心理解剖在农村青年自杀身亡者中的应用研究

发布时间：2018-05-30 07:23

本文选题：心理解剖 + 自杀　；参考：《山东大学》2009年博士论文

【摘要】：研究背景自杀,尤其是农村青年的自杀,是我国目前面临的重大公共卫生问题和社会问题。有效的自杀预防和干预策略的制订,需要深入的自杀身亡研究作为其可靠的科学依据。而自杀死亡的研究,却面临着“如何有效地收集死者的全面信息”这一方法学问题。心理解剖的发展与应用,则为该问题的解决提供了一个有效的途径。心理解剖通过访谈“了解死者信息的其他人”等方式,可有效地重建与自杀死亡者有关的心理、社会因素,更好地了解其自杀原因,是目前最具有科学性和可操作性的自杀死亡研究方法。心理解剖在我国自杀研究中的应用仍处于起步阶段,而且目前尚无专门针对我国农村青年自杀死亡的心理解剖研究。而有效的自杀干预,则需要根据不同人群的特征量身定制相应的策略。为了更好地开展针对我国农村青年的自杀预防与干预工作,运用心理解剖方法对农村青年的自杀死亡特征进行深入探讨是非常必要的。心理解剖的信息来源主要以信息人为主,而单一的信息来源往往不能够全面地反映目标人的情况,所以在人力物力及其它条件许可的情况下,可争取选择多位信息人作为心理解剖的信息来源。而当心理解剖存在多位信息人,且不同信息来源所提供的信息不一致时,如何进行信息的合成,到目前为止国际上尚无统一的操作规范,也没有确凿的证据来证明哪一种信息合成技术相对较佳。所以,如何运用多位信息人所提供的代理信息进行死者或其它目标人的信息重构,而且该技术所提供的代理信息能否有效地代表目标人的实际信息,己成为心理解剖研究迫切需要解决的问题之一。心理解剖的应用,不应该仅仅停留在理论上的探讨,还应该将之运用于自杀研究的实践中,使之更好地服务于自杀的预防和干预工作。以往的心理解剖研究,主要侧重自杀者的“单一特征”的识别,而非“特征群”的识别。特征群的分析,通常采用分类回归树进行探讨,它在自杀危险因素的探讨及其它自杀研究中上已体现出其独特的优越性。而且众多研究一致表明：自杀不是单一危险因素作用的结果,自杀者通常同时具有多种危险因素和保护因素。因此,运用心理解剖方法进行自杀者“特征群”的探讨比“单一特征”的识别,对于自杀的预防和干预将更具指导意义和可操作性。针对上述多个问题,本研究为每例目标人选择了两位信息人进行心理解剖研究,并在国际上首次全方位地对不同信息重构技术进行效度探讨,然后应用已被证明相对较优的心理解剖信息重构技术,对我国农村青年自杀身亡者的特征群进行分类树分析,以期为我国自杀预防与干预工作提供科学依据。研究目的 1.探讨不同信息重构技术的效度,为心理解剖的信息收集和目标人的信息构建提供科学依据； 2.探索我国农村青年自杀死亡的高危特征群,为自杀预防和干预策略的制订提供有效参考。研究方法在山东、辽宁和湖南三个省随机抽取16个县或县级市,然后在上述16个县市通过序贯抽样的方法抽取2005年10月1日至2008年6月30日期间自杀死亡的15-34岁农村居民作为研究对象,同时随机抽取等数量同年龄段的农村居民作为自杀身亡者的对照纳入本次研究。自杀死亡者名单来自各县级疾病预防控制中心死因监测系统。此次调查共获得392例自杀死亡者和416例对照的有效资料。采用对照组的资料,以对照本人所提供的信息作为金标准,进行心理解剖信息重构技术的效度评价。采用配对t检验并结合组内相关系数,对1st、2nd、A、B、C、D、E类共七种不同信息重构技术在绝望感、社会支持、冲动性、特质焦虑和应对技能等数值型资料上的效度进行评价；采用McNemar检验并结合灵敏度、特异度、约登指数、Kappa值等指标,对1st、2nd、D、E、G、H类共六种不同信息重构技术在二分类变量如生活事件的发生情况上的效度进行评价。将“各信息人在绝望、社会支持、冲动性、特质焦虑和应对技能量表的分量表上对目标人的量表评分”与“相应分量表目标人自我评估得分”的“差值的绝对值”的总分作为因变量,信息人的“与目标人的关系”、“对目标人的了解程度”、性别、年龄、婚姻状况、职业、受教育程度、宗教信仰、家庭年收入、健康状况和CES-D抑郁评分作为自变量,以“比例偏差”和"Pearson卡方”作为模型筛选和评估参数,进行广义线性模型的拟合,筛选对代理信息效度有影响的信息人特征变量。分别采用相应的最优或相对较优的心理解剖信息重构技术,对392例自杀死亡者和416例对照的不同方面的信息进行构建。然后在校正信息人特征对代理效度的影响的基础上,采用病例对照研究的方法,进行我国农村青年自杀身亡影响因素的探讨,并应用分类树分析对他们的高危特征群进行探索。主要结果 1.心理解剖不同信息重构技术效度评价 lst、2nd、A、B、C、D、E、G、H类这九种信息重构技术所构建的信息的组内相关系数或Kappa值均小于0.6,代理信息与对照本人所提供的信息的一致性均未达到良好的程度,即代理信息的精确度均不高。 lst、2nd、A、D和E类这五种信息重构技术所构建的信息在绝望量表总量表及其三个分量表、社会支持量表的社会交往及工具性支持两个分量表、冲动性量表的功能性冲动分量表、特质焦虑量表、应对技能量表的“积极行动应对维度及其四个分量表”和“消极回避应对维度及其两个分量表(认知回避、寻求代偿)”上与对照本人信息均无统计学差异。