国境口岸严重精神病攻击性行为风险评估模型的建立

Establishment of the risk assessment model for the serious mental illness aggressive behavior at frontier ports

目的筛查精神疾病患者攻击性行为的有效预测因子,建立精神疾病患者攻击性行为风险评估数学模型。方法采用整群抽样的方法,选取2015年10月—2016年9月在上海市精神卫生中心门诊或住院接受治疗的各类精神疾病患者800例。根据目标危险因子编制出入境人群精神类疾病攻击性行为风险评估问卷,对800例患者进行调查。结果单因素分析显示,年龄、性别等30项指标与攻击性行为发生相关(P<0.05);多因素分析显示,性别、收入、精神分裂、暴力史、睾酮(>10)、精神疾病阳性、阴性症状量表(PANSS)阳性得分(≥25)、6项生存质量量表(QOL-6)得分(≥20)、治疗依从性(<6)、临床疗效总评量表-疗效总评(CGI-GI)得分(≥4)是攻击性行为发生的独立危险因子。模型样本数据分析显示,当危险评分界值为28时,危险灵敏度和特异度分别为75.7%和72.4%,该公式具有一定的预测效果。结论建立的评估模型对精神疾病患者的攻击性行为风险进行评估有一定意义。

Objective To study the effective predictive factors of aggressive behavior in patients with mental illness, to build and evaluate the mathematical model of the risk assessment for aggressive behavior of these patients. Methods A total of 800 patients who receive treatment in Shanghai mental disease center during Oct. 2015 to Sept. 2016 were selected using cluster sampling method. A risk assessment questionnaire for aggressive behavior of psychiatric disorders was worked out,the patients were investigated using the questionnaire. Results Single factor analysis showed that 30 items were related with aggressive behavior including sex,age,etc.. Multiple factor analysis showed that sex,income,schizophrenia,history of aggressive behavior,testosterone （〉10）,PANSS scale positive score （ ≥ 25）, QOL-6 score （ ≥ 20）,treatment compliance （〈6）, CGI-GI score （≥4） was an independent risk factor for ag- gressive behavior. Based on the various coefficient detected during multiple factors analysis,a mathematical model was finally built. When the value of risk score boundary was 28,the sensitivity and specificity of the model was 75.7% and 72.4% respectively,which showed that the model had a good predictive effect. Conclusion The model built had predictive effect for aggressive behavior evaluation of patients with mental disease.