471","added_by":"auto","created_at":"2025-10-08 16:38:01","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":192325,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/2401fc28d6395614c8475d9c.jpeg"},{"id":93062534,"identity":"026aa427-e355-4ed5-be9b-46790da744df","added_by":"auto","created_at":"2025-10-08 16:37:59","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":153078,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/d5d1857b9195c59f05bb9a44.jpeg"},{"id":93062435,"identity":"c58babf0-88a9-45a1-a50a-4e0bd3f9fa02","added_by":"auto","created_at":"2025-10-08 16:37:56","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":22764,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/5742f757a54065c3af422ebe.png"},{"id":93062583,"identity":"735741a7-423e-45f7-9492-5ef8884f1b7f","added_by":"auto","created_at":"2025-10-08 16:38:00","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":88974,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/ba9f6d5c1b78a3723e493ee5.png"},{"id":93062861,"identity":"4e6f8f2a-801c-406e-8098-fa13a83cbc0d","added_by":"auto","created_at":"2025-10-08 16:38:04","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":55537,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/719b1b31a9e0a575fbefb24e.png"},{"id":93062714,"identity":"b3a350e8-ce38-4d2a-95e4-52e07f14ac7e","added_by":"auto","created_at":"2025-10-08 16:38:02","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53454,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/3e93b5705303dc65fbd136ce.png"},{"id":93062520,"identity":"3ee93b2a-f3a7-425e-bfe7-6cb8b960a475","added_by":"auto","created_at":"2025-10-08 16:37:58","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11927,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/f14967a945f086d858c8b9af.png"},{"id":93063068,"identity":"93ba9e2c-ae60-4cf5-868a-b7ec67033ee2","added_by":"auto","created_at":"2025-10-08 16:38:07","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":155307,"visible":true,"origin":"","legend":"","description":"","filename":"faaa7f3ab2da49b5819934c7b6aa7b331structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/3a12393d206fc6dcf332c538.xml"},{"id":93062885,"identity":"0fa7cd57-e9e6-443e-9571-39b2a70350ba","added_by":"auto","created_at":"2025-10-08 16:38:04","extension":"html","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":163400,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/a3017cee720cb33392b2d8fa.html"},{"id":93063205,"identity":"7c469e02-e3ee-45fe-a30c-031b0d3bb46d","added_by":"auto","created_at":"2025-10-08 16:38:09","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":155622,"visible":true,"origin":"","legend":"\u003cp\u003eA flowchart for the study. AUC, area under the curve; DT, decision tree; KNN, k-nearest neighbors; LR, logistic regression; NB, naïve bayes; RF, random forest; ROS, random over-sampling; RUS, random under-sampling; SMOTE, synthetic minority over-sampling technique; SVM, support vector machine; XGBoost, extreme gradient boosting.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/727c19fbc031f2a9bd848fd5.jpeg"},{"id":93062726,"identity":"665d1bf0-002d-488d-800e-92892455ff04","added_by":"auto","created_at":"2025-10-08 16:38:02","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":192325,"visible":true,"origin":"","legend":"\u003cp\u003eROCs curve for ML models based on training data. \u003cstrong\u003eA\u003c/strong\u003e ROCs on ROS processing data. \u003cstrong\u003eB\u003c/strong\u003e ROCs on RUS processing data. \u003cstrong\u003eC\u003c/strong\u003e ROCs on SMOTE processing data.\u003c/p\u003e\n\u003cp\u003eAUC, area under the curve; DT, decision tree; KNN, k-nearest neighbors; LR, logistic regression; NB, naïve bayes; RF, random forest; ROC, receiver operating characteristic curve; ROS, random over-sampling; RUS, random under-sampling; SMOTE, synthetic minority over-sampling technique; SVM, support vector machine; XGBoost, extreme gradient boosting.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/87f4c7802950ef77753982de.jpeg"},{"id":93062526,"identity":"270cf08e-e438-4a75-a0c6-74cbda855fd7","added_by":"auto","created_at":"2025-10-08 16:37:59","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":153078,"visible":true,"origin":"","legend":"\u003cp\u003eROCs curve for ML models based on test data. \u003cstrong\u003eA\u003c/strong\u003eROCs on ROS processing data. \u003cstrong\u003eB\u003c/strong\u003e ROCs on RUS processing data. \u003cstrong\u003eC\u003c/strong\u003eROCs on SMOTE processing data.\u003c/p\u003e\n\u003cp\u003eAUC, area under the curve; DT, decision tree; KNN, k-nearest neighbors; LR, logistic regression; NB, naïve bayes; RF, random forest; ROC, receiver operating characteristic curve; ROS, random over-sampling; RUS, random under-sampling; SMOTE, synthetic minority over-sampling technique; SVM, support vector machine; XGBoost, extreme gradient boosting.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/f24a796552327ce3b8d54d53.jpeg"},{"id":93063073,"identity":"9cdd047f-9966-49e1-878d-9cdc2923e671","added_by":"auto","created_at":"2025-10-08 16:38:07","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":23123,"visible":true,"origin":"","legend":"\u003cp\u003ePredictors importance ranking for the XGBoost model based on RUS processing data.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/911ff789b58d5024e03524e4.png"},{"id":93064266,"identity":"ab12ad2e-1996-47ad-b226-a1e021e8089e","added_by":"auto","created_at":"2025-10-08 16:41:12","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1685363,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/1652cc4c-531f-4faf-bb9b-eaeda2c88707.pdf"},{"id":93062901,"identity":"7773f3ce-1b85-4a72-b166-0d2997ece74a","added_by":"auto","created_at":"2025-10-08 16:38:05","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":45102,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial1.docx","url":"https://assets-eu.researchsquare.com/files/rs-7260288/v1/8e69ecc7ec5b2ad2834020d7.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Unintentional Injury Predicton Among Children aged 0-10 years old in Shenzhen, China: Based on Machine Learning Models","fulltext":[{"header":"Background","content":"\u003cp\u003eInjury is a global public health challenge that involves various age groups and different gender groups[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Existing research on trends in population health has demonstrated that injury is a substantial contributor to the overall loss of population health[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Injury has emerged as the fifth leading cause of death among people in China[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. It is estimated that injury causes around five million fatalities annually, accounting for 9% of total deaths worldwide[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Similarly, child injury is a public health and social problem to be solved urgently, and more than 95% of deaths from child injuries occur in low-income and middle-income countries[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. However, injury is an important preventable and controllable cause of premature death[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. The current challenges in the field of child injury prevention include the diverse types of injury, the complexity of their causes, and the difficulty in identifying high-risk groups in advance. Hence, it is crucial to develop machine learning (ML) models capable of predicting child injury in order to facilitate monitoring and enhance the effectiveness of preventative measures.\u003c/p\u003e\u003cp\u003eIn the medical field, artificial intelligence (AI) has emerged as a critical tool for tapping into the profound value of diagnostic and treatment data. Machine learning (ML) constitutes a key and central process, aimed at \u0026ldquo;learning\u0026rdquo; from past data to form a model that can be applied to predict future events, particularly for data analysis of non-linear associations[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Compared to classical regression prediction models, ML models have relatively less strict distribution of data and can handle large-scale and complex data, especially unstructured data (images, text). Meanwhile, through techniques such as automatic feature selection, feature combination, and cross validation, machine learning models can achieve good predictive performance on unknown data. Some studies reported that machine learning algorithms were able to manage more predictors and outperform classical regression models[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. The methods used are primarily supervised and unsupervised learning. Representative supervised learning algorithms include Logistic Regression (LR), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), Na\u0026iuml;ve Bayes (NB), K-Nearest Neighbors (KNN), and Extreme Gradient Boosting (XGBoost), while representative unsupervised learning algorithms include K-Means[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. In recent years, supervised learning algorithms have been widely applied and extensively investigated in medicine and healthcare[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. The study by Tiruneh1[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] showed that RF and boosting-type algorithms had the better predictive performance comparing to classical regression models in predicting pre-eclampsia. Researchers had also conducted a systematic review of ML models for neurological disorder diagnosis based on genetic and molecular pathways, emphasizing the need to promote existing models in real-world clinical settings[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. ML models have been applied to predict fall occurrence in older adults, highlighting the importance of incorporating fall history, physical function, psychological factors, and home environment into prevention strategies[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. However, the application of ML models in the field of child unintentional injury research is still uncommon.