用线性结构方程模型对居民两周患病与就诊影响因素的分析
|
摘要
目的
研究人群卫生服务需求和利用及其影响因素是卫生服务研究的重要课题，近些年来，我国居民卫生服务的需求和利用发生了十分显著的变化。两周患病与就诊是反映居民医疗服务需求和利用的重要指标，影响两周患病与就诊的因素多且复杂，国内外对此进行了一定的研究，许多因素在已有的研究中结果并不一致。
线性结构方程模型是一种线性统计建模技术，能将显在变量作为隐变量的标识，并允许这两种变量存在测量误差。它不但能研究标志变量，又可研究不能直接观测的隐变量，既可研究变量间的直接作用，又可研究变量间的间接作用。本研究拟采用线性结构方程模型对居民两周患病与就诊的影响因素及其潜在变量与居民两周患病与就诊之间的关系进行分析，一方面，探讨此方法应用于卫生服务研究的适宜性，另一方面，为卫生部门采取对策，改善医疗保健工作提供参考依据。
资料来源与方法
资料来源于1998年浙江省卫生服务调查家庭健康询问调查。该调查按照四阶段整群随机抽样法从全省共抽取12个县7200户住户，本次研究对象为其中15岁及以上居民，城市2906人，农村16278人。家庭健康询问调查的内容包括(1)调查家庭成员的基本情况；(2)家庭经济、住房情况及环境状况；(3)家庭成员享受的医疗保障制度和卫生服务可得性；(4)家庭成员的两周内病伤、半年内慢性病患病及就诊情况等。
本文将整个人群分为城市地区居民和农村地区屠民两个人群，初始模型共设
有生理状态、社会因素、经济状况、居住卫生条件、生活习惯、病残状态、卫生
可得性、两周患病、两周就诊等九个隐变量，分别用年龄、性别等二十个标志变
量来标识。采用SAS6．12对数据进行线性结构方程模型拟合，采用ADF法进行参
数估计。
结 呆
在城市模型中，病残状态和经济状况对两周患病有直接影响，其标化效应系
数分别为0．3854和－0．0604社会因素和生理状态通过病残状态对两周患病有间
接影响，其标化效应系数分别为－0．2012和 0．0513。两周就诊仅受两周患病的直
接影响，其标化效应系数为0．6663；其它各因素均为间接影响，其影响从大到
小依次为病残状态、社会因素、经济状况和生理状态，对两周就诊的标化效应系
数分别为 0．2568、－0．1341、－0．0402和 0．0342。
在农村模型中，病残状态、居住卫生条件和社会因素均为直接影响两周患病
的主要因素，其标化效应系数分别为0．叨15、0．1410和－0．0田6：经济状况除对
两周患病有直接影响外，还通过居住卫生条件和病残状态对两周患病产生间接作
用，其总的标化效应系数为－0．1242；生理状态通过病残状态对两周患病产生间
接影响，其标化效应系数为 0．1581。影响两周就诊的各因素中，两周患病与之
直接相关，其标化效应系数为0．7552；经济状况同时具有直接和间接作用，其
总的标化效应系数为－0．11 54；其它各因素均为间接影响，其影响从大到小依次
为病残状态、生理状态、居住卫生条件和社会因素，对两周就诊的标化效应系数
分别为 0．3334、0．1194、0．1065和－0．0443。
结 论
本文利用线性结构方程模型对居民两周患病与就诊的影响因素进行分析，探
索了各因素之间的直接与间接作用。在城市模型中，病残状态和经济状况对两周
患病起着直接作用，而生理状态和社会因素通过病残状态对两周患病起着间接作
用；各因素通过两周患病对两周就诊间接产生影响。在农村，影响两周患病的直
接因素除与城市相同的病残状态和经济状况外，还有社会因素和居住卫生条件，
间接因素除生理状态与城市相同外，还有经济状况；除了两周患病外经济状况对
两周就诊也起了直接作用外，其它因素对两周就诊的影响主要是通过两周患病起
间接作用。线性结构方程模型在卫生服务研究领域中有一定的适用性。
Objective
It is important to study the requirements and utilizations of health service including their affecting factors. In recent years, the requirements and utilizations of health service have changed remarkably. Reporting illness and seeking medical care within 2 weeks are important indexes of health service requirements and utilizations, which have many affecting factors and their relationships are complicated. Some researches have been done internal and external but many affecting factors are inconsistent in those researches.
