【作者】 刘兆兰；

【导师】 徐飚；

【作者基本信息】 复旦大学 ， 流行病与卫生统计学， 2009， 博士

【摘要】 我国糖尿病患者数增长迅速，1980年仅有糖尿病患者480万，2003年则达2300万，2007年达到3980万，预计到2025年我国糖尿病患者数将达到4600万。我国糖尿病患者数仅次于印度，居世界第2位。2型糖尿病（type 2 diabetes mellitus，T2DM）患者约占糖尿病患者总数的90-95％。随年龄增加，糖尿病患病率逐渐增加，因目前我国正处于老龄化社会的加速发展阶段，65岁以上的老年人口达1亿以上，随着老年人口的增加和老年人口糖尿病患病率的升高，我国老年糖尿病患者数将不断增加。糖尿病患者的分布与经济发展水平密切相关，具有明显的城乡差别；我国城市糖尿病患病率显著高于农村，且城市糖尿病患病率增长较快。由于目前T2DM在临床上缺乏有效的治愈手段，临床治疗目的主要是控制血糖，延缓或减少合并症的发生，对已有合并症的患者进行对症治疗。随着医学技术和医疗理念的发展，T2DM的临床治疗和管理方案正在逐步改进；同时，T2DM患者的教育和自我管理越来越受重视。随着健康理念和生命价值观的不断发展，T2DM患者的治疗满意度、生存质量等也越来越引起医务科研工作者以及政府和社会的关注。而另一个不可忽视的方面是T2DM造成了严重的医疗及社会问题，即T2DM患病和早死给病人、家庭和社会造成了巨大的社会经济成本。我国是世界糖尿病第2大国，超过3000万的糖尿病患者的管理已经成为我国一个重要的医疗卫生问题和社会问题。糖尿病的控制和管理需要以设计科学，实施严格，多学科交叉的大样本流行病学和临床研究结果为依据，但目前国内这方面的系统研究还比较少。本课题的研究现场为分别位于我国东、南、西、北的四个大城市（上海、广州、成都、北京），研究对象为各城市4-7家二、三级医院门诊和住院T2DM患者。本课题采用横断面研究设计，将多学科（流行病学、内分泌学、卫生经济学、行为心理测量学）研究方法相结合，首先了解我国大城市T2DM患者的临床治疗和自我管理现状、血糖控制情况，并对影响T2DM控制的相关因素进行探讨；估计合并症患病率以及合并症对T2DM患者生存质量和疾病经济负担的影响；通过对患者当前治疗满意度、生存质量和疾病经济负担的评价，了解我国T2DM患者的治疗满意度、生存质量现状和疾病经济负担。主要研究方法和结果如下：1．2型糖尿病门诊患者综合管理现况研究。以4个城市15家医院连续纳入的1524名T2DM门诊病例为研究对象；通过对T2DM门诊患者的治疗、自我管理及相关因素的现况调查，了解我国大城市T2DM门诊患者的综合管理现状。结果发现：9例（0.6％）患者目前采用单纯生活方式干预策略控制血糖；1005例（66.1％）目前采用单纯服用口服降糖药治疗方案控制血糖，其中635例（41.7％）应用2种或2种以上口服降糖药治疗；506例（33.3％）患者目前应用胰岛素控制血糖。随着病程延长，患者胰岛素治疗比例逐渐提高（χ²趋势=132.670，P=0.000）。自我管理方面：1207（79.3％）例患者报告了解糖尿病相关知识；89.1％（1356）的患者报告在药物治疗的同时控制饮食；1121例（73.7％）患者报告监测空腹血糖（fasting blood glucose，FBG）或餐后血糖（postprandial blood glucose，PBG），844（55.5％）例监测频率≤4次／月（FBG和／或PBG），277例（18.2％）监测频率＞4次／月（FBG和／或PBG）；仅45.6％（691例）的患者报告检测过糖化血红蛋白（glycosylated hemoglogin Alc，HbAlc）；58.8％（890）的患者报告规律参加体育锻炼（≥1次／周）；92.9％（1400例）的患者报告遵从医嘱。对T2DM门诊患者自我管理相关因素进行分析，结果如下：在调整了年龄、性别等因素后，1)随病程延长，患者糖尿病知识知晓率逐渐提高（aOR=1.06，95％CI：1.03-1.09）；患者文化程度高（aOR=2.02，95％CI：1.74-2.35）、有T2DM家族史（aOR=1.64，95％CI：1.19-2.27）、首诊（诊断出糖尿病）医院为三级（aOR=1.54，95％CI：1.12-2.10）、医生给予健康教育（aOR=1.48，95％CI：1.04-2.10）、就诊时与医生交流（aOR=3.13，95％CI：2.31-4.23）是患者知晓糖尿病知识的促进因素；2)随年龄增加，患者饮食控制率提高（aOR=1.05，95％CI：1.03-1.07），了解糖尿病知识（aOR=2.8，95％CI：1.85-3.88）是患者饮食控制的促进因素；3)较非胰岛素治疗患者，胰岛素治疗患者更倾向于监测血糖（aOR=1.92，95％CI：1.42-2.60）。文化程度高（aOR=1.27，95％CI：1.12-1.45）、态度积极（认为促进血糖控制能减少合并症）（aOR=1.51，95％CI：1.03-2.22）、知晓糖尿病知识（aOR=2.09，95％CI：1.54-2.85）能够促进患者监测血糖；4)随年龄增加，患者规律体育锻炼率升高（aOR=1.02，95％CI：1.01-1.04），知晓糖尿病知识（aOR=1.78，95％CI：1.34-2.37）是患者规律体育锻炼的促进因素。2．2型糖尿病门诊患者血糖控制及相关因素研究。