\u003c/p\u003e\u003cp\u003eIn medical datasets, cases of certain diseases are typically far fewer than those of healthy populations, leading to suboptimal performance of machine learning models in predicting these minority-class diseases[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Data balancing techniques aim to enhance the model\u0026rsquo;s identification performance for minority classes by adjusting the proportion of samples across different categories, thereby improving overall prediction performance[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Data balancing techniques, including random over-sampling (ROS), random under-sampling (RUS), and the synthetic minority over-sampling technique (SMOTE), have emerged as critical approaches for addressing imbalanced datasets in disease prediction[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. ROS balances datasets by simply replicating minority-class instances[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. RUS achieves data balance by randomly removing majority-class samples[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. SMOTE is an advanced oversampling technique that balances data by synthesizing new minority-class samples[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Unlike simple replication, SMOTE generates novel synthetic samples through interpolation based on existing minority-class features, thereby mitigating overfitting to some extent[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. Research demonstrates that applying data balancing techniques enhances model identification performance for individuals at risk of cardiovascular diseases and diabetic patients[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Through prudent selection and application of these techniques, models can achieve significant improvement in minority-class identification performance, thereby strengthening diagnostic efficacy for early-stage disease detection and therapeutic intervention.\u003c/p\u003e\u003cp\u003eThe study aimed to build a predictive model for child injury in Shenzhen using ML methods, identify the influencing factors with and formulate corresponding prevention strategies.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eStudy Design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis cross-sectional study was conducted in Shenzhen, as part of the \u0026ldquo;2016\u0026ndash;2020 Child Injury Prevention Program Child Injury Special Survey Plan\u0026rdquo; by the National Center for Chronic and Noncommunicable Disease Control and Prevention of the Chinese Center for Disease Control and Prevention in China[22]. Participants were selected through multistage cluster sampling in Bao\u0026rsquo;an, Longgang, and Longhua Districts. Ultimately, 16 schools, 8 kindergartens and 8 communities were included in the investigation sites. Children aged 6-10 years and 3-6 years in this study were recruited from selected schools and kindergartens, respectively. Children under the age of 3 were randomly selected from the chosen communities. The inclusion criteria were as follows:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e(1) Children who had lived in Bao\u0026rsquo;an/Longgang/Longhua for a cumulative period of more than 6 months within the past 12 months,\u003c/p\u003e\n\u003cp\u003e(2) Children aged 0 to 10 years,\u003c/p\u003e\n\u003cp\u003e(3) Participation of guardians in the study.\u003c/p\u003e\n\u003cp\u003eBased on the formula:\u0026nbsp;\u0026ldquo;N=deff*Z\u003csup\u003e2\u003c/sup\u003ep(1-p)/d\u003csup\u003e2\u003c/sup\u003e\u0026rdquo;\u0026nbsp;(deff=1.5, Z\u003csub\u003e1-\u0026alpha;/2\u003c/sub\u003e=1.96, p=16.5%[23], q=1-p=83.5%, d=0.20p, and 10% non-response rate),, the minimum sample size required was 11,490.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Collection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA self-administered questionnaire (Supplementary Material 1) including demographic characteristics, knowledge-attitude-practice (KAP) of injury prevention, and injury occurrence in the past 12 months was developed by the National Center for Chronic and Noncommunicable Disease Control and Prevention. We provided a standardized training programm to ensure that the children\u0026rsquo;s guardians understood the study and the questionnaire\u0026rsquo;s content and to confirm that they had appropriate interview skills. The questionnaires were completed by the children\u0026rsquo;s guardians with the assistance of trained personnel on site.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCase definition\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWithin the past 12 months, children who have suffered from and been diagnosed with a particular injury, or who have taken a leave of absence (from school or for rest) for more than one day as a result of that injury[24].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModels and Predictors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLR, SVM, DT, RF, NB, KNN, and XGBoost models were used to predict child injury with 13 predictors, including administrative district, Children\u0026rsquo;s age, Children\u0026rsquo;s gender, children\u0026rsquo;s grade, permanent registered residence, living arrangement in kindergarten/school, primary caregiver and primary caregivers\u0026rsquo; age, primary caregivers\u0026rsquo; educational attainment, parents\u0026rsquo; working status, parent-child communication duration, injury prevention knowledge and attitude score, injury prevention practice score. The detailed variable assignments were shown in Table 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1\u0026nbsp;\u003c/strong\u003ePredictor assignment situation\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePredictor\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAssignment\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003eAdministrative district\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=Bao\u0026rsquo;an, 2=Longgang, 3=Longhua\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003eChildren\u0026rsquo;s Age\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=0-3, 2=4-6, 3=7-10\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003eChildren\u0026rsquo;s gender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=Boy, 2=Girl\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003eChildren\u0026rsquo;s grade\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=Not enrolled in kindergarten/school, 2=Kindergarten, 3=Grade 1, 4=Grade 2, 5=Grade 3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003eRegistered permanent residence\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=Shenzhen, 2=Other cities in Guangdong, 3=Other provinces in China, 4=Foreign nationality\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003eLiving arrangement in kindergarten/school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=Not enrolled in kindergarten/school, 2=Boarding at kindergarten/school, 3=Attending a non-boarding, 4=kindergarten/school\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003eParents\u0026rsquo; working status\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=Father works away from home, 2=Mother works away from home, 3=Both parents work away from home, 4=Neither parent works away from home\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003ePrimary caregiver\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=Parents, 2=Stepparents, 3=Grandparents, 4=Other\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 170px;\"\u003e\n \u003cp\u003ePrimary caregiver\u0026rsquo;s age\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 398px;\"\u003e\n \u003cp\u003e1=\u0026lt;30, 2=30-40, 3=40-50, 4=\u0026gt;50\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePrimary caregiver\u0026rsquo;s educational attainment\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1=Primary school or below, 2=Junior high school, 3=High school/technical school/technical secondary school, 4=Junior college, 5=Bachelor degree or above\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eParent-child communication duration\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1=Never, 2=Within 10 minutes, 3=10-30 minutes, 4=30-60 minutes, 5=More than 60 minutes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eKnowledge and attitude score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1=\u0026lt;15, 2=\u0026ge;15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePractice score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1=\u0026lt;50, 2=\u0026ge;50\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eModel Derivation and Validation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data were randomly divided into a training set for model derivation and a test set for model validation in a 7:3 ratio. On the training set, ROS, RUS and SMOTE were used to address the imbalanced data between injury and non-injury participants(approximate 1:16). A 10-fold cross-validation strategy was applied for ML model derivation and validation. All participants were randomly divided into 10 subgroups with similar injury reporting incidence, 7 of which were combined as the training set, and the remaining part as the test set. This process was repeated 10 times, and finally, the means of the results were estimated to evaluate the prediction effect (\u003cstrong\u003eFig.1\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data were statistically analyzed using \u003cem\u003eR\u003c/em\u003e package version 4.2.2. Continuous variables were represented by mean\u0026plusmn;SD. Categorical variables were described using frequency and percentage. Student\u0026rsquo;s \u003cem\u003et\u003c/em\u003e-test, chi-square test, or \u003cem\u003eFisher\u003c/em\u003e exact test were used to compare participants between injury and non-injury groups. A two-sided test with a p-value of less than 0.05 was considered statistically significant. For variables with missing values\u0026lt;5%, mode/mean imputation was used.