而它们在社会支持量表的可觉察社会支持分量表上的信息均欠佳,均高于对照本人信息(P0.05)；在冲动性量表的非功能性冲动分量表、消极回避应对维度的发泄情绪分量表上则均低于对照本人信息(P0.05)。上述五种信息重构技术在冲动性总量表、应对技能量表的消极回避应对维度的“接受或放弃”分量表与对照本人信息的差异性则不尽相同：(1)在冲动性总量表上,1 st、2nd与D类这三类代理信息与对照本人无差异,而A和E类的代理信息则均小于对照本人信息(P0.05)；(2)在“接受或放弃”分量表上,1 st与D类信息重构技术所提供的信息与对照本人无显著性差异,2nd、A和E类这三种代理信息则均小于对照本人信息(P0.05)。另外,B类与C类信息重构技术所构建的信息在上述量表与对照本人信息均有统计学差异,前者的信息均比对照本人信息大,而后者则均比对照小。总体而言,在上述各量表上,1st与D类信息重构技术的效度比其它五类(2nd、A、B、C和E类)信息重构技术略好。在“婚姻相关问题”、“工作和学习相关问题”、“健康相关问题”和“法律相关问题及其它”这四大类生活事件上,“取阳性值”的G类信息重构技术所提供的信息与对照本人的信息无统计学差异,P值依次为0.810、0.363、0.534、0.477；而在“家庭生活相关问题”这一类生活事件上,1 st类代理信息与对照本人无显著性差异(P=0.828)。其他四种信息重构技术(2nd, D, E, H类)在上述生活事件上所构建的信息与本人信息差异较大。另外,这六种信息重构技术所提供的信息在生活事件量表上的灵敏度、特异度及约登指数均有显著性差别；其中,“取阳性’值”的G类信息重构技术所提供的信息的灵敏度及约登指数均最高,但特异度最差。2.我国农村青年自杀死亡的高危特征群探索在我国农村青年中,受教育程度越高、躯体健康状况越好、寻求指引与支持的能力越好、过去一年中发生的与婚姻相关的负性生活事件越少,自杀风险越低；愿望未实现、患有精神疾病是自杀的危险因素,其OR值分别为4.19、9.95；绝望量表的“对未来的感觉”和“对未来的期望”分量表得分越高,自杀风险越高；消极回避应对技能的“发泄情绪”分量表得分越高、“自杀家族史”阳性,均呈现增加自杀风险的趋势,但无统计学意义(OR=1.09、4.50,P=0.183、0.157)。社会支持、冲动性、特质焦虑均不是我国农村青年自杀身亡的重要影响因素。我国农村青年自杀的高危特征群主要为：(1)绝望感分量表“对未来的感觉”得分大于17.5,而且“对未来的期望”得分大于12.5；(2)绝望感分量表“对未来的感觉”得分小于或等于17.5,但患有精神疾病；(3)绝望感分量表“对未来的感觉”得分小于或等于17.5,并未患有精神疾病、“寻求指引和支持”的能力相对尚可,但受教育年限较低、“发泄情绪”量表得分大于5.5；(4)绝望感分量表“对未来的感觉”得分小于或等于17.5,不患有精神疾病,但“寻求指引和支持”的能力差；(5)绝望感分量表“对未来的期望”得分小于或等于12.5,但“对未来的感觉”得分大于17.5。训练样本的结果表明,该自杀身亡分类树的灵敏度为84.34%,特异度为92.08%。检验样本对上述分类树进行再验证,结果表明：该自杀身亡分类树的灵敏度为76.47%,特异度为85.92%。训练样本与检验样本的灵敏度和特异度一致表明：该自杀身亡分类树的判别效果较好。结论与意义尽管心理解剖代理信息的精确度不高,但是合适的信息重构技术仍可有效地估计目标人的信息。在心理解剖研究中,以亲属为主的第一信息人足以有效地构建目标人的绝大部分信息。而以朋友、邻居为主的第二信息人的增加,对于提高绝大部分代理信息的效度未见显著功效。对于生活事件信息,采访两位信息人可达到更好的效度,但必须借助于合适的信息重构技术。增加第二信息人却采取不合适的信息合成方式,如对数值型资料取极大值或极小值,将会导致代理信息严重偏离目标人信息。我国农村青年存在多种不同的自杀高危特征群,它们并不只包含自杀高危因素,可能还会伴有其他自杀保护因素,或者只是多个自杀低危因素的集合体。不能因为某些个体伴有自杀保护因素、或不具备一些常见自杀高危特征,而忽视其潜在的自杀高风险。有效的预防自杀,有赖于自杀高危特征群的识别、相应筛查工具的研制、个体自杀风险的综合评估。本研究在国际上首次对心理解剖不同信息重构技术的效度进行全方位的探讨,为心理解剖信息收集和数据合成提供了科学依据,也为今后心理解剖信息构建的标准化奠定了部分基础；本研究还率先对我国农村青年自杀死亡的高危特征群进行有益探索,为我国自杀干预策略的制订提供了有效的参考。
[Abstract]:Research background
