【作者】 董四平；

【导师】 方鹏骞；

【作者基本信息】 华中科技大学 ， 社会医学与卫生事业管理， 2010， 博士

【摘要】 一、研究目的与意义针对我国综合医院床位规模不断扩张的现实问题,对县级综合医院规模经济效率进行测量和评价,探讨综合医院发展是否存在规模经济现象,并进一步分析医院规模的影响因素。研究我国医院规模经济效率及其影响因素具有重要意义：一方面可以为医院建设和发展提供理论支持,另一方面可以为卫生行政部门制定区域卫生规划和监管医院运营提供决策参考和依据。二、研究目标与内容(一)我国综合医院规模发展的特点与趋势：对现阶段我国综合医院床位规模现状进行描述性统计分析,探寻医院规模发展的特点和趋势。(二)综合医院规模与财务风险的相关关系：定量分析分析床位医院规模与财务风险的相关关系,揭示医院规模扩张的潜在风险。(三)综合医院规模经济效率的实证研究：对不同规模医院的DEA效率得分进行分析,检验不同组别之间效率得分的差异程度,探寻医院规模与规模效率之间的证据,并初步界定湖北省县级综合医院的适宜规模。(四)综合医院规模的影响和决定因素研究：定性探讨并定量测量综合医院规模的内外部主要影响因素,以及影响因素的程度及其相互关系。(五)我国综合医院规模发展和供给行为的理论解释模型：在实证研究基础上提出医院规模发展和供给行为的理论解释模型,并初步提出实现医院规模经济发展的治理对策。三、研究方案与方法(一)研究对象本研究采取典型调查和整群抽样的方式确定研究样本。以湖北省所有61所县(含县级市)人民医院作为研究对象,根据研究目标将样本医院按实际开放床位分为小规模、中小规模、中等规模、中大规模、大规模五组。(二)资料来源(1)文献法：利用二次文献收集研究资料,收集样本医院2004、2005和2006三个年度的规模外部影响因素平行数据(Panel Data)。为保证数据的准确和真实性,主要收集来源于官方统计资料的数据。(2)现场调查：通过现场调查获取本研究所需数据和资料,包括医院三年的基本情况,内部影响因素指标等。(3) Delphi专家咨询法：建立综合医院规模影响因素指标体系,选取国内15名专家开展专家咨询,确定床位规模的内外部影响因素及其影响程度。(4)知情人深度访谈：利用知情人深度访谈等社会学调查方法进一步获取和补充现场调查未能涉及到的潜在信息,探究研究对象所面临的深层次问题。(三)调查和测量工具(1)《医院基本情况调查表》：包括医院基本情况指标,床位、人员、设备、房屋等投入指标,医疗服务量、业务收入等产出指标,医院规模内部影响因素指标。(2)《医院外部环境调查表》：主要内容包括影响医院规模和供给行为的测量变量。(3)《医院规模影响因素专家咨询表》：包括医院规模的一级影响因素维度和二级影响因素指标。(4)《医疗服务供给行为调查问卷》：调查内容包括对医院规模适宜程度的判断、医院规模的决定和影响因素、医院是否存在诱导需求及其程度、我国医院供给行为的适用模型等。(四)资料分析方法1、医院规模经济理论分析：利用文献研究内容分析(content analysis)和二次分析(secondary)对医院规模经济的研究理论进行系统总结和分析。2、医院规模发展现状与趋势分析：对现阶段我国综合医院床位规模现状进行描述性统计分析,探寻医院规模发展的特点和趋势。分析工具为SPSS V15. 0。3、医院床位规模与财务风险的相关关系分析：在运用因子分析(Factor analysis)建立医院财务风险评估模型并对其财务风险进行定量测量的基础上,应用回归法(Regression)分析医院规模与财务风险的相关关系。分析工具为SPSS V15. 0。4、医院规模经济效率实证研究：首先利用系统分析和文献优选法(Optimum seeking)建立综合医院规模经济效率测量指标体系,其次利用数据包络分析方法(DEA) CCR和BCC模型对样本医院行相对效率分析,获得医院综合效率、技术效率、规模效率和规模报酬状况的量化效率得分,最后对不同组别医院的DEA效率得分进行方差分析,检验不同组别之间效率得分的差异程度。分析工具为DEAP(V2.1)和SPSS(V15.0)。5、医院规模的影响因素研究：一方面通过专家咨询方法(Delphi)定性研究方法建立医院规模影响因素指标体系,明确影响因素的主要维度和权重；另一方面,以医院床位规模为因变量,内外部影响因素作为测量变量,构建医院规模影响因素的结构方程模型(SEM),定量刻画医院床位规模的主要影响因素、影响程度及其相互关系。应用工具为AMOS (V7.0)。6、医院供给行为模型分析：对通过问卷调查获取的有关供给行为和诱导需求的信息进行分析统计,验证国外学者提出的医院供给行为理论模型(包括效用最大化、收入最大化、医生控制模型等)。四、研究结果(一)我国综合医院规模发展描述性分析我国医院床位规模总体水平居世界平均水平,但综合医院平均床位数量明显高于世界平均水平,截止2008年,800张床位以上医院数量达到418张,县级综合医院平均床位数量也明显超出其他国家同级别医院。(二)医院床位规模与财务风险之间的关系湖北省县级综合医院实际开放床位与财务状况综合评分之间存在三次方曲线回归关系,回归方程为：Y=-1.19×10-7xX3+0.046×X-4.867(三)基于DEA方法的医院规模经济效率研究1、县级综合医院存在规模经济现象,医院效率随着床位的增加呈现增长的趋势,但在达到一定规模之后,效率开始下降。2、县级综合医院的适宜规模：以DEA相对效率的测量结果为基础,结合医院规模和财务风险回归方程,初步确定县级综合医院实际开放床位最佳规模范围在250至300张之间。3、县级综合医院规模经济收益情况：绝大多数县级大型综合医院(床位大于335张)处于规模报酬递减状态,同时绝大多数小型医院(床位小于200张)呈现规模报酬递增状态。4、DEA投入、产出投影分析结果显示：县级综合医院在实际开放床位和人员方面存在较大过剩,特别是大规模医院过剩比例高达22.7%,平均可以减少97张床位。(四)医院规模影响因素定性研究结果医院规模影响因素包括五个一级影响因素和30个二级影响因素：一级影响因素依据其影响权重排序分别是：医院内部因素(22.03%)、区域经济水平(20.53%)、卫生资源(19.77%)、社会状况(19.02%)和竞争环境(19.02%)；二级影响程度权重排在前10位的指标分别是：区域内总人口、医院领导决策类型、城镇居民人均可支配收入、农村居民人均纯收入、医院人均总收入、城镇人口比例、农村居民人均生活费总支出、全社会从业人员比例和区域卫生规划执行情况。(五)医院规模影响因素定量研究结果1、区域内经济人口状况是开放床位的最重要影响因素,其标准化回归系数为0.477,其代表指标是农村居民纯收入这一指标。2、医院内部经营绩效是医院规模的重要影响因素,其回归系数为0.324,人均业务收入和人均业务支出对医院绩效水平存在共同影响,且影响程度相当。3、医院的竞争环境和能力是医院实际开放床位数的中度影响因素,表明市场竞争是影响县级综合床位规模的影响因素,但尚不能对医院规模发展构成重要影响。4、卫生资源存量水平对县级人民医院规模发展存在较轻程度的负向影响,即随着区域内卫生资源配置水平提高,县级人民医院床位数略有减少。5、医院每百门诊人次住院比率、病床使用率、医院人均业务收入三者之间存在密切的关系。医院每百门诊人次住院比率可作为判断医院是否存在诱导需求及其诱导程度的敏感指标。(六)医院供给行为及其理论解释模型93.3%的咨询专家认为我国非营利性医院存在供方诱导需求现象,其中认为十分严重和比较严重的比例分别为28.6%和64.3%,二者累计比例达到92.9%；咨询专家认为可以解释我国非营利性医院供给行为的模型所占比例分别是：医院利润最大化模型(42.86%),两个群体模型(28.57%),而社会效用最大化模型仅为(14.29%)。五、研究结论1、我国不同级别综合医院平均床位数量明显超出世界主要国家水平,并具有进一步扩大的趋势。2、湖北省县级综合医院财务状况与实际开放床位在一定区间内呈非线性的三次方回归关系,当床位在280张左右时财务风险最小,应适当控制县级人民医院规模。3、湖北省县级综合医院存在规模经济效率现象,即医院效率随着床位的增加呈现增长的趋势,但在达到一定规模之后,综合效率和规模效率开始下降。4、湖北省县级医院床位规模整体过剩,初步确定县级综合医院实际开放床位适宜规模范围在250至300张之间。5、县级综合医院床位规模的最主要的影响因素包括三个方面：医院内部因素、区域经济水平和社会状况。