\u003c/p\u003e\n\u003cp\u003eThe area under the receiver operating characteristic curve (ROC)and the area under the curve (AUC)were used to compare the model\u0026rsquo;s ability to predict injury. At the same time, the performance of the model derivation was measured by its accuracy, sensitivity, specificity. The VarImp function was employed to evaluate the importance of all predictors. Higher scores indicate more influential predictors.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eCharacteristics of Participants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study conducted a comparative analysis of demographic characteristics and family factors between 915 injured children and 14,881 non-injured children, revealing statistically significant differences (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05) across multiple indicators \u0026nbsp;(Table 2). Geographical variations were observed across administrative districts (\u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=8.804, \u003cem\u003eP\u003c/em\u003e=0.012), with Longhua District showing the highest injury incidence rate (6.51%). Boys had a significantly higher incidence of injury than girls (6.36% vs 5.10%; \u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=9.671, \u003cem\u003eP\u003c/em\u003e=0.002). Injury rates varied substantially by educational stage (\u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=28.939, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001), peaking in kindergarten attendees (7.08%). Children with foreign nationality registration demonstrated elevated injury risk (7.41%, \u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=8.516, \u003cem\u003eP\u003c/em\u003e=0.036), while boarding school arrangements were associated with a higher incidence of injury (7.26%, \u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=10.009, \u003cem\u003eP\u003c/em\u003e=0.007). Parental employment patterns revealed particular vulnerability when mothers worked remotely (9.50%, \u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=18.683, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001). A higher incidence of injury was associated with caregivers who had primary education or less (8.63%, \u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=21.936, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001). Inverse relationships emerged between injury and parent-child communication duration (13.33%, \u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=32.088, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001). Low knowledge/attitude scores (8.87%, \u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=27.829, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001) and poor practice scores (7.12%, \u003cem\u003ec\u003c/em\u003e\u003cem\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e=23.423, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001) were correlated with a higher incidence of injury.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2\u003c/strong\u003e Characteristics of the study participants.\u003c/p\u003e\n\u003cdiv align=\"center\"\u003e\n \u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"806\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eCharacteristics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003eInjury (n=915)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eNon-injury (n=14,881)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u003cimg width=\"15\" height=\"17\" src=\"https://myfiles.space/user_files/58895_8739fc6c57c1c19a/58895_custom_files/img1759918912.gif\" alt=\"image\"\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u003cem\u003eP\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eAdministrative district, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e8.804\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eBao\u0026rsquo;an\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e182(4.95%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e3498(95.05%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eLonggang\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e460(5.81%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e7463(94.19%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eLonghua\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e273(6.51%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e3920(93.49%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eChildren\u0026rsquo;s Age, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.374\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e0.503\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e0-3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e192(5.60%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e3239(94.40%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e4-6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e435(5.68%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e7227(94.32%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e7-10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e288(6.12%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e4415(93.88%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eChildren\u0026rsquo;s Gender, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e9.671\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eBoy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e554(6.36%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e8157(93.64%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eGirl\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e361(5.10%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e6724(94.90%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eChildren\u0026rsquo;s Grade, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e28.939\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eNot enrolled in kindergarten/school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e42(3.47%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1170(96.53%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eKindergarten\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e303(7.08%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e3976(92.92%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eGrade 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e176(5.15%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e3240(94.85%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eGrade 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e192(5.41%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e3355(94.59%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eGrade 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e202(6.04%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e3140(93.96%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eRegistered permanent residence, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e8.516\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e0.036\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eShenzhen\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e373(5.40%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e6531(94.60%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eOther cities in Guangdong\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e199(5.46%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e3446(94.54%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eOther provinces in China\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e317(6.47%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e4579(93.53%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eForeign nationality\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e26(7.41%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e325(92.59%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eLiving arrangement in kindergarten/school, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e10.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e0.007\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eNot enrolled in kindergarten/school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e70(4.35%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1540(95.65%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eBoarding at kindergarten/school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e69(7.26%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e882(92.74%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eAttending a non-boarding kindergarten/school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e776(5.86%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e12459(94.14%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eParents\u0026rsquo; working status, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e18.683\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eFather works away from home\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e116(7.48%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1435(92.52%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eMother works away from home\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e17(9.50%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e162(90.50%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eBoth parents work away from home\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e130(6.62%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1834(93.38%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eNeither parent works away from home\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e652(5.39%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e11450(94.61%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003ePrimary