Suicide, especially the suicide of the young people in the countryside, is a major public health and social problem in our country. Effective suicide prevention and intervention strategies are made, and the study of suicide is a reliable scientific basis. The study of suicide death is faced with "how to effectively collect the comprehensive information of the dead". The development and application of psychological anatomy provides an effective way to solve this problem. Psychological anatomy can effectively reconstruct the psychological, social factors and better understand the cause of suicide by interviewing "the others of the deceased" and so on. It is the most scientific at present. Sexual and operable methods of suicide research.
The application of psychosocial anatomy in China's suicide study is still in its infancy, and there is no psychosocial research on suicide deaths in rural youth in China. Effective suicide intervention needs to be tailored according to the characteristics of different populations. In order to better carry out suicide prevention against young people in rural areas of China It is very necessary to conduct in-depth discussions on the characteristics of suicide among rural youth by means of psychological intervention and intervention.
The information source of the psychological anatomy is mainly based on the information man, and the single information source often cannot reflect the situation of the target person in an all-round way. So, in the case of the human resources and other conditions, the multiple information people can be selected as the information source of the psychological dissection. When the information provided by the source is inconsistent, the information is synthesized, and so far there is no unified operating specification in the world, and there is no conclusive evidence to prove that the information synthesis technology is relatively good. So, how to use the proxy information provided by the multi information person to reconstruct the information of the deceased or other target people, And whether the proxy information provided by the technology can effectively represent the actual information of the target person has become one of the urgent problems to be solved in the study of psychological anatomy.
