基于潜在剖面分析的药物滥用患者报告结局异质性研究  

Heterogeneity of reported outcomes in drug abuse patients based on latent profile analysis

作　　者：姚秀坤 张娇 任飞[5] 胡晓娟 张岩波 尉志文[1,2] 曹虹艳 贠克明 YAO Xiukun;ZHANG Jiao;REN Fei;HU Xiaojuan;ZHANG Yanbo;WEI Zhiwen;CAO Hongyan;YUN Keming(School of Forensic Medicine,Shanxi Medical University,Taiyuan 030001,China;Key Laboratory of Forensic Toxicological Analysis of the Ministry of Public Security;Academy of Medical Sciences,Shanxi Medical University;Department of Health Statistics,School of Public Health,Shanxi Medical University;Institute of Forensic Science Public Security Bureau of Taiyuan(IFSTY);School of Manage-ment,Shanxi Medical University)

机构地区：[1]山西医科大学法医学院,太原030001 [2]法庭毒物分析公安部重点实验室 [3]山西医科大学医学科学院 [4]山西医科大学公共卫生学院卫生统计学教研室 [5]山西省太原市公安局司法鉴定中心 [6]山西医科大学管理学院

出　　处：《山西医科大学学报》2025年第3期320-328,共9页Journal of Shanxi Medical University

基　　金：国家自然科学基金资助项目(82130056,82101980,82072116);国家重点研发计划项目(2022YFC3300903,2022YFC3302003);山西省基础研究计划项目(202103021224233)。

摘　　要：目的通过潜在剖面分析方法对药物滥用患者进行分型,探索不同类型患者在各领域的表现差异,为制定个体化干预策略提供依据。方法选择全国3所强制戒毒中心和6个社区戒毒及康复中心的药物依赖患者为调查对象,使用基础资料调查问卷和药物滥用患者报告结局量表(DA-PROM)收集数据。通过潜在剖面分析将患者分为不同类别(Class),根据其在量表各领域的得分表现命名,并采用方差分析和卡方检验评估可能影响潜在类别的因素。结果共纳入研究对象611名,根据潜在剖面分析,药物滥用患者可分为3种潜在类别:类别1(Class1)、类别2(Class2)、类别1(Class3)。根据其在各领域的得分表现依次命名为:综合表现良好型(18.49%,Class1)、低功能型(53.36%,Class2)和高恢复型(28.15%,Class3)。在生理、心理和康复领域,各类别之间的得分差异均有统计学意义(P<0.05),高恢复型、综合表现良好型、低功能型得分依次降低;然而,在社会领域,高恢复型与综合表现良好型之间的得分差异无统计学意义(P>0.05)。患者分型与性别、收入、婚姻状况、药物滥用类型、居住条件、居家治疗经历及既往治疗环境相关(P<0.05)。具体而言,低功能型患者中女性、月收入<5000元、未婚/离异/丧偶、使用传统毒品、无稳定住所、无居家治疗经历以及强制治疗人群占比较高;高恢复型患者中月收入>5000元、混合使用毒品人群占比较高。结论通过潜在剖面分析方法将药物滥用患者分为3种类别,且不同类别在生理、心理、康复领域表现存在显著差异,患者分型与性别、收入等多种因素相关。针对女性、低收入、婚姻不稳定、缺乏稳定住所、吸食传统毒品以及有居家治疗或强制治疗经历的低功能型患者,应依据其特点制定个体化干预策略,以促进其康复,实现对不同类型药物滥用患者的精准管理。Objective To classify drug abuse patients into subtypes and explore their domain-specific variations using latent profile analysis(LPA)for providing a basis in design of individualized intervention strategies.Methods Drug abuse patients from three compulsory drug treatment centers and six community-based rehabilitation centers were enrolled.Data were collected through a basic information questionnaire and the drug abuse patient-reported outcome measures(DA-PROM).LPA was employed to categorize patients into different latent classes based on standardized domain scores.ANOVA and Chi-square tests were used to analyze factors influencing classification.Results A total of 611 participants were included in the study.The LPA results showed that drug abuse patients were divided into three latent classes:Class 1,Class 2,and Class 3,named as good comprehensive performance type(18.49%,Class1),low function type(53.36%,Class2),and high recovery type(28.15%,Class3).Significant differences were observed in physiological domain,psychological domain,and rehabilitation domain(P<0.05),and the scores in patients with high recovery type,good comprehensive performance type,and low function type decreased gradually.No significant difference was found in the social domain between high recovery type and good comprehensive performance type(P>0.05).Patient classification was correlated with gender,income,marital status,type of drug abuse,living conditions,home treatment experience,and previous treatment environment(P<0.05).Specifically,the low-function type patients(Class 2)exhibited higher proportions of female,low-income individuals(monthly income<5000 RMB),unmarried/divorced/widowed individuals,traditional drug users,and those without stable housing or homebased treatment experience.The high-recovery patients(Class 3)showed higher proportions of high-income individuals(monthly income≥5000 RMB)and polydrug users.Conclusion Drug abuse patients are classified into three categories by LPA,and the performance of different categories in the

关 键 词：潜在剖面分析 药物滥用 患者报告结局 异质性 潜在类别 个体化干预 精准医疗 

分 类 号：R749.6[医药卫生—神经病学与精神病学]