以4个城市15家医院连续纳入的1511名合格T2DM门诊病例为研究对象；通过对T2DM门诊患者进行HbAlc检测，以及社会人口学因素、管理相关因素的现况调查，了解我国大城市T2DM门诊患者的血糖控制现状以及社会人口学因素、管理相关因素和血糖控制水平的关系。1511例T2DM门诊患者的HbAlc平均值为8.11±1.65％，其中206例血糖控制理想（＜6.5％），占13.6％；408例血糖控制良好（6.5％-7.5％），占27.0％；897例血糖控制差（＞7.5％），占59.4％。随病程延长，患者血糖控制差的比例逐渐升高，控制理想和良好的比例逐渐降低（χ²趋势=101.670，P=0.000）。在调整了城市、年龄、病程、性别和是否接受胰岛素治疗后，知晓糖尿病相关知识（aOR=0.60，95％CI：0.46-0.80）、遵从医嘱（aOR=0.63，95％CI：0.40-0.98）和控制饮食均有助于患者的血糖控制（aOR=0.49，95％CI：0.34-0.72）；BMI控制差的患者血糖控制良好的可能性较BMI控制良好的患者低（aOR=1.40，95％CI：1.06-1.85）；血糖监测频率＞4次／月（aOR=0.51，95％CI：0.36-0.73）和血糖监测频率≤4次／月（aOR=0.66，95％CI：0.50-0.87）的患者血糖控制良好的可能性均高于不监测血糖的患者；检测HbAlc（≥3次／年；0-3次／年）有助于患者的血糖控制（aOR=0.33，95％CI：0.23-0.48；aOR=0.57，95％CI：0.43-0.74）。3．2型糖尿病门诊患者合并症患病率研究。以4个城市15家医院连续纳入的1524名合格T2DM门诊患者为研究对象；通过对T2DM门诊患者合并症患病情况以及社会人口学因素的现况调查，了解我国大城市T2DM门诊患者主要慢性合并症的患病率、合并症患病率的地区和人口学分布并检测合并症患病率与年龄和病程的关系。1524例患者中，792（52.0％）例患者合并1个或多个合并症；心血管系统、脑血管系统、周围神经病变、眼部病变、肾脏病变、足部病变患病率分别为：30.1％、14.8％、17.8％、10.7％、10.7％、0.8％。女性合并症总患病率高于男性（χ²=9.751，P=0.002）；4城市患者合并症患病率不同（χ²=8.763，P=0.033）；总合并症患病率(χ²_trend=91.899，P=0.000)和各系统合并症患病率（P值均＜0.05）均随年龄增加而显著增加。调整年龄后，总合并症患病率（χ²=106.290，P=0.000）和各系统合并症患病率均（P值均＜0.05）随病程延长逐渐增加。4．2型糖尿病门诊患者治疗满意度研究。本研究首次对糖尿病治疗满意度问卷当前版（diabetes treatment satisfaction questionnaire status version，DTSQs）在我国T2DM患者中应用的信用和效度进行初步评价。以上海市3家医院和广州市4家医院749例门诊T2DM病例为研究对象，采用自行翻译的中文版DTSQs进行问卷调查、同时进行实验室检测收集资料，应用信度分析、因子分析结合结构方程模型、多元协方差分析分别评价DTSQs的信度、效度和反应度。结果发现，DTSQs满意度维度分半信度系数为0.807，Cronbachα系数为0.717；探索性因子分析结果显示，3个公因子累计方差贡献率为67.656％；拟合度检验结果显示，相对χ²=4.95，DK=0.18；不同性别、不同血糖控制水平组患者DTSQs3个维度得分向量不同。研究数据支持DTSQs的理论假设，中文版DTSQs具有较高的信度和结构效度。结果初步表明DTSQs中文版可以在中国T2DM患者中应用。应用DTSQs测评1524例T2DM门诊患者的治疗满意度，结果表明患者治疗满意度平均得分23.49±6.33；高血糖和低血糖条目得分中位数分别为3（P25-P75：1-4）和1（P25-P75：0-2）。多因素分析结果显示：男性患者（P=0.024）、HbAlc控制较好患者（P=0.000）、与医生交流患者（P=0.000）治疗满意度得分较高，满意度得分与患者医嘱遵从率正相关（aOR=1.08，95％CI：1.05-1.12）。5．2型糖尿病门诊患者生存质量研究。以4个城市15家医院连续纳入的1524例T2DM门诊病例为研究对象，应用SF-36简明量表测量T2DM门诊患者的生存质量。结果发现，研究对象生理职能维度，社会功能维度平均得分高于80分；总体健康维度平均得分低于50分；其余维度得分在60-80分之间；大部分维度得分低于国内和国际常模，且躯体相关维度（physical component scales，PCS）、精神相关维度（mental component scales，MCS）总分均小于50，初步表明T2DM门诊患者生存质量低于常模人群。有合并症患者生存质量各维度得分均低于无合并症患者（得分低4.68-16.06），合并症对患者生存质量各维度得分影响程度不同，对生理职能（降低22.45％）和情感职能（降低16.28％）影响较大，对生理机能（降低5.61％）和精神健康（降低6.65％）影响相对较小。随合并症数量增加，患者生存质量逐渐降低，患者每增加1个合并症，其生存质量各维度得分平均降低2.82-10.33分。知晓糖尿病相关知识的患者生存质量各维度得分及PCS和MCS得分分别较不知晓糖尿病相关知识的患者高1.03-11.26％；控制饮食组患者较不控制饮食患者精力、精神健康维度得分和MCS得分高2.22-3.06％；规律运动的患者生理机能、生理职能、总体健康、精力、社会功能、情感职能维度得分及PCS和MCS得分分别较不规律运动患者高4.11-20.05％；监测血糖的患者精力和精神健康维度得分显著较不监测血糖的患者高5.03％和4.27％，生理机能维度得分显著较不监测血糖的患者高2.86％。