市场竞争环境是医院规模的中度影响因素,但尚不能对医院规模发展构成重要影响,县人民医院在当地医疗市场处于龙头和垄断地位；卫生资源存量水平对县级人民医院规模发展影响程度不大,表明县级医院的发展并没有考虑当地卫生资源的存量水平。6、我国综合医院床位规模过剩,医疗服务市场存在较为严重的诱导需求,其内在的理论解释是：医院为追求平均业务收入最大化,通过诱导需求提高病床使用率,病床使用率的提高给医院决策者提供了进一步扩大规模的理由和假象,床位规模扩大之后,诱导需求现象更加严重,由此形成恶性循环。7、为遏制我国综合医院规模扩张,建议采取以下政策措施：(1)建立医院的合理补偿机制消除医院的逐利动机；(2)取消医院的经营剩余自主分配权；(3)严格医疗机构设置规划形成制度性“硬”约束；(4)提高医院内部决策和管理水平；(5)建立医疗信息披露制度增加医院诱导需求成本；(6)改革医疗服务提供方式,促进非营利性医院经营方式转变。六、研究创新与价值(一)理论创新通过对湖北省所有61家县(包括县级市)人民医院的实证研究,证实了县级综合医院存在规模经济现象,初步澄清了我国医院是否存在规模经济这一重大理论问题。本研究还从定性和定量研究两个方面探寻了影响医院规模的因素,解释了我国医院规模扩张的动机,并进一步提出了我国医疗服务的供给行为理论解释模型。(二)方法创新本研究以跨学科的视角,将管理学、经济学、社会学、会计学、计量经济学和统计学多学科研究方法引入医院规模经济效率及其影响因素研究,特别是综合运用定性的Delphi专家咨询法和定量的结构方程模型对医院规模的影响因素进行了深入系统分析,这一研究方法克服了局限于某一方面方法研究的局限性,研究结果可以相互验证和补充,因此研究结果和结论更为可信。(三)政策价值从理论方面来看,县级综合医院规模经济效率的证实和适宜规模的界定,不仅可以为医院发展提供理论支持,避免扩张的盲目性,提高医院医疗资源的规模效率,而且有利于政府对医疗资源的配置提供决策依据,避免卫生资源的浪费。从应用方面来看,本研究的定量研究结果和结论,包括医院规模与财务风险的非线性回归方程、医院规模影响因素指标体系和结构方程模型等,可以作为医院管理者和卫生行政部门的决策依据。七、研究局限与不足(一)本研究以湖北省县级医院为研究对象得出了综合医院存在规模经济效率的结论,但以此推论我国综合医院存在规模经济效率现象证据尚不充分。今后还需要扩大研究对象,进一步研究省部级、地市级综合医院的规模经济效率问题,以证实规模经济效率现象的存在。(二)本研究所收集的医院规模影响因素尚不十分完备,在今后的研究中应注重探索医院规模更多潜在的影响因素。更多还原

【Abstract】 1. Research purposes and significanceIn mind the existing expansion of beds in general hospital, the economies of scale in efficiency of county general hospital are measured to clarify the existence of economies of scale in general hospitals, and the impacts of hospital size were explored. This Study is important:on the one hand from a theoretical point of view, it help to provide theoretical support for of the hospital’s development; the other hand; from a practical point of view, the scale factors research could provide direct policy and references for the health administration in the health planning and supervising for hospitals.2. Research objectives and content2.1 The present status of general hospital beds was decrypted by statistical analysis to explore the characteristics and trends of the scale of Chinese general hospital.2.2 The correlation between of bed size and financial risk of general hospital was quantitatively analyzed to reveal the potential risk for the scale expansion of hospital.2.3 An empirical study of efficiency by DEA analysis with different sizes hospitals was conducted to test the significant differences of scale between them.2.4 The qualitative method and quantitative measurement were employed to certain the internal and external factors of general hospital size, their weights and their relationships.2.5 On the basis of the empirical study, one theory mode was put forward to explain the supply behaviors of general hospital, and the countermeasures were proposed to achieve economies of scale in the development of hospital.3 Research programs and methods3.1 Sampling and CasesThe typical investigation and the cluster sampling were employed to determine the study sample. As the research cases, all 61 county People’s Hospital in Hubei province(including county-level city) was divided into five groups(small, medium and small scale, medium scale, medium scale, large-scale group) according to the bed scale.3.2 Data source(1) Literature method:Use secondary sources to collect research data (including the external factor indicators, the data years are 2004,2005 and 2006 (Panel Data). In order to ensure the accuracy and authenticity, the data are mainly from official statistics.(2) Field survey:Collect the panel data and information of 2004,2005 and 2006 including basic situation, the internal factors indicators through field investigation.(3) Delphi expert consultation:The nationwide 15 experts were selected to develop the system of influencing factors for the bed scale of general hospital.