caregiver, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eParents\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e768(5.64%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e12846(94.36%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eStepparents\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e4(25.00%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e12(75.00%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eGrandparents\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e121(6.23%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1821(93.77%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eOther\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e22(9.82%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e202(90.18%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003ePrimary caregiver\u0026rsquo;s age, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e3.324\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e0.344\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e\u0026lt;30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e169(5.72%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e2786(94.28%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e30-40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e529(5.59%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e8935(94.41%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e40-50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e96(6.29%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1431(93.71%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e\u0026gt;50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e121(6.54%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1729(93.46%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003ePrimary caregiver\u0026rsquo;s educational attainment, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e21.936\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003ePrimary school or below\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e97(8.63%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1027(91.37%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eJunior high school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e253(6.00%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e3962(94.00%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eHigh school/technical school/technical secondary school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e250(5.79%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e4067(94.21%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eJunior college\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e163(5.21%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e2964(94.79%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eBachelor degree or above\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e152(5.04%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e2861(94.96%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eParent-child communication duration, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e32.088\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eNever\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e14(13.33%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e91(86.67%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eWithin 10 minutes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e54(8.02%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e619(91.98%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e10-30 minutes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e215(7.02%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e2849(92.98%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e30-60 minutes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e213(5.60%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e3592(94.40%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eMore than 60 minutes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e419(5.14%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e7730(94.86%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003eKnowledge and attitude score, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e27.829\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e\u0026lt;15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e129(8.87%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e1325(91.13%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e\u0026ge;15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e786(5.48%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e13556(94.52%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003ePractice score, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e23.423\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e\u0026lt;50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e354(7.12%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e4618(92.88%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 332px;\"\u003e\n \u003cp\u003e\u0026ge;50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 103px;\"\u003e\n \u003cp\u003e561(5.18%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e10263(94.82%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 91px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eEpidemiological Characteristics of Injury\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOut of 15,796 participants, 915 experienced injuries in the past 12 months, accounting for a total of 1,076 injury episodes. The incidence of injury in children aged 0-10 years old in Shenzhen was 5.79%, while the cumulative injury incidence rate is 6.81%. Out of 790 participants experienced one injury, 102 participants experienced two injuries, and 23 participants experienced three or more injuries. Of these injuries, 45.72% and 53.07% were found in the afternoon (12:01\u0026ndash;18:00), and at home, respectively. The top three causes of injury were falls (60.22%), blunt force trauma (11.06%), and knife or sharp force injuries (6.97%). Injuries to the head, upper limbs, and lower limbs accounted for 25.65%, 21.93%, and 29.93% of cases, respectively. Contusions and bruises were the most common types of injury, comprising 42.01% of cases. Of all injuries, 85.87% were successfully treated, while the remainder were either still undergoing treatment or resulted in conditions such as disability (Table 3).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 3\u003c/strong\u003e Characteristics of the 1076 unintentional injuries cases\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 375px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariables\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNumber\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProportion(%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eTime of injury\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003e00:01-06:00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e5.95\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003e06:01-12:00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e260\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e24.16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003e12:01-18:00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e492\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e45.73\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003e18:01-24:00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e260\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e24.16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCause of injury\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eFall\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e648\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e60.22\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eBlunt force trauma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e119\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e11.06\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eKnife or sharp force injury\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e6.97\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eOthers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e234\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e21.75\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eLocation of injury\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eHome\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e571\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e53.06\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eSchool/Kindergarten\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e224\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e20.82\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eOutside home/school/Kindergarten\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e281\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e26.12\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eNature of injury\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eSprain/strain\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e146\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e13.57\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eOpen injury\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e163\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e15.15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eContusion/abrasion\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e452\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e42.01\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eOthers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e29.27\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eArea of injury\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eHead\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e276\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e25.65\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eUpper limb\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e236\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e21.93\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eLower limb\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e322\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e29.93\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eTorso\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e4.46\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eMulti-area\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e2.70\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eOthers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e165\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e15.32\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eResult