Structural equation model is a method of linear statistical modeling, which makes the observed variables indicate the latent variables and allows the observed and latent variables have measurement errors. It can research not only the observed and latent variables but also the direct and indirect effect between these variables. In this study, structural equation model was used to analyze the relationship between the reporting illness > seeking medical care within 2 weeks and their affecting factors ^ latent variables.
Materials and methods
The data in this study came from the family health survey of Zhejiang Province in 1998. 7200 families from 12 counties were selected by four-stage cluster sampling. A total of 2906 residents lived in urban area and 16278 residents lived in rural area aged 15 or over were included for analysis. The content of the survey was composed of (1) general characteristics of family members; (2) economic state > housing conditions and surroundings of the family; (3) health insurance situation and
acquirement of health service of family members; (4) circumstance of family members about whether they had illness within 2 weeks and chronic disease within recent half year and whether they went to see doctors.
The whole participators were divided into urban population and rural population in analysis. The original model had 9 latent variables, which were indicated by 20 observed variables. SAS 6.12 was used to fit the structural equation model and the parameters were estimated by ADF method.
Result
In urban model, morbid status and economic conditions had direct effect on reporting illness within 2 weeks and the standardized coefficients were 0.3854 and -0.0604. Social factor and physiological conditions influenced reporting illness within 2 weeks indirectly through morbid status, the standardized coefficients were -0.2012 and 0.0513 respectively. Residents' seeking medical care within 2 weeks was influenced indirectly by all factors except the reporting illness within 2 weeks. The standardized coefficient of reporting illness to it was 0.6663. According to their influence to seeking medical care, those indirect factors were morbid status > social factor ., economy conditions and physiological conditions orderly and their standardized coefficient were 0.2568 > -0.1341 > -0.0402 and 0.0342 respectively.
In rural model, morbid status x sanitary conditions of family and social factor influenced the reporting illness within 2 weeks directly and their standardized coefficients were 0.4415s 0.1410 and -0.0586 respectively. Economic conditions not only had direct effect but also had indirect effect on residents' reporting illness through sanitary conditions of family and morbid status. The total standardized coefficient of economy conditions was -0.1242. Physiological conditions influenced residents' reporting illness within 2 weeks indirectly through morbid status and the standardized coefficient was 0.1581. Residents' seeking medical care was influenced by all factors. Reporting illness had direct relationship to it, the standardized coefficient was 0.7552. Economy conditions had both direct and indirect effect on it, the total standardized coefficient was -0.1154. Other factors had indirect effect on it.
According to their influence, those indirect factors were morbid status, physiology condition, Sanitary conditions of family and social factor orderly and their standardized coefficients were 0.3334, 0.1194> 0.1065 and -0.0443.
Conclusion
In th
研究人群卫生服务需求和利用及其影响因素是卫生服务研究的重要课题，近些年来，我国居民卫生服务的需求和利用发生了十分显著的变化。两周患病与就诊是反映居民医疗服务需求和利用的重要指标，影响两周患病与就诊的因素多且复杂，国内外对此进行了一定的研究，许多因素在已有的研究中结果并不一致。
线性结构方程模型是一种线性统计建模技术，能将显在变量作为隐变量的标识，并允许这两种变量存在测量误差。它不但能研究标志变量，又可研究不能直接观测的隐变量，既可研究变量间的直接作用，又可研究变量间的间接作用。本研究拟采用线性结构方程模型对居民两周患病与就诊的影响因素及其潜在变量与居民两周患病与就诊之间的关系进行分析，一方面，探讨此方法应用于卫生服务研究的适宜性，另一方面，为卫生部门采取对策，改善医疗保健工作提供参考依据。
资料来源与方法
资料来源于1998年浙江省卫生服务调查家庭健康询问调查。该调查按照四阶段整群随机抽样法从全省共抽取12个县7200户住户，本次研究对象为其中15岁及以上居民，城市2906人，农村16278人。家庭健康询问调查的内容包括(1)调查家庭成员的基本情况；(2)家庭经济、住房情况及环境状况；(3)家庭成员享受的医疗保障制度和卫生服务可得性；(4)家庭成员的两周内病伤、半年内慢性病患病及就诊情况等。
