Analysis On Influencing Factors And Prediction Model Of Traumatic Events In Left-Behind And Non-Left-Behind Middle School Students In China

Objective:To understand the occurrence status and influencing factors of traumatic events of left-behind and non-left-behind middle school students,construct a prediction model of traumatic events,effectively identify high-risk groups of traumatic events,and put forward targeted suggestions for the prevention of traumatic events of left-behind and non-left-behind middle school students in the future.Methods:Using the method of combining typical sampling and multi-stage cluster sampling,5 provinces were randomly selected from the 7 provinces with a large number of left-behind children,and 1 county was selected from each of the 5provinces.Each county selected a junior high school,an ordinary high school and a secondary vocational school,and all the students in the selected schools were selected as research objects.Students’personal basic information,family intimate relationship and growth environment were investigated by self-reported questionnaire,and the occurrence of traumatic events was investigated by standardized traumatic event screening scale.Traumatic events were categorized by Latent class analysis,influential factors were analyzed byχ² test,t test,and unqualified Logistic regression,and the prediction model was constructed.The model was evaluated by ROC curve and Hosmer-Lemshow goodness of fit test.Results:（1）Traumatic events occurrence condition:A total of 13678 middle school students were surveyed in this study,including 7807 left-behind middle school students and 5871 non-left-behind middle school students.The incidence of traumatic events of left-behind middle school students was higher than that of non-left-behind middle school students（54.0%vs.39.3%,p<0.001）.The top three traumatic events for middle school students were witnessed violence outside the home（18.7%）,suffered domestic violence（15.7%）,saw dead people（13.4%）,and earthquake（0.8%）was the lowest.（2）Classification of traumatic events:Traumatic events were divided into low exposure and high exposure sub-types through Latent class analysis.Among the left-behind middle school students,5475（70.1%）were in the low exposure group and2332（29.9%）were in the high exposure group.Among non-left-behind middle school students,the number of low exposure group and high exposure group were 4749（80.9%）and 1122（19.1%）,respectively.The proportion of high exposure among left-behind middle school students was higher than that of non-left-behind middle school students（χ²=205.532,P<0.001）.（3）Influencing factors of traumatic events:Among the left-behind middle school students,the gender was female（OR=0.703,95%CI:0.631-0.782）,and the earliest left-behind age were 4～5 years old and 6～10 years old（OR=0.803,95%CI:0.656-0.985;OR=0.804,95%CI:0.663-0.974）,average and close relationship with father（OR=0.609,95%CI:0.499-0.745;OR=0.450,95%CI:0.364-0.555）,close relationship with caretakers（OR=0.627,95%CI:0.430-0.914）;the family economic conditions were medium（OR=0.756,95%CI:0.667-0.856）,the security near the school were general and good（OR=0.661,95%CI:0.494-0.885;OR=0.397,95%CI:0.298-0.529）were protective factors for high exposure to traumatic events.Older age（OR=1.057,95%CI:1.024-1.092）,family location in rural areas（OR=1.183,95%CI:1.061-1.319）,cumulative staying time of 2-3 years,4-5 years,6-10 years（OR=1.462,95%CI:1.228-1.740;OR=1.368,95%CI:1.151-1.625;OR=1.331,95%CI:1.111-1.595）,the family atmosphere were average,poor（OR=1.494,95%CI:1.323-1.688;OR=2.632,95%CI:2.077-3.335）,unstable caretakers（OR=1.545,95%CI:1.347-1.773）,scolding,spoiling and corporal punishment（OR=1.789,95%CI:1.530-2.091;OR=1.546,95%CI:1.287-1.858;OR=2.659,95%CI:2.042-3.462）were risk factors for high exposure to traumatic events.Among the non-left-behind middle school students,the gender was female（OR=0.790,95%CI:0.6855-0.912）,and the relationship with their father were normal and close（OR=0.556,95%CI:0.415-0.746;OR=0.403,95%CI:0.298-0.545）,moderate family economic conditions（OR=0.695,95%CI:0.582-0.830）,and good security near school（OR=0.440,95%CI:0.300-0.645）were protective factors for high exposure to traumatic events.The family was located in the rural area（OR=1.320,95%CI:1.134-1.536）,the emotional status of the parents were average,poor（OR=1.325,95%CI:1.081-1.624;OR=1.415,95%CI:1.027-1.948）,the family atmosphere were average or poor（OR=1.586,95%CI:1.296-1.942;OR=1.869,95%CI:1.237-2.822）,the parenting style was scolding,corporal punishment,neglect,other（OR=3.089,95%CI:2.538-3.761;OR=3.686,95%CI:2.685-5.058;OR=1.885,95%CI:1.123-3.164;OR=1.462,95%CI:1.123-1.902）,poor academic performance（OR=1.488,95%CI:1.145-1.935）were risk factors for high exposure to traumatic events.（4）Prediction model of high exposure to traumatic events:The prediction model of high exposure to traumatic events of left-behind and non-left-behind middle school students was effective in diagnosis(P_all<0.001),the Hosmer-Lemshow goodness of fit test(P_all>0.05),and the area under ROC curve were 0.712 and 0.734,respectively.All the models had moderate prediction value.Conclusions:The incidence of traumatic events of Chinese middle school students was high,most of them were at low exposure level,and the proportion of left-behind middle school students in high exposure was higher than that of non-left-behind middle school students.Although there were many common factors affecting the high exposure to traumatic events of left-behind and non-left-behind middle school students,there were also significant differences.Therefore,targeted intervention measures should be taken to prevent the high exposure to traumatic events of left-behind and non-left-behind middle school students.