of injury\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eDuring the treatment period\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e8.28\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eRecover\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e924\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e85.87\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003ePhysical disability\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e0.37\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 225px;\"\u003e\n \u003cp\u003eOthers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 146px;\"\u003e\n \u003cp\u003e59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 162px;\"\u003e\n \u003cp\u003e5.48\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003ePerformance of ML Models\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWithout applying any data balancing methods, all models achieved identical training and test set accuracy (0.942); however, their sensitivity was zero or near zero indicating a complete failure to detect the minority class despite perfect specificity. The overall performance of the machine learning methods improved, \u0026nbsp;particularly in the classification accuracy of positive samples, after applying data balancing methods. The RF model consistently demonstrated the best training performance on the training set (AUC: 0.962-0.986). However, on the test set, the LR model with ROS, the XGBoost model with RUS, and the XGBoost model with SMOTE achieved the best predictive effects for injuryoccurrence (with AUCs of 0.586, 0.593, and 0.569 respectively). However, with SMOTE, the sensitivity and specificity of XGBoost model were 0.900 and 0.182, respectively (Table 4). The AUCs of the training set and test set of each model under three data balancing methods were shown in \u003cstrong\u003eFig.2 and Fig.3\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 4\u003c/strong\u003e Performance of ML models in different data sets.\u003c/p\u003e\n\u003cdiv align=\"center\"\u003e\n \u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"782\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBalance method\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"5\" style=\"width: 308px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTraining set\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"4\" style=\"width: 313px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTest set\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSensitivity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 83px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSpecificity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAUC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSensitivity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSpecificity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAUC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eLR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.625\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.583\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eSVM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.500\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.500\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eDT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.500\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.500\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.758\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.576\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eNB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.602\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.579\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eKNN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.003\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.792\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.004\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.519\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.619\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.942\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.572\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eLR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.597\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.599\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.596\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.629\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.595\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.511\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.600\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.586\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eSVM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.591\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.624\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.559\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.591\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.584\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.536\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.587\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.562\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eDT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.559\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.460\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.658\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.564\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.438\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.672\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.424\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.565\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.933\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.972\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.893\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.962\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.818\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.131\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.860\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.542\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eNB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.546\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.255\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.837\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.587\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.770\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.292\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.800\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.582\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eKNN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.817\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.972\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.662\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.954\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.598\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.394\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.610\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.496\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.727\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.743\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.710\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.809\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.675\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.383\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.692\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.560\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eLR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eRUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.609\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.593\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.626\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.647\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.565\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.544\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.566\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.588\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eSVM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eRUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.612\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.577\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.647\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.612\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.603\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.496\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.609\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.553\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eDT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eRUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.562\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.457\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.666\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.564\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.575\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.511\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.579\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.539\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eRUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.945\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.930\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.961\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.982\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.630\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.460\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.640\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.569\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eNB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eRUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.500\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e1.