治疗满意度得分与生存质量各维度得分存存显著正相关关系（相关系数：0.091-0.289），与高血糖得分（相关系数：0.056-0.206）、低血糖得分（相关系数：0.060-0.122）存在显著负相关关系。6．2型糖尿病患者卫生服务利用和疾病经济负担研究。以4个城市20家医院连续纳入的1524例T2DM门诊病例和516例住院病例为研究对象；通过对2040例患者本次就诊／住院费用，最近1个月、最近6个月就诊次数以及就诊费用，最近1年住院次数以及住院费用的调查；了解我国大城市T2DM患者卫生服务利用和疾病经济负担现状。门诊患者最近1个月半均就诊1.15±1.34次，住院患者本次入院前1个月平均就诊1.42±1.97次；门诊患者最近6个月平均就诊7.33±7.21次，住院患者入院前6个月平均就诊8.52±11.81次。门诊患者最近1年平均住院0.19±0.48次，住院患者最近1年平均住院0.56±1.13次。有合并症患者最近1个月就诊次数（Z=-3.578，P=-0.000）、最近6个月就诊次数（Z=-2.472，P=0.000）、最近1年住院次数（Z=-9.296，P=0.000）均高于无合并症患者。门诊患者年人均总费用中位数为5992元，均数为9963元；住院患者年人均总费用中位数为8355元，均数为23231元。所有研究对象（门诊和住院患者）总费用中位数为6110元，均数为12523元；其中直接医疗费用占77.7％，直接非医疗费用占17.2％，间接费用占5.1％。有合并症患者年人均总费用（Z=-9.270，P=0.000），年人均直接医疗费用（Z=-8.317，P=0.000）均高于无合并症患者。随合并症数量的增多，患者年人均直接医疗费用和年人均总费用逐渐增加。更多还原

【Abstract】 The number of diabetic patients increased rapidly in China, from 4.8 million（1980） to 23.8 million（2003）, and to 39.8 million（2007）. In 2025, it was predicted tobe 46 million. Type 2 diabetes constitutes about 90 to 95 percents of all diabetes.China carries the next highest diabetes burden in the world after India having thehighest number of diabetic patients. Type 2 diabetes is highly prevalent in the agedpopulation. Since more than 100 million Chinese are at or after age of 65 years andthe number increases continuously, there will be a rapid growth of old diabeticpatients. Higher prevalence of diabetes mellitus are usually found in big cities and theprevalence increases much more significantly in big cities than in small cities or ruralareas.Since type 2 diabetes is incurable, the main roles of diabetic management are toreach aggressive control hyperglycemia and to prevent or delay the occurrence ofchronic complications. With the development of medicine, the treatment regimen andmanagement algorithm become more important and were focused on by healthcarecare systems. While, it is not neglected that type 2 diabetes exerts a considerable tollon health resources of the country. The chronicity of the disease and diabeticcomplications also place a heavy burden on diabetic patients, their families and thesociety. Equally important, the quality of life of diabetic patients needs much moreemphasis due to the pursuing of fulfillment