【作者】 王芳；

【导师】 卢祖洵；

【作者基本信息】 华中科技大学 ， 社会医学与卫生事业管理， 2006， 博士

【摘要】 研究目的 自2000年世界卫生组织(WHO)在卫生系统倡导卫生服务绩效评价以来，其思想和方法已引起了各国政府的高度重视。社区卫生服务系统作为卫生系统的一个子系统，正在我国得到迅速的发展，对该系统的绩效进行科学化的评价将有助于发现其中存在的问题并促进其健康发展。 本研究旨在通过采用定性和定量的研究方法建立一套切合实际的、具有较高效度、信度和区分度的社区卫生服务绩效评价指标体系以提高社区卫生服务管理的规范化、科学化程度，同时探讨社区卫生服务评价指标的评价方法、方式以及绩效综合评价模型，为将来社区卫生服务绩效评价实践提供一些方法学基础和建议。此项研究对于我国社区卫生服务的持续健康发展具有重要的现实意义，同时为相关政策的制定提供坚实的理论和方法学基础。 研究方法 本研究采用了文献复习法、专家论证法、Delphi专家咨询法用于建立指标体系并采用模糊数学综合评判法计算指标权重系数。利用现场调研的方法探讨了指标的综合评价方法、快速评价方法和间接评估方式。采用因子分析法、克朗巴哈α(Cronbach’s alpha)系数计算、聚类分析和方差分析等数理统计方法进行指标体系的效度、信度和区分度的检验。最后采用加权累加综合评分法、Topsis法和秩和比法等综合评价方法对社区卫生服务绩效进行了实证分析并采用平均值法、Borda法、Copeland法和模糊Borda法等组合评价方法对综合评价结果进行了组合，探讨了组合评价结果的一致性与合理性。 研究内容 本研究主要以行政区为评价对象，从第三方评价的角度对社区卫生服务系统的绩效进行评价指标体系的研究。主要内容包括：(1) 通过查阅和研究系统论、绩效与绩效评价理论、社区卫生服务理论以及国内外相关研究成果等，确立指标体系的基本框架和结构；(2)通过专家论证和Delphi专家咨询法进一步筛选指标并采用模糊综合评判法计算指标的权重；(3)利用深圳现场调查资料探讨指标评价方法如洛伦茨曲线、基尼系数、Topsis法和模糊数学综合评价法用于评价资源配置公平性、满意度和反应性等客观指标；(4)利用湖北省、重庆和贵州等社区卫生服务现场评价活动探讨社区卫生服务快速评价方法和方式，如皮质量保证抽样法(LQAS)、模拟病人法、拦截调查等；(5)采用数理统计方法评价指标体系的效度、信度和区分度等；(6)探讨加权累加综合评分法、Topsis法和秩和比法等综合评价方法在社区卫生服务绩效评价中的应用；(7)利用组合评价方法对综合评价结果进行组合，探讨组合评价结果的一致性与合理性。 研究结果 通过上述一系列的研究活动，本次研究取得了如下研究结果：(1)构建了以支持、服务功能落实和绩效三大类指标为主要结构的具有层次性的指标体系，包括一级指标三个、二级指标16个、三级指标56个。其中支持类指标主要包括政策支持、资金投入、服务机构与网络建设以及人员配备等四个二级指标；服务功能落实包括社区卫生的医疗、预防、保健、康复、健康教育和计划生育等六大功能指标；绩效指标包括效率、效益、卫生服务利用、健康改善、感受性指标以及公平性等6个二级指标；由于三级指标数目较多，在此不进行详述。(2)确定了一、二、三级各指标的权重值，其中三个一级指标的权重值分别为0.3396737、0.336193、0.32417；16个二级指标的权重分别为0.082637、0.086619、0.083327、0.087053、0.060558、0.059508、0.055785、0.058028、0.052899、0.049415、0.053786、0.05649、0.055146、0.053182、0.052968、0.052597。(3)二级指标因子分析结果显示，共有三个公因子提取，其累计贡献率达到78.556％，且三个公因子的特征根均大于1。其中F1受变量8、11、12、13、15、16的影响，可看作社区卫生服务成绩和效果；F2受变量5、6、7、9、10、14的影响，可看作社区卫生服务的过程和质量；F3受1、2、3、4的影响，可看作社区卫生服务的支持和网络建设。分析结果与预设结构基本相符，具有较好的效度。(4)通过计算克朗巴哈系数，二级指标的克朗巴哈α标准值为0.919(F=92.154，P=0.0000)，三级指标的克朗巴哈α标准值为0.959(F=31.238，P=0.0000)，结果显示指标的重测信度较好。(5)根据聚类分析，十二个评价地区分为三类比较合适，方差分析结果显示三类之间和每两类之间的差异具有统计学意义(P＜0.05)，指标体系具有较好的区分度。(6)分析不同地区二级指标加权得分的差异发现，政策支持、资金投入、服务机构和网络建设、人员配备、社区康复、社区计划生育、社区卫生服务效益和感受性指标这几项的得分是造成地区总体绩效差异的主要影响因素。(7)现场调查结果显示，洛伦茨曲线和基尼系数评价可用于社区卫生服务资源配置公平性评价，而Topsis法和模糊数学综合评判法可分别用于评价满意度和反应性等指标，所得结果可靠、准确。(8)采取定性和小样本抽样的LQAS流行病学快速评价法同样可用于社区卫生服务质量指标的考评，具有快速、便捷的特点。(9)国家级社区卫生服务评价现场结果显示，间接评估方式与直接评估具有互补作用，对于发现社区卫生服务中存在的问题以及获得真实的细节具有非常显著的效果。(10)通过实证研究发现，加权累加综合评分法、Topsis法和秩和比法三种不同的评价模型所得结果具有较好的一致性，Spearman等级相关系数分别为0.9371和0.8531。(11)对综合评价结果进行组合后发现，四种组合评价结果同样具有一致性，Kendall(肯达尔)一致性协和系数为W~a=0.994(P=0.000)。与单一综合评价结果相比，组合评价结果更加合理。 结论 基于上述研究结果，本研究得出了以下几点结论与相关建议：(1)建立的指标体系覆盖面广，有效、可信；(2)评估指标的筛选方法是合理的；(3)确定指标权重系数的方法是科学的，结果是相对准确的；(4)定性与定量的有机结合应该是指标体系研究的发展方向；(5)各个综合评估模型所产生的结果基本一致，组合评价结果将误差降至最低，结果可信；(6)采用何种评价方法应该根据具体情况和条件选择应用；(7)社区卫生服务绩效受到多种因素的影响，提升社区卫生服务的总体绩效水平有待于全面提升和加强各项能力，特别是关键要素；(8)根据研究结果，基本可以确定各项指标的重要性、灵敏性、可操作性以及指标的信度和效度等。更多还原

【Abstract】 Research PurposesSince performance assessment of health service has been advocated by world health organization (WHO) in the year 2000, the ideas and methods have been highly paid more attention by the countries around the world. As one of branches of health system, community health service (CHS) has being developed rapidly, a scientific evaluation on this system will be helpful for us to find the problems and promote its further development.This study was aimed to use qualitative and quantitative methods to establish an indicator system with highly practicability, validity, reliability and differentiation for evaluating performance of CHS to improve the degree of management of CHS in terms of standardization and scientificity. In the meantime, the methods, manners and the models of synthetical evaluation were also explored in this study to provide some of basis in methodology for evaluation practice. This study has important