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.614\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.084\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.978\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.029\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.567\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eKNN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eRUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.657\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.633\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.680\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.724\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.591\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.471\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.598\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.544\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eRUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.612\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.604\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.619\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.637\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.579\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.540\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.582\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.593\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eLR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eSMOTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.718\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.667\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.765\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.787\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.742\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.310\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.768\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.554\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eSVM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eSMOTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.720\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.661\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.776\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.718\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.762\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.288\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.791\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.539\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eDT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eSMOTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.653\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.655\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.651\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.680\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.636\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.445\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.648\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.550\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eSMOTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.972\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.955\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.988\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.986\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.752\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.307\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.780\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.565\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eNB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eSMOTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.660\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.475\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.833\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.757\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.814\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.245\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.849\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.560\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eKNN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eSMOTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.789\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.610\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.957\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.915\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.881\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.117\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.928\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.550\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 86px;\"\u003e\n \u003cp\u003eSMOTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e0.861\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.793\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\n \u003cp\u003e0.925\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003e0.923\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\n \u003cp\u003e0.859\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\n \u003cp\u003e0.182\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.900\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.569\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 82px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 1px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 81px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 86px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eAUC, area under the curve; DT, decision tree; KNN, k-nearest neighbors; LR, logistic regression; NB, na\u0026iuml;ve bayes; RF, random forest; ROS, random over-sampling; RUS, random under-sampling; SMOTE, synthetic minority over-sampling technique; SVM, support vector machine; XGBoost, extreme gradient boosting.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImportant Predictors for Injury\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAccording to the\u0026nbsp;XGBoost\u0026nbsp;model with RUS, the most important factors influencing injury among children aged 0-10 in Shenzhen were children\u0026rsquo;s grade (with a score of 100.00), practice score (69.87), children\u0026rsquo;s age (56.64), administrative district (39.66), and registered permanent residence (37.94), respectively \u003cstrong\u003e(Fig.4)\u003c/strong\u003e.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eInjury in childhood contribute substantially to global mortality and morbidity rates, with children residing in low- and middle-income countries being particularly susceptible[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. According to epidemiological survey data, the incidence of child injury in Shenzhen (5.79%) was lower than that in Shanghai (9.3%) and Suzhou (7.7%)[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Compared to specific high-risk groups, Shenzhen\u0026rsquo;s rate is significantly lower than Guangzhou\u0026rsquo;s preschool-aged children (34.24%), and left-behind children in rural poverty-alleviation regions (15.6%)[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. In cross-national comparisons, child injury incidence in Shenzhen exceeded that of Nepal\u0026rsquo;s Makwanpur District (2.46%), Thailand (1.5%), and India (3.42%)[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Regional disparities arose from differences in survey sample sources, survey periods, and research scopes. Falling was the primary cause. These findings were similar to other studies[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Another significant fact was fall as the primary contributor to traumatic brain injury (TBI) in children, comprising 90% of TBI cases in children aged 0 to 4 years and 42% of those in children between 5 and 14 years old[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Child injury constitutes a critical public health issue due to its substantial disease burden and economic impact, necessitating effective prevention and control measures..\u003c/p\u003e\u003cp\u003eIn our study, the injury reporting incidence was low (imbalance). The distribution of classes in datasets was highly unequal. It is a common challenge in machine learning, particularly in classification problems. Resampling techniques and synthetic data generation were used in our analysis to handle these data[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Among them, ROS was performed by randomly duplicating samples in a minority class to match the number of samples in a majority class[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. SMOTE is easy to synthesize \u0026ldquo;noisy sample\u0026rdquo;and\u0026ldquo;boundary sample\u0026rdquo;[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Findings emphasized the importance of balancing data in ML applications.