(4) Insider interviews:The sociology of insider interviews and other investigative methods were used to gather potential information for the study.3.3 Survey and measurement tools(1) Questionnaire for Basic information of hospital;(2) Questionnaire for the external environment of hospital;(3) Expert consultancy List of factors for scale of Hospital;(4) Questionnaire for supply behavior of medical services in hospital;3.4 Data analysis(1) The literature content analysis and secondary analysis were used to summarize the theory of economies of scale in hospital(2) The descriptive statistical analysis was employed to describe the tendency of beds scale in general hospital for exploring its characteristics and tendency in software SPSS (V15).(3) On the base of quantitative measurement of hospital financial risk by factor analysis, the regression method was used to the correlation between hospital beds and financial risk in software SPSS (V15).(4) The empirical study of the scale efficiency in general hospital:first, the use of literature-optimum-seeking to establish an index system of efficiency measurement; second, using data envelopment analysis (DEA) of the relative efficiency to calculate the efficiency score; and finally, the analysis of variance was used to test the significance different among different groups. This was finished in DEAP (V2.1) and SPSS (V15).(5) The study of factors affecting hospital scale:on the one hand, the method of Delphi was employed to establish the influencing factors system for bed scale of the hospital to determent the factors dimensions and their weights; the other hand, with the hospital beds as the dependent variable, the internal and external factors as measured variables, the structural equation modeling (SEM) of scale factors was built to quantitatively descript the main factors, their degree and their relationship. The application tool is AMOS (V7.0).4. Results4.1 The overall level of hospital beds per capital is equivalent to the world average, but the average number of beds in general hospital was significantly higher than the world average. By 2008, the number of hospitals over 800 beds to 418 and the average number of beds at County General Hospital was also higher than the hospital of other countries4.2 The relation between the financial status and beds size of County General Hospital could be estimated with cubic regression curve, and the regression equation is: Y=-1.19×10-7×X3+0.046×X-4.8674.3 The findings of DEA efficiency analysis in hospitals:(1) Hospital efficiency increased as a growing trend with the scale of beds, but when the scale reached a certain size, the efficiency started to decline.(2) Based on the scores of DEA results and combined with regression equation between the hospital size and financial risk, the range of optimal beds size has been determined as 250 to 300 in county general hospital.(3) The majority of large general hospitals at county level (greater than 335 beds) in a state of decreasing returns to scale, while the vast majority of small hospitals (less than 200 beds) showed increasing returns to scale status.