practical significance in promoting sustainable development of CHS and providing theoretical and methodological basis for policy making.Research MethodsIn this study, literature searching, expert discussion and Delphi method were used to establish the indicator system, and fuzzy mathematical method was used to calculate the weight of the indicators. The synthetical evaluation methods of indicator, rapid assessment methods and indirect evaluation manners were also explored by utilizing field survey data. Factor analysis, Cronbach’s alpha coefficient, cluster analysis and ANOVA analysis were employed to test the validity, reliability anddifferentiation of indicator system. Lastly, accumulative synthtical grading evaluation , Topsis method and Rank Sum Ratio (RSR) method were applied to evaluate the performance of CHS synthetically at more than 10 districts in Hubei province, meanwhile, four combination methods, namely average method, Borda method, Copeland method and fuzzy Borda method, were used to combine the results of the four synthetical evaluation, and the consistency and rationality were also explored.ContentsIn this study, the administration district in city was seen as evaluation object, and the performance was evaluated from the third side perspectives. The contents include: (1) reviewing system theory, performance and performance assessment theory, CHS theory and some of relevant researches and establishing the basic framework of indicator system; (2) by using experts discussion and Delphi method to filter the indicators and calculate the weight of indicators by using fuzzy mathematical evaluation method; (3) through Shenzhen field survey, applying and exploring some of methods for indicator evaluation like equity, satisfaction and responsiveness, those methods include Lorenz Curve, Gini coefficient, Topsis method and fuzzy synthetical evaluation and etc.;(4) through the field activities in Hubei province, Chongqing and Guizhou province, the methods of rapid evaluation and indirect evaluation manners such as LQAS, patient imitation and heading off; (5) validity, reliability and differentiation were tested by using mathematical statistical methods; (6) exploring the application of some of synthetical evaluation methods such as accumulative synthetical grading, Topsis and RSR in CHS performance assessment; (7) combining the results of four synthetical evaluation and discussing the consistency and rationality of the results of combination.Resultsthe results of this study are as follows: (1) having constructed a hierachical indicator system composed of three groups, namely supporting, service process and performance, including three first class indicators, 16 second ones and 56 third ones. The supporting group is composed of policy support, financial input, organization andnetwork construction and health personnel and etc. service process group is composed of six functions of CHS, namely medical service, prevention, health care, health education, rehabilitation and family planning; the performance group is composed of the indicators of efficiency, benefits, service utilization, health improving, equity, satisfaction and responsiveness and etc.; (2) having confirmed the weight of the indicators of three classes. The weight of three indicators of the first class are 0.3396737, 0.336193 and 0.32417 respectively; the weight of 16 second class indicators are 0.082637, 0.086619, 0.083327, 0.087053, 0.060558, 0.059508, 0.055785, 0.058028, 0.052899, 0.049415, 0.053786, 0.05649, 0.055146, 0.053182, 0.052968, 0.052597 respectively. (3) the factor analysis of the second class indicators showed that totally three common factors were extracted, the