The application of psychological anatomy should not only stay in theory, but should also be used in the practice of suicide research to better serve the prevention and intervention of suicide. Previous psychological anatomy studies mainly focus on the "single characteristics" of suicides, rather than the identification of "characteristic groups", and the analysis of characteristic groups, The classification regression tree is usually used to explore the unique advantages of the suicidal risk factors and other suicide studies. Moreover, many studies have shown that suicide is not the result of a single risk factor, and the suicides usually have many risk factors and protective factors at the same time. Therefore, the use of psychological anatomy is used. It is more instructive and feasible to identify the "feature group" of suicide than "single feature" in the prevention and intervention of suicide.
In view of the above problems, this study has chosen two information persons for each person to carry out a psychological anatomy study, and the validity of different information reconstruction techniques in the world for the first time, and then the technology of psychological anatomy information reconstruction which has been proved relatively superior to the characteristic groups of the suicide of the young people in China's rural areas Classification tree analysis was conducted in order to provide a scientific basis for the prevention and intervention of suicide in China.
research objective
1. to explore the validity of different information reconstruction techniques, and to provide scientific evidence for the collection of psychological anatomic information and the information construction of target persons.
2. explore the high risk characteristics of suicide deaths among rural youth in China, so as to provide an effective reference for the development of suicide prevention and intervention strategies.
research method
In three provinces of Shandong, Liaoning and Hunan, 16 counties or county-level cities were randomly selected, and then 16 counties and cities were selected by sequential sampling to study the suicide deaths of 15-34 year old rural residents during the period from October 1, 2005 to June 30, 2008 as the research object, and the rural residents of the same age group were randomly selected as suicide victims. The list of suicidal deaths came from the CDC death surveillance system at the county level. The survey obtained 392 suicidal deaths and 416 control data.
The data of the control group were used as the gold standard to evaluate the validity of the psychological dissection information reconstruction technology. Seven different information reconstruction techniques, such as 1st, 2nd, A, B, C, D, E, were measured by paired t test and combined with the correlation coefficient in the group, and the number of different information reconfiguration techniques were in the sense of expectation, social support, impulsiveness, trait anxiety and coping skills. The validity of the value type data was evaluated, and the validity of six different information reconstruction techniques, such as 1st, 2nd, D, E, G, H, was evaluated by McNemar test and the sensitivity, specificity, the Kappa index, and the value of the value of the class, such as 1st, 2nd, D, E, G, H.
A total score of "the absolute value of the difference between the target person's scale score" and "the target person self evaluation score of the corresponding subscale target person" is taken as the dependent variable, the "relation of the target person", "the understanding of the target person", "the relationship between the target person" and the target person's understanding of "the information person's subscale on the scale of despair, social support, impulse, trait anxiety and coping skills scale" Degree ", sex, age, marital status, occupation, education, religious belief, family income, health status and CES-D depression score as independent variables, using" proportional deviation "and" Pearson card "as model screening and evaluation parameters, fitting the generalized linear model to screen the informational characteristics that have influence on the validity of agent information. Variable.
Using the corresponding optimal or relatively superior psychological anatomy information reconstruction techniques, the information of 392 cases of suicide deaths and 416 cases of the control were constructed. Then, on the basis of correcting the influence of the information human characteristics on the agent validity, the case control study was adopted to carry out the influence of the suicide of the rural youth in our country. We also explored their high risk characteristics using classification tree analysis.
Main results
1. validity evaluation of different psychological dissection information reconstruction techniques
In LST, 2nd, A, B, C, D, E, G, and H, the intra group correlation coefficient or Kappa value of the nine kinds of information reconstruction techniques are less than 0.6. The consistency of proxy information and the information provided by the control is not good, that is, the accuracy of proxy information is not high.