\u003c/p\u003e\u003cp\u003eOur comprehensive evaluation of the results-based on accuracy, AUC, sensitivity, specificity, indicated that the RF model with SMOTE outperformed the other ML models (AUC\u0026thinsp;=\u0026thinsp;0.986, sensitivity\u0026thinsp;=\u0026thinsp;0.955, specificity\u0026thinsp;=\u0026thinsp;0.988) on training set. RF is one of the state-of-the-art ML methods for developing predictive models. It is a non-parametric methodology capable of handling diverse response types, encompassing categorical, quantitative outcomes, as well as survival times[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. RF excels in machine learning due to its ensemble learning framework, which aggregates multiple decision trees through bagging and random feature selection, effectively reducing overfitting while maintaining high predictive accuracy[\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. It demonstrates robustness to high-dimensional data and noise by automatically evaluating feature importance, enabling efficient handling of complex datasets without requiring prior feature scaling[\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. For instance, in a study comparing supervised ML algorithms, a total of 48 articles were included for disease prediction, among which 15 applied the RF, and 9 of these achieved the highest accuracy[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. On the test set, the XGBoost model combined with RUS (AUC\u0026thinsp;=\u0026thinsp;0.593, sensitivity\u0026thinsp;=\u0026thinsp;0.540, specificity\u0026thinsp;=\u0026thinsp;0.582) and the LR model combined with ROS (AUC\u0026thinsp;=\u0026thinsp;0.586, sensitivity\u0026thinsp;=\u0026thinsp;0.511, specificity\u0026thinsp;=\u0026thinsp;0.600) were superior to other ML models. The XGBoost model also demonstrated excellent performance in the risk prediction of coronary heart disease[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. The discrepancy between the training set and test set results may arise from data balancing applied to the training set while the test set retains its original imbalanced distribution. Additionally, the suboptimal predictive performance could be attributed to the limited number of variables included in the model, particularly the absence of critical factors such as household environment characteristics and external environment variables.\u003c/p\u003e\u003cp\u003eThe analysis of predictor importance revealed that the children\u0026rsquo;s grade was the most significant factor influencing child injury, And the other important predictors included practice score, children\u0026rsquo;s age, administrative district, and registered permanent residence. These findings emphasized the need for comprehensive injury prevention programs that target not only children but also their caregivers, focusing on increasing the time spent together, improving communication, and addressing the unique challenges faced by families across diverse socioeconomic and demographic backgrounds. Modern injury prevention and control aims to prevent, limit, or control injuries through the 4 Es of injury prevention: engineering, enforcement, education, and economic[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe results of this study had important implications for the development of effective injury prevention strategies in Shenzhen. First, targeted interventions aimed at improving working status, the educational level and safety knowledge of primary caregivers, particularly in low-income and under-resourced communities, could help reduce the incidence of child injury. Second, promoting regular and meaningful communication between parents and children, especially regarding safety issues, could also contribute to injury prevention. Third, the appropriate use of ML models facilitated the identification of high-risk groups and informed the allocation of resources and interventions.\u003c/p\u003e\u003cp\u003eThis study had several limitations. First, The generalization of findings was limited due to the study being conducted in a developed megacity. Multi-center research in the future should be considered to validate the results in diverse populations. Second, Fewer variables could lead to lower predictive performance in our study. More variables should be designed and included.. Finally, while the ML models demonstrated promising predictive performance, further validation in an independent dataset is necessary to confirm their robustness and generalizability.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eAlthough the injury reporting incidence in Shenzhen was relatively low, it remains a challenge in the field of public health. By identifying key predictors and developing effective ML models, the factors influencing child injury became clearer, enabling the development of targeted prevention strategies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eAI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eArtificial intelligence\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eAUC,\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003earea under the curve\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eDT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eDecision tree\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eKNN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eK-nearest neighbors\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eLR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eLogistic regression\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eML\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eMachine learning\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eNB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eNa\u0026iuml;ve bayes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eRandom forest\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eROC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eReceiver operating characteristic curve\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eRandom over-sampling\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eRUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eRandom under-sampling\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eSMOTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eSynthetic minority over-sampling technique\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eSVM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eSupport vector machine\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 195px;\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 374px;\"\u003e\n \u003cp\u003eExtreme gradient boosting\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEthical approval was granted by the Ethics Committee of National Center for Chronic and Noncommunicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention (No. 201713). The study adhered to the Declaration of Helsinki, with written informed consent obtained from all participants and/or their legal guardians.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analyzed during the current study are available from the corresponding authors on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by United Nations International Children\u0026rsquo;s Emergency Fund (Grant No. 2022XQLH115) and the District-Level Scientific Research Project of Shenzhen Longhua District Medical Institutions (Grant No. 2023048).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eXF and CN designed the study. SY, XY and NP analyzed the data and wrote the manuscript. YQ, CN, JR and MX participated in on-site investigations and collected and organized data. SJ and JD helped perform the analysis with constructive discussions. LL revised the manuscript and edited the English language. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGratitude is expressed to the United Nations International Children\u0026rsquo;s Emergency Fund (UNICEF), and National Center for Chronic and Noncommunicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention for financial support, and to all child and their guardians with their involvement.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor details\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e1\u003c/sup\u003eLonghua Center for Chronic Disease Control, Shenzhen, Guangdong 518110, China\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e2\u003c/sup\u003eBaoan Chronic Diseases Prevent and Cure Hospital, Shenzhen, Guangdong 518102, China\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e3\u003c/sup\u003eLonggang Center for Chronic Disease Control, Shenzhen，Guangdong 518100，China\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e4\u003c/sup\u003eShenzhen Center for Chronic Disease Control, Shenzhen 518020, China\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe questionnaire used to collect data in this study was compiled by the National Center for Chronic and Noncommunicable Disease Control and Prevention. Sections of this questionnaire were utilized in the research by Xie et al.[43], and the full questionnaire is detailed in \u003cstrong\u003eSupplementary Material 1\u003c/strong\u003e.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eDuan LL, Ye PP, Haagsma JA, Jin Y, Wang Y, Er YL, Deng X, Gao X, Ji CR, Linhong Wang LH, et al. The burden of injury in China, 1990-2017: findings from the Global Burden of Disease Study 2017. Lancet Public Health. 2019; 4:e449-61.\u003c/li\u003e\n\u003cli\u003eJames SL, Castle CD, Dingels ZV, Fox JT, Hamilton EB, Liu Z, NL SR, Sylte DO, Henry NJ, LeGrand KE et al. Global injury morbidity and mortality from 1990 to 2017: results from the Global Burden of Disease Study 2017. Inj Prev. 2020; 26:i96-114.\u003c/li\u003e\n\u003cli\u003eWorld Health Organization. Injuries and violence: the facts 2014. Geneva: World Health Organization. 2014.\u003c/li\u003e\n\u003cli\u003eTang EHM, Bedford LE, Yu EYT, Tse ETY, Dong W, Wu T, Cheung BMY, Wong CKH, Lam CLK. Unintentional Injury Burden in Hong Kong: Results from a Representative Population-Based Survey. Int J Environ Res Public Health. 2021; 18.\u003c/li\u003e\n\u003cli\u003ePeden M, Oyegbite K, Ozanne-Smith J, Hyder AA, Branche C, Rahman A, Rivara F, Bartolomeos K. WHO Guidelines Approved by the Guidelines Review Committee. World Report on Child Injury Prevention.2008.\u003c/li\u003e\n\u003cli\u003eJullien S. Prevention of unintentional injuries in children under five years. BMC Pediatr. 2021; 21:311.\u003c/li\u003e\n\u003cli\u003eHuang S, Cai N, Pacheco PP, Narrandes S, Wang Y, Xu W. Applications of Support Vector Machine (SVM) Learning in Cancer Genomics. Cancer Genomics Proteomics. 2018; 15:41-51.\u003c/li\u003e\n\u003cli\u003eClift AK, Dodwell D, Lord S, Petrou S, Brady M, Collins GS, Hippisley-Cox J. Development and internal-external validation of statistical and machine learning models for breast cancer prognostication: cohort study. Bmj. 2023; 381:e073800.