accumulative variance reached 78.556% and initial eigenvalues of these three common factors are bigger than 1. The common factor 1 (F1) is mainly influenced by variable 8,11,12,13,15 andl6 , it could be named as the achievement and effectiveness of CHS; F2 is mainly influenced by variable 5,6,7,9,10 and 14, named as service process and quality; F3 influenced by variable 1,2,3 and 4, and named as supporting and network construction. The results of factor analysis are accordant with the expected. On the whole, the indicator system has good validity. (4) Cronbach’s alpha coefficient of the second class indicators is 0.919 ( F = 92.154, P = 0.0000) and the third class is 0.959 ( F = 31.238, P = 0.0000) , it was showed that the reliability of the indicators is very good. (5) according to the cluster analysis, it is appropriate to divide the 12 districts into 3 groups, and the ANOVA analysis showed that the differences among the three groups and between every two groups are statistical significant ( P<0.05 ) , the differentiation of the indicators is good. (6) it was found by analyzing the difference of the scores of the second class indicators among the three groups that policy supporting, financial input, organization and network construction, health personnel, community rehabilitation, family planning, benefits, satisfaction and responsiveness are the key factors influencing the performance of CHS in the 12 districts. (7) the results of field surveys showed that Lorenz Curve and Gini coefficient could be used to evaluate the equity of resource allocation, and it is reliable to use Topsis method and fuzzymathematical method to evaluate the satisfaction and responsiveness; (8) some of qualitative and quantitative methods (LQAS) could be used for the rapid evaluation of some indicators, it has the characters of rapidness and convenience;(9) the field activity of CHS evaluation on national level showed that the manners of indirect evaluation was important and effective for us to find the problems and acquire the details, direct and indrect manners can make up for each other to assure the truth of the information; (10) three evaluation models, namely accumulative synthetical grading method, Topsis method and RSR method had highly consistency, the Spearman rank correlation coefficients were 0.9371 and 0.8531 respectively. (11) it was found by combining evaluation results that the four combination results have highly consistency, the Kendall coefficient is W~a = 0.994 ( P = 0.000 ). Compared with the separate evaluation, the combination result is much reasonable.ConclusionsBased on the results of the research, the conclusions and recommendations could be put forward. (1) the scope covered by the indicator system is broad, and the whole system is valid and realiable; (2) the method for filtering indicators is reasonable; (3) the method of confirming the weight of indicators is scientific, and the result is relatively accurate; (4) the organic integration of quantitative and qualitative methods is the direction of indicator system research in the future; (5) the results from, different synthetical evaluation models are basically consistent, the combination evaluation of the results has reduced the error lowest and the combination results are realiable and reasonable; (6) the method chosen for evaluation is depended on the practical situation and conditions; (7) CHS performance is influenced by lots of factors, the stronger the ability of all aspects, especially the key factors, the better the performance; (8) based on the research results, the importance, maneuverability, sensitivity, validity and reliability of the indicators could be summarized.更多还原