The five information reconstruction techniques of LST, 2nd, A, D and E are constructed with the information of the despair scale and its three subscales, the social interaction and instrumental support of the social support scale, the functional impulse subscale of the impulsive scale, the trait anxiety scale, and the "positive action response dimension and the four" of the skill scale. There was no statistical difference between the subscales and the negative avoidance response dimensions and their two subscales (cognitive avoidance, seeking compensation). The information on the perceived social support subscale of the social support scale was poor, higher than the control personal information (P0.05), and the non functional impulse of the impulse scale. Dynamic subscales, negative avoidance response dimensions were lower than the control personal information (P0.05). The above five information reconstruction techniques were in the impulsivity table, and the difference of "acceptance or abandonment" of the negative avoidance response dimension of the coping skill scale was different from the control information: (1) impulsiveness On the aggregate table, the three kinds of agent information of 1 st, 2nd and D are not different, while the proxy information of the A and E class is less than the control information (P0.05); (2) on the "acceptance or abandonment" subscale, there is no significant difference between the information provided by the 1 st and D information reconfiguration technology and the contrast of the control, 2nd, A and E, the three proxy information In addition, the information of the information reconstruction technology of class B and C has statistical difference between the above scale and the control information. The information of the former is larger than the control, and the latter is smaller than the control. In general, the validity of the 1st and D information reconstruction techniques is better than the others on the above scales. The five classes of information refactoring (2nd, A, B, C, and E) are slightly better.
In the "marriage related issues", "work and learning related issues", "health related issues" and "legal related issues and other" four major life events, the information provided by the G information reconstruction technology of "positive value" is not statistically different from that of the control itself, and the value of P is 0.810,0.363,0.534,0.477 in turn. There is no significant difference between the 1 st agent information and the control (P=0.828). The other four information reconstruction techniques (2nd, D, E, H) are different from the personal information. In addition, the information provided by these six information reconstruction techniques is in life. There are significant differences between the sensitivity, the specificity and the Jordan index on the scale. Among them, the information provided by the G information reconstruction technology of "positive 'value" is the highest, but the difference is the worst.2. in the high risk group of suicide death in the rural youth of China
In the rural youth of China, the higher education level, the better the physical health, the better the ability to seek guidance and support, the less negative life events associated with marriage in the past year, the lower the risk of suicide, the lack of aspiration, the mental illness as a risk factor for suicide, the OR value of 4.19,9.95, and the despair scale. The higher the score of "feeling for the future" and "expectation for the future", the higher the score, the higher the risk of suicide; the higher the score of the "vent emotion" subscale of the negative avoidance coping skills, the positive family history, and the tendency to increase the risk of suicide, but there was no statistical meaning (OR=1.09,4.50, P=0.183,0.157). Social support, impulse. Sex and trait anxiety are not important factors affecting the suicide of rural youth in China.
The high risk groups of suicide in China's rural youth are as follows: (1) the subscale of the sense of despair "feeling for the future" is more than 17.5, and the score of "expectation for the future" is greater than 12.5; (2) the subscale of the sense of despair "feeling for the future" is less than or equal to 17.5, but there is a mental illness; (3) the sense of despair is "the sense of the future." The score was less than or equal to 17.5, not suffering from mental illness. The ability to "seek guidance and support" was relatively available, but the number of years of education was lower and the score of the "vent emotion" scale was greater than 5.5; (4) the subscale of the sense of despair was less than or equal to 17.5 for the sense of the future, without mental illness, but "seek guidance and support". Poor ability; (5) the sense of despair subscale "expectation for the future" score is less than or equal to 12.5, but the result of "feeling for the future" greater than the 17.5. training sample shows that the sensitivity of the suicide classification tree is 84.34%, and the specificity is 92.08%. test samples to re verify the above classification tree, the result shows that the suicide score is the result. The sensitivity of the class tree is 76.47%, and the specificity of the 85.92%. training sample is consistent with the sensitivity and the specificity of the test sample, which shows that the discriminant effect of the suicide classification tree is better.
Conclusion and significance
Although the accuracy of the psychosocial proxy information is not high, the appropriate information reconstruction technology can still effectively estimate the information of the target person. In the psychosocial study, the first information of the first person is sufficient to effectively construct the vast majority of the information of the target person. The increase of the second information based on friends and neighbours is improved. The validity of most of the agent information is not significant. For the information of life events, the two people can achieve better validity, but it is necessary to use the appropriate information reconstruction technology. The increase of second information is taken.
【学位授予单位】：山东大学
【学位级别】：博士
【学位授予年份】：2009
【分类号】：B846

【参考文献】