本文将整个人群分为城市地区居民和农村地区屠民两个人群，初始模型共设
有生理状态、社会因素、经济状况、居住卫生条件、生活习惯、病残状态、卫生
可得性、两周患病、两周就诊等九个隐变量，分别用年龄、性别等二十个标志变
量来标识。采用SAS6．12对数据进行线性结构方程模型拟合，采用ADF法进行参
数估计。
结 呆
在城市模型中，病残状态和经济状况对两周患病有直接影响，其标化效应系
数分别为0．3854和－0．0604社会因素和生理状态通过病残状态对两周患病有间
接影响，其标化效应系数分别为－0．2012和 0．0513。两周就诊仅受两周患病的直
接影响，其标化效应系数为0．6663；其它各因素均为间接影响，其影响从大到
小依次为病残状态、社会因素、经济状况和生理状态，对两周就诊的标化效应系
数分别为 0．2568、－0．1341、－0．0402和 0．0342。
在农村模型中，病残状态、居住卫生条件和社会因素均为直接影响两周患病
的主要因素，其标化效应系数分别为0．叨15、0．1410和－0．0田6：经济状况除对
两周患病有直接影响外，还通过居住卫生条件和病残状态对两周患病产生间接作
用，其总的标化效应系数为－0．1242；生理状态通过病残状态对两周患病产生间
接影响，其标化效应系数为 0．1581。影响两周就诊的各因素中，两周患病与之
直接相关，其标化效应系数为0．7552；经济状况同时具有直接和间接作用，其
总的标化效应系数为－0．11 54；其它各因素均为间接影响，其影响从大到小依次
为病残状态、生理状态、居住卫生条件和社会因素，对两周就诊的标化效应系数
分别为 0．3334、0．1194、0．1065和－0．0443。
结 论
本文利用线性结构方程模型对居民两周患病与就诊的影响因素进行分析，探
索了各因素之间的直接与间接作用。在城市模型中，病残状态和经济状况对两周
患病起着直接作用，而生理状态和社会因素通过病残状态对两周患病起着间接作
用；各因素通过两周患病对两周就诊间接产生影响。在农村，影响两周患病的直
接因素除与城市相同的病残状态和经济状况外，还有社会因素和居住卫生条件，
间接因素除生理状态与城市相同外，还有经济状况；除了两周患病外经济状况对
两周就诊也起了直接作用外，其它因素对两周就诊的影响主要是通过两周患病起
间接作用。线性结构方程模型在卫生服务研究领域中有一定的适用性。
Objective
It is important to study the requirements and utilizations of health service including their affecting factors. In recent years, the requirements and utilizations of health service have changed remarkably. Reporting illness and seeking medical care within 2 weeks are important indexes of health service requirements and utilizations, which have many affecting factors and their relationships are complicated. Some researches have been done internal and external but many affecting factors are inconsistent in those researches.
Structural equation model is a method of linear statistical modeling, which makes the observed variables indicate the latent variables and allows the observed and latent variables have measurement errors. It can research not only the observed and latent variables but also the direct and indirect effect between these variables. In this study, structural equation model was used to analyze the relationship between the reporting illness > seeking medical care within 2 weeks and their affecting factors ^ latent variables.
Materials and methods
The data in this study came from the family health survey of Zhejiang Province in 1998. 7200 families from 12 counties were selected by four-stage cluster sampling. A total of 2906 residents lived in urban area and 16278 residents lived in rural area aged 15 or over were included for analysis. The content of the survey was composed of (1) general characteristics of family members; (2) economic state > housing conditions and surroundings of the family; (3) health insurance situation and
acquirement of health service of family members; (4) circumstance of family members about whether they had illness within 2 weeks and chronic disease within recent half year and whether they went to see doctors.
The whole participators were divided into urban population and rural population in analysis. The original model had 9 latent variables, which were indicated by 20 observed variables. SAS 6.12 was used to fit the structural equation model and the parameters were estimated by ADF method.
Result
In urban model, morbid status and economic conditions had direct effect on reporting illness within 2 weeks and the standardized coefficients were 0.3854 and -0.0604. Social factor and physiological conditions influenced reporting illness within 2 weeks indirectly through morbid status, the standardized coefficients were -0.2012 and 0.0513 respectively. Residents' seeking medical care within 2 weeks was influenced indirectly by all factors except the reporting illness within 2 weeks. The standardized coefficient of reporting illness to it was 0.6663. According to their influence to seeking medical care, those indirect factors were morbid status > social factor ., economy conditions and physiological conditions orderly and their standardized coefficient were 0.2568 > -0.1341 > -0.0402 and 0.0342 respectively.