(4) DEA input and output projection analysis showed that county general hospital beds and staff in are actual surplus, especially in large-scale hospital surplus as high as 22.7%, the average can be reduced by 97 beds.4.4 The factors analysis of Hospital-scale by qualitative methodThe factors of Hospital-scale include 5 first-level factors and 30 second-level factors. The first-level factors and their impact weights are: hospital internal factors (22.03%), regional economic level (20.53%), health resources (19.77%), social status (19.02%) and the competitive environment (19.02%). The top 10 indicators of second-level factors and their weights are: the region’s total population, the hospital leadership decision-making type, per capita disposable income of urban residents, per capita net income of rural residents, hospitals per capita income, the ratio of urban population, the total expenditure of rural residents per capita cost of living, the proportion of the whole society practitioners and regional health planning implementation.4.5 The factors analysis of Hospital-scale by Quantitative study(1) The demographic and economy situation in the region is the most important factors for beds size, and its standardized regression coefficient was 0.477. In addition, its representative indicator is the net income of rural residents.(2) Hospital performance is also a major factor for scale, and its regression coefficient was 0.324. The per capita revenue and per capita expenditure has common effects on hospital performance.(3) The hospital’s competitive environment is the moderate factors for hospital beds, which shows the competition is a factor but not yet play an important influence to the scale of the hospital scale.(4) As the stock level of health resources on the scale of the development of county general hospital, there is lesser degree of negative impact, that is, with the allocation of health resources within the region raise the level of beds decreased slightly.(5) There are close relationships among the ratio of inpatient per 100-outpatient, occupancy rate of beds and hospital income per capita. The ratio of inpatient per 100-outpatient can be used as a sensitive indicator to judge the existence and the degree of induced demand in hospital.(6) 93.3%of the consulting experts believe that there exists the phenomenon of supplier-induced demand in China’s non-profit hospitals, which is very serious and more serious that the proportions were 28.6%and 64.3%respectively, and the cumulative percentage of them reached 92.9%; the proportion of model that can explain the supply behavior of non-profit hospitals are:the model of hospital profit maximization (42.86%), two groups (28.57%), and social utility maximization only (14.29%).5. Conclusions5.1 The beds size of general hospital in China at different levels significantly exceeded the average level of the world’s major countries, and the explosion has continued further.5.2 There is nonlinear parabola regression relationship between financial status and open beds of county general hospital of Hubei Province. Therefore, from a financial point of view to reduce the risk, have to control the size of county-level hospital.5.3 The economies of scale has been confirmed in county general hospital of Hubei Province:the efficiency of hospital increased with the beds growing trend, but reached a certain scale, the overall efficiency and scale efficiency began to decline.5.4 The overall size of county surplus of hospital beds is redundant and the optimal size is in the range of 250-3005.5 The most important factor foe beds scale of county general hospital include three aspects:the hospital internal factors, the level of regional economic and social conditions. The market competition is a moderate factor but can not yet a significant impact, which shows the county people’s hospital is in leading and monopoly status in local health care market. The stock levels of health resource has little impact and hinted that the development of