A Predictive Model of Non-suicidal Self-Injury -a Study Based on the Construction and Validation of a Nomogram

YuJie Liu ^1,2 武 泰民 ² Yan Shu ² Lianzhong Liu ^3* Yang Zhou ^2*

• ¹ School of Medicine, Jianghan University, Wuhan, Hebei Province, China
• ² Wuhan Mental Health Center, Wuhan, Hubei Province, China
• ³ Wudong Hospital, Wuhan, China

The issue of psychological maladjustment, particularly Non-Suicidal Self-Injury (NSSI), is prevalent among vocational high school students. Therefore, timely identification of high-risk individuals is important in providing further intervention. Methods: A survey was conducted among 2081 students from a vocational high school in Wuhan, China. The students were divided into two groups: those who had engaged in Non-Suicidal Self-Injury (NSSI) within the past two weeks and those who had not. Lasso regression and logistic regression were employed to identify significant risk factors associated with NSSI. Subsequently, a nomogram was developed to enhance the accuracy and efficiency of identifying individuals at high risk for NSSI. The performance of the model was assessed through various validation methods including Area Under the Curve (AUC), calibration curves, and Decision Curve Analysis (DCA). Results: The significant predictors of NSSI encompassed gender, problem behavior, depressive mood, and borderline personality tendencies. Based on these predictors, a nomogram was constructed. The model's accuracy was validated using AUC, calibration curves, and DCA, showing high accuracy. Conclusion: A nomogram prediction tool for NSSI among vocational high school students was constructed, providing an accurate and quick method for predicting adolescent NSSI behavior.