In rural model, morbid status x sanitary conditions of family and social factor influenced the reporting illness within 2 weeks directly and their standardized coefficients were 0.4415s 0.1410 and -0.0586 respectively. Economic conditions not only had direct effect but also had indirect effect on residents' reporting illness through sanitary conditions of family and morbid status. The total standardized coefficient of economy conditions was -0.1242. Physiological conditions influenced residents' reporting illness within 2 weeks indirectly through morbid status and the standardized coefficient was 0.1581. Residents' seeking medical care was influenced by all factors. Reporting illness had direct relationship to it, the standardized coefficient was 0.7552. Economy conditions had both direct and indirect effect on it, the total standardized coefficient was -0.1154. Other factors had indirect effect on it.
According to their influence, those indirect factors were morbid status, physiology condition, Sanitary conditions of family and social factor orderly and their standardized coefficients were 0.3334, 0.1194> 0.1065 and -0.0443.
Conclusion
In th
引文
1．龚幼龙．卫生服务研究．上海：上海医科大学出版社，1987：123～140
2．饶克勤，李青．多项式logistic回归分析在患者就诊行为影响因素中的应用．中国卫生统计，1999，16(2)：72～75
3．段明月，陈育德．北京东城区朝阳门医院保健地段0～14岁儿童患病与就诊的影响因素分析．中国卫生统计，1994，11(4)：8～10
4．黄方湧，宋沈超，张婷芬．贵阳社区儿童两周患病率及其影响因素．贵阳医学院学报，1998，23(1)：32～34
5．黄方湧，宋沈超，张婷芬．贵阳社区儿童两周就诊率影响因素分析．中国初级卫生保健，1998，12(5)：24～25
6．冯学山，范群辉，曹素华等．上海万航地段离退休人员医疗服务需求量分析．中国卫生事业管理，1992，8(8)：504～506
7．田俊，孙昌盛，张镜源等．福州市中小学教师卫生服务需求分析．中国学校卫生，1997，18(6)：462～464
8．田俊，孙昌盛，张镜源等．福州市中小学教师病残状态及其影响因素调查分析．中国公共卫生，1996，12(6)：257～259
9．张晓萍，张春江，张拓红等．中关村社区医疗保健卫生服务需要与利用调查分析(Ⅱ)．中国初级卫生保健，1997，11(4)：37～38
10．李士雪，成昌慧，刘兴柱等．山东省农村居民门诊卫生服务利用及其影响因素分析．中国卫生事业管理，1996，12(11)：611～614
11．曹素华，冯学山，翁仲华等．两周患病率影响因素分析．中国卫生事业管理，1991，7(7)：333～335
12．陈静，程茂金，陈翔等．深圳市职工卫生服务利用分析．中国卫生经济，1998，17(8)：34～35
13.Peek MK. Structural equation modeling and rehabilitation research. Am J Phys Med Rehabil., 2000, 79(3): 301-9
14.Hoyle RH. Introduction to the special section: Structural equation modeling in clinical research. Journal of Consulting and Clinical Psychology, 1994, 62: 427～521
15.Reis HT, Stiller J. Publication trends in JPSP: A three-decade review. Personality and Social Psychology Bulletin, 1992, 18:465～472
16.Fisher JO, Mitchell DC, Smiciklas-Wright H, et al. Maternal milk consumption predicts the tradeoff between milk and soft drinks in young girls'diets. Community and International Nutrition, 2000, 246～250
17.Noll RB, Gartstein MA, Vannatta, et al. Social, emotional, and behavioral functioning of children with cancer. Rediatrics, 1999, 103(1): 71～78
18.armelli D, Fabsitz RR, Swan GE, et al. Contribution of genetic and environmental influences to ankle-brachial blood pressure index in the NHLBI twin study. Am J Epidemiol, 2000, 151(5): 452～458
19.Faith MS, Pietrobelli A, Nunez C, et al. Evidence for independent genetic influences on fat mass and body mass index in a pediatric twin sample. Pediatrics, 1999, 104(1): 61～67