county hospitals do not take account of local health resources stock levels.5.5 The intrinsic theoretical explanation for the redundant in scale of beds and the serious induce demand in China’s general hospital is:hospital in pursuit of revenue maximization, utilization rate of bed increasing by inducing demand, hospital decision-makers further expand the scale of bed based on the reason and illusion of bed occupying rate, and the induced demand behavior is more serious, thus a vicious circle.5.6 The following policy measures were recommended to curb the scale expansion of China’s general hospital t:(1) establish the hospital’s reasonable compensation mechanism to eliminate its profit motive; (2) cancel the hospital’s independent right of sharing economic benefits; (3) strict medical institutions planning to set "hard" constraints; (4) improve the internal decision-making and management of hospitals; (5) establish medical information disclosure system to increase the cost of hospital’s behaviour of induce demand; (6) reform of health care delivery to promote non-profit hospital management pattern.6. Innovations and merits6.1 Theoretical InnovationThe empirical research with all 61 county general hospital of Hubei province (including county-level city) confirms the existence of economies of scale in general hospital, initially clarifies whether there are economies of scale in Chinese hospitals, which is an important theoretical issue. The study also explored factors that influence the size of the hospital from both qualitative and quantitative research, explained the motivation for the expansion of hospital size, and further developed an explain model theory for supply behavior of medical services in China.6.2 Innovation of methodsFrom an interdisciplinary perspective, the multi-disciplinary methods, such as management, economics, sociology, accounting, econometrics and statistical method, are employed to research the economies of scale in hospital and its impact factors. Particularly, the comprehensive utilization of methods with qualitative Delphi method and quantitative structural equation model was used to exploring the factors influencing the size of bed. This method overcomes the limitations of the single methods, the results can be verified and mutual complemented, and then the conclusions are more credible.6.3 The merits of researchFrom the theoretical point of view, this study confirmed the existing of economies of scale in hospital bed, not only providing theoretical support for the development of hospital to avoid blind expansion, but also helping the Government provide health care resource allocation decisions to avoid the waste of health resources. From the application point of view, the quantitative results and the conclusions of this study, including nonlinear regression equation between hospital size and financial risks, the Delphi system and structural equation models of scale factors, can be used as basis of decision making for hospital administrators and health administrative departments7. Research limitations7.1 There is not sufficient evidence to proof the existence of economies of scale in hospital, for this study is under the background of county general hospital in Hubei Province. The future research cases need to be expanded to the provincial, municipal general hospitals for confirming the existence of economy of scale in China general hospitals.7.2 The factors for bed scale of hospital in this study is not yet complete, so in future research should focus on more potential factor to explore the impacts of hospital scale.更多还原