\u003c/li\u003e\n\u003cli\u003eGoldstein BA, Navar AM, Carter RE. Moving beyond regression techniques in cardiovascular risk prediction: applying machine learning to address analytic challenges. Eur Heart J. 2017; 38:1805-14.\u003c/li\u003e\n\u003cli\u003eBeam AL, Kohane IS. Big Data and Machine Learning in Health Care. Jama. 2018; 319:1317-18.\u003c/li\u003e\n\u003cli\u003eFu P, Lu J. Research progress of artificial intelligence in ischemic stroke imaging. Chin Geriatr Heart Brain Vessel Dis. 2020;22:1110-12.\u003c/li\u003e\n\u003cli\u003eKrishnan R, Rajpurkar P, Topol EJ. Self-supervised learning in medicine and healthcare. Nat Biomed Eng. 2022; 6:1346-52.\u003c/li\u003e\n\u003cli\u003eTiruneh SA, Vu TTT, Rolnik DL, Teede HJ, Enticott J. Machine Learning Algorithms Versus Classical Regression Models in Pre-Eclampsia Prediction: A Systematic Review. Curr Hypertens Rep. 2024; 26:309-23.\u003c/li\u003e\n\u003cli\u003eAljarallah NA, Dutta AK, Sait ARW. A Systematic Review of Genetics- and Molecular-Pathway-Based Machine Learning Models for Neurological Disorder Diagnosis. Int J Mol Sci. 2024; 25.\u003c/li\u003e\n\u003cli\u003eChen X, He L, Shi K, Wu Y, Lin S, Fang Y. Interpretable Machine Learning for Fall Prediction Among Older Adults in China. Am J Prev Med. 2023; 65:579-86.\u003c/li\u003e\n\u003cli\u003eGentili E, Franchini G, Zese R, Alberti M, Ferrara M, Domenicano I, Grassi L. Machine learning from real data: A mental health registry case study. CMPB Update. 2024; 5:100132.\u003c/li\u003e\n\u003cli\u003eJadhav A, M. Mostafa SM, Elmannai H, Karim FK. An Empirical Assessment of Performance of Data Balancing Techniques in Classification Task. Appl Sci. 2022; 12:3928.\u003c/li\u003e\n\u003cli\u003eHasanah U, Soleh AM, Sadik K. Effect of Random Under sampling, Oversampling, and SMOTE on the Performance of Cardiovascular Disease Prediction Models. JMSK. 2024; 21:88-102.\u003c/li\u003e\n\u003cli\u003eKhan AA, Chaudhari O, Chandra R. A review of ensemble learning and data augmentation models for class imbalanced problems: Combination, implementation and evaluation. Expert Syst Appl. 2024; 244:122778.\u003c/li\u003e\n\u003cli\u003eZhou H, Tong J, Liu Y, Zheng K, Cao C: An oversampling FCM-KSMOTE algorithm for imbalanced data classification. J King Saud Univ-Com. 2024; 36:102248.\u003c/li\u003e\n\u003cli\u003eZhang L, Wang RY, Yang H, Zhu SL. Application of resampling technique in the classification of imbalanced diabetes data in middle-aged and elderly residents. Mod Prev Med. 2023; 50:1339-44.\u003c/li\u003e\n\u003cli\u003eGeng X. 2016-2020 Childhood Injury Prevention Program launched (in Chinese). Disease Surveillance. 2017; 32:485.\u003c/li\u003e\n\u003cli\u003eDuan LL,Yang Y, Zhang R, Wu F. Analysis of childhood unintentional injury situation in the three municipalities in China. Chin J Health Educ. 2006; 23:248-50.\u003c/li\u003e\n\u003cli\u003eDuan LL, Wang LH. Theory and methods for injury and violence prevention and control. People\u0026apos;s Medical Publishing House. 2020.\u003c/li\u003e\n\u003cli\u003eTupetz A, Friedman K, Zhao D, Liao H, Isenburg MV, Keating EM, Vissoci JRN, Staton CA. Prevention of childhood unintentional injuries in low- and middle-income countries: A systematic review. PLoS One. 2020; 15:e0243464.\u003c/li\u003e\n\u003cli\u003ePeng JJ, Gao N, Yu Y, Zhou DD, Su HJ, Xu NT, Shi Y, Zhong WJ. Epidemiological survey on non-fatal injuries among students in Shanghai.Shanghai J Prev Med. 2018; 30:723-9.\u003c/li\u003e\n\u003cli\u003eDai NB, Wang J, Gong T, Huang QL. Analysis of injury epidemiological characteristics in children aged 0-14 years in Suzhou. Chin J Dis Control Prev. 2019; 23:299-303.\u003c/li\u003e\n\u003cli\u003eWei YH, Wang DH, Jing QL, Lu JY, Hu HH, Chang HM, Liu WJ. Epidemiological analysis of primary students unintentional injury in Guangzhou area. J Med Pest Control. 2014; 30:711-13.\u003c/li\u003e\n\u003cli\u003eHe JZ, Chen YL, Liu YF, Du WJ. Analysis on the influencing factors of injuries among left-behind children in the first grade of junior middle. Chin J Health Educ. 2022; 38:438-46.\u003c/li\u003e\n\u003cli\u003ePant PR, Towner E, Ellis M, Manandhar D, Pilkington P, Mytton J. Epidemiology of Unintentional Child Injuries in the Makwanpur District of Nepal: A Household Survey. Int J Environ Res Public Health. 2015; 12:15118-28.\u003c/li\u003e\n\u003cli\u003ePant PR, Towner E, Pilkington P, Ellis M. Epidemiology of unintentional child injuries in the South-East Asia Region: a systematic review. Int J Inj Contr Saf Promot. 2015; 22:24-32.\u003c/li\u003e\n\u003cli\u003eBao W, Li H, Li N, Sun JX, Zhang XL, An Hl. Analysis of children injuries in Hebei Province from 2011 to 2015. Chin J Sch Health. 2018; 39:709-12+15.\u003c/li\u003e\n\u003cli\u003eCenters for Disease Control and Prevention. Injury prevention and control: WISQARS injury statistics. 2019.\u003c/li\u003e\n\u003cli\u003eTaylor CA, Bell JM, Breiding MJ, Xu L. Traumatic Brain Injury-Related Emergency Department Visits, Hospitalizations, and Deaths - United States, 2007 and 2013. MMWR Surveill Summ. 2017; 66:1-16.\u003c/li\u003e\n\u003cli\u003eHe H, Garcia EA. Learning from Imbalanced Data. IEEE Trans. 2009; 21:1263-84.\u003c/li\u003e\n\u003cli\u003eChawla NV, Bowyer KW, Hall LO, et al. SMOTE: synthetic minority oversampling technique. J Artif Intell Res. 2002; 16:321-57.\u003c/li\u003e\n\u003cli\u003eHan H, Wang WY, Mao BH. Borderline-SMOTE: a new over-sampling method in imbalanced datasets learning.International conference on intelligent computing, Springer, Berlin, Heidelberg. 2005.\u003c/li\u003e\n\u003cli\u003eHu J, Szymczak S. A review on longitudinal data analysis with random forest. Brief Bioinform. 2023; 24.\u003c/li\u003e\n\u003cli\u003eBreiman L, Cutler A. Random Forests. Berkeley: University of California. 2004.\u003c/li\u003e\n\u003cli\u003eUddin S, Khan A, Hossain ME, Moni MA. Comparing different supervised machine learning algorithms for disease prediction. BMC Med Inform Decis Mak. 2019; 19:281.\u003c/li\u003e\n\u003cli\u003eYue HT, He CC, Cheng YY, Zhang SC, Wu Y, Ma J. Coronary heart disease risk prediction model based on machine learning. Chin Gen Pract. 2025; 28:499-509.\u003c/li\u003e\n\u003cli\u003eMace SE, Gerardi MJ, Dietrich AM, Knazik SR, Mulligan-Smith D, Sweeney RL, Warden CR. Injury prevention and control in children. Ann Emerg Med. 2001; 38:405-14.\u003c/li\u003e\n\u003cli\u003eXie Y, Yu XY, Wu XY, Zhang WY, Feng ZL, Xiao F, Deng X, Dai WJ, Zhao SJ. Association between the guardians\u0026apos;educational levels and unintentional injuries in children aged 6-18 in Shenzhen, China. BMC Public Health. 2024; 24:2344.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Child injury, Machine learning, Prediction, Risk factors, Prevention","lastPublishedDoi":"10.21203/rs.3.rs-7260288/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7260288/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eUnintentional injury is a leading cause of death and years of healthy life lost due to disability among children. This study aimed to develop machine learning (ML) models to predict the occurrence of unintentional injury in children aged 0\u0026ndash;10 years in Shenzhen, China, and to identify the associated influencing factors.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThis cross-sectional study recruited 15,796 children aged 0\u0026ndash;10 years in Shenzhen during 2017\u0026ndash;2018. Information on the children and guardians were collected, and unintentional injuries in the past year was examined by using two nested questions. The dataset was randomly partitioned into training and test sets at a 7:3 ratio. Class imbalance in the training set was addressed using data balancing techniques. Subsequently, seven machine learning algorithms were employed to establish a pediatric injury risk prediction model. The top five predictors for injury were ultimately identified based on the optimal performing model.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eAmong 15,796 children surveyed, 915 (5.79%) experienced at least one unintentional injury within the past year. Falls constituted a leading cause of injuries, with 53.06% of such incidents occurring at home. On the training set, The Random Forest (RF) model integrated with SMOTE demonstrated optimal performance, achieving an AUC of 0.986, sensitivity of 0.955, and specificity of 0.988. On the test set, the prediction effect of the Extreme Gradient Boosting model combined with Random Under-Sampling (RUS) outperformed other algorithms, yielding an AUC of 0.593, sensitivity of 0.540, and specificity of 0.582. The children\u0026rsquo;s grade was the most important predictor of child injury.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eOur research indicated that ML models, when coupled with data balancing techniques, prove to be potent instruments for predicting child injury in scenarios involving imbalanced datasets.\u003c/p\u003e","manuscriptTitle":"Unintentional Injury Predicton Among Children aged 0-10 years old in Shenzhen, China: Based on Machine Learning Models","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-08 15:56:51","doi":"10.21203/rs.3.rs-7260288/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"238595923744162052822029717787672850012","date":"2025-09-26T15:33:41+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-25T09:39:36+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-14T12:52:44+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-08-11T12:11:54+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-11T08:34:49+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Public Health","date":"2025-08-11T08:31:24+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"16a006a6-4ce1-4b49-a531-7cc457fb6b0e","owner":[],"postedDate":"October 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-10-08T15:56:51+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-08 15:56:51","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7260288","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7260288","identity":"rs-7260288","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}