20．刘大维．结构方程模型在跨文化心理学研究中的应用．心理学杂志，1999，7(2)：48～51
21．张建新，张妙清，梁觉．殊化信任与泛化信任在人际信任行为路径模型中的作用．心理学报，2000，32(3)：311～316
22．陈冠民，倪宗瓒，张菊英等．Ⅱ型糖尿病危险因素的线性结构关系模型分析．中国卫生统计，1998，15(2)：12～16
23．王建琼，张菊英，倪宗瓒．骨质疏松症成因的证实性研究．华西医大学报，1997，28(1)：66～68
24．王海军，杭结城，贾秀敏等．农村老年人生活质量构成指标关系的探讨．中国老年学杂志，1995，15(2)：69～71
25.Cribbie RA. Evaluating the importance of individual parameters in structural equation modeling: the need for type I error control. Personality and Individual Differences, 2000, 29(3): 567-577
26．张一先，陈荣昭，董雪芳．城市大气环境质量比较评价分级方法．中国环境科学，1997，17(4)：357～359
27．张颖，倪宗瓒，姚树祥等．Box-Cox变换在估计危险因素平均暴露剂量中的应用．现代预防医学，2000，27(2)：180～181
28．高惠璇．SAS／STAT软件使用手册．北京：中国统计出版社，1997，76～110，228～248
29．郭志刚．社会统计分析方法——SPSS软件应用．北京：中国人民大学出版社，1999，339～371
30．孙尚拱．医学多变量统计与统计软件．北京：北京医科大学出版社，2000，163～216
31．陈冠民，张选群，陈华．多序列相关系数及其估计．数理医药学杂志，1999，12(2)：101～102
32．陈冠民，陈华，宋麒．多项相关系数及其估计．数理医药学杂志，2000，13(1)：26～27
33.Epstein JA, Griffin KW, Botvin GJ. A model of smoking among inner-city adolenscents: The role of personal competence and perceived social benefits of smoking. Prevention Medicine, 2000, 31:107～114
34.Fernandez E, Schiaffino A, Rajmil L, et al. Gender inequalities in health and health care services use in Catalonia(Spain). J Epidemiol Community Health, 1999, 53(4):218～222
35．杨辉．14个项目县的医疗服务需要量和卫生服务利用状况．中国农村卫生事业管理，1996，16(12)：29～53
36．高军，谢双保，高志强．国家级贫困县居民医疗服务需求及利用影响因素的分析．中国卫生统计，1998，15(6)：20～22
37.Fern(?)ndez de la Hoz k, Leon DA. Self-perceived health status and inequalities in use of health services in spain. Int J Epidemiol, 1996,25(3): 593～603
38.Henderson G, Akin J, Li ZM, et al. Equity and the utilization of heath services: report of an eight province survey in China. Social science and Medicine, 1994,39(5): 687～699
39.Mbugua Jk, Bloom GH, Segall MM. Impact of user charges on vulnerable group: the case of Kibwezi in rural Kenya. Social Science and Medicine, 1995,41(6): 829～835
40．国家卫生服务总调查报告——家庭健康询问调查部分初步报告．中国初级卫生保健，1995，9(4)：6～9
41．李筱蕾．家庭经济收入与医疗费用支出关系的调查研究．国外医学卫生经济分册，1995，15(1)：15～19
42．王建琼，倪宗瓒．潜在变量度量模型研究．中国卫生统计，1997，14(5)：21～22
2．饶克勤，李青．多项式logistic回归分析在患者就诊行为影响因素中的应用．中国卫生统计，1999，16(2)：72～75
3．段明月，陈育德．北京东城区朝阳门医院保健地段0～14岁儿童患病与就诊的影响因素分析．中国卫生统计，1994，11(4)：8～10
4．黄方湧，宋沈超，张婷芬．贵阳社区儿童两周患病率及其影响因素．贵阳医学院学报，1998，23(1)：32～34
5．黄方湧，宋沈超，张婷芬．贵阳社区儿童两周就诊率影响因素分析．中国初级卫生保健，1998，12(5)：24～25
6．冯学山，范群辉，曹素华等．上海万航地段离退休人员医疗服务需求量分析．中国卫生事业管理，1992，8(8)：504～506
7．田俊，孙昌盛，张镜源等．福州市中小学教师卫生服务需求分析．中国学校卫生，1997，18(6)：462～464
8．田俊，孙昌盛，张镜源等．福州市中小学教师病残状态及其影响因素调查分析．中国公共卫生，1996，12(6)：257～259
9．张晓萍，张春江，张拓红等．中关村社区医疗保健卫生服务需要与利用调查分析(Ⅱ)．中国初级卫生保健，1997，11(4)：37～38
10．李士雪，成昌慧，刘兴柱等．山东省农村居民门诊卫生服务利用及其影响因素分析．中国卫生事业管理，1996，12(11)：611～614
11．曹素华，冯学山，翁仲华等．两周患病率影响因素分析．中国卫生事业管理，1991，7(7)：333～335
12．陈静，程茂金，陈翔等．深圳市职工卫生服务利用分析．中国卫生经济，1998，17(8)：34～35
13.Peek MK. Structural equation modeling and rehabilitation research. Am J Phys Med Rehabil., 2000, 79(3): 301-9
14.Hoyle RH. Introduction to the special section: Structural equation modeling in clinical research. Journal of Consulting and Clinical Psychology, 1994, 62: 427～521
15.Reis HT, Stiller J. Publication trends in JPSP: A three-decade review. Personality and Social Psychology Bulletin, 1992, 18:465～472
16.Fisher JO, Mitchell DC, Smiciklas-Wright H, et al. Maternal milk consumption predicts the tradeoff between milk and soft drinks in young girls'diets. Community and International Nutrition, 2000, 246～250
17.Noll RB, Gartstein MA, Vannatta, et al. Social, emotional, and behavioral functioning of children with cancer. Rediatrics, 1999, 103(1): 71～78
18.armelli D, Fabsitz RR, Swan GE, et al. Contribution of genetic and environmental influences to ankle-brachial blood pressure index in the NHLBI twin study. Am J Epidemiol, 2000, 151(5): 452～458
19.Faith MS, Pietrobelli A, Nunez C, et al. Evidence for independent genetic influences on fat mass and body mass index in a pediatric twin sample. Pediatrics, 1999, 104(1): 61～67
20．刘大维．结构方程模型在跨文化心理学研究中的应用．心理学杂志，1999，7(2)：48～51
21．张建新，张妙清，梁觉．殊化信任与泛化信任在人际信任行为路径模型中的作用．心理学报，2000，32(3)：311～316
22．陈冠民，倪宗瓒，张菊英等．Ⅱ型糖尿病危险因素的线性结构关系模型分析．中国卫生统计，1998，15(2)：12～16
23．王建琼，张菊英，倪宗瓒．骨质疏松症成因的证实性研究．华西医大学报，1997，28(1)：66～68
24．王海军，杭结城，贾秀敏等．农村老年人生活质量构成指标关系的探讨．中国老年学杂志，1995，15(2)：69～71
25.Cribbie RA. Evaluating the importance of individual parameters in structural equation modeling: the need for type I error control. Personality and Individual Differences, 2000, 29(3): 567-577
26．张一先，陈荣昭，董雪芳．城市大气环境质量比较评价分级方法．中国环境科学，1997，17(4)：357～359
27．张颖，倪宗瓒，姚树祥等．Box-Cox变换在估计危险因素平均暴露剂量中的应用．现代预防医学，2000，27(2)：180～181
28．高惠璇．SAS／STAT软件使用手册．北京：中国统计出版社，1997，76～110，228～248
29．郭志刚．社会统计分析方法——SPSS软件应用．北京：中国人民大学出版社，1999，339～371
30．孙尚拱．医学多变量统计与统计软件．北京：北京医科大学出版社，2000，163～216
31．陈冠民，张选群，陈华．多序列相关系数及其估计．数理医药学杂志，1999，12(2)：101～102
32．陈冠民，陈华，宋麒．多项相关系数及其估计．数理医药学杂志，2000，13(1)：26～27
33.Epstein JA, Griffin KW, Botvin GJ. A model of smoking among inner-city adolenscents: The role of personal competence and perceived social benefits of smoking. Prevention Medicine, 2000, 31:107～114
34.Fernandez E, Schiaffino A, Rajmil L, et al. Gender inequalities in health and health care services use in Catalonia(Spain). J Epidemiol Community Health, 1999, 53(4):218～222
35．杨辉．14个项目县的医疗服务需要量和卫生服务利用状况．中国农村卫生事业管理，1996，16(12)：29～53
36．高军，谢双保，高志强．国家级贫困县居民医疗服务需求及利用影响因素的分析．中国卫生统计，1998，15(6)：20～22
37.Fern(?)ndez de la Hoz k, Leon DA. Self-perceived health status and inequalities in use of health services in spain. Int J Epidemiol, 1996,25(3): 593～603
38.Henderson G, Akin J, Li ZM, et al. Equity and the utilization of heath services: report of an eight province survey in China. Social science and Medicine, 1994,39(5): 687～699
39.Mbugua Jk, Bloom GH, Segall MM. Impact of user charges on vulnerable group: the case of Kibwezi in rural Kenya. Social Science and Medicine, 1995,41(6): 829～835
40．国家卫生服务总调查报告——家庭健康询问调查部分初步报告．中国初级卫生保健，1995，9(4)：6～9
41．李筱蕾．家庭经济收入与医疗费用支出关系的调查研究．国外医学卫生经济分册，1995，15(1)：15～19
42．王建琼，倪宗瓒．潜在变量度量模型研究．中国卫生统计，1997，14(5)：21～22