of well-being.The management of more than 30 million diabetic patients has become a toughsocial and medical problem to our country. And the management of DM should beevidenced by results from well-designed, strict-implemented, multidisciplinary,big-sampled epidemiologic surveys and clinical trials. While, there are not manyevidences at hand now in China. To provide clues and give evidences to managementand control of type 2 diabetes, we carried out this multi-center and cross-sectionalstudy in four major Chinese cities representing the east, west, north and south ofChina’s mainland: Shanghai, Chengdu, Beijing and Guangzhou, respectively. The study subjects were 1524 outpatients and 516 inpatients from clinics or wards of a total of20 hospitals, using a two-phase subject enrolment process, by face-to-face interviewwith a unique questionnaire. Using multidisciplinary methods（epidemiology,endocrinology, sociology, and health economics）, this study is to know themanagement status and blood glucose control, to estimate the prevalence of chroniccomplication, to assess the treatment satisfaction and quality of life of T2DM patients,and to examine the economic burden of type 2 diabetes mellitus with or withoutcomplications.Specifically,1. A cross-sectional study was conducted to examine the management status oftype 2 diabetic outpatients. The study subjects were 1524 outpatients continuouslyrecruited from 15 hospitals in 4 major cities of China. The subjects were interviewedface-to-face by trained interviewers using a questionnaire capturing information ontreatment regimens, self-management and related factors. Among the 1524 outpatients,9（0.6%） took the first line therapy of lifestyle intervention, 1005（66.1%） took oralhyperglycemic agents and 635 of the 1005 took two or more agents combinationtherapy, and 506（33.3%） had insulin injected. The proportion of outpatients acceptinginsulin therapy increased with prolong of diabetic duration(χ²_trend=132.670, P=0.000).1207（79.3%） outpatients reported that they were knowledgeable about type 2 diabetesand diabetic management. 1356（89.1%） outpatients complied with diet control.1121（73.7%） outpatients monitored FBG or PBG, 844（55.5%） had their blood glucoseexamined no more than 4 times per month and 277（18.2%） more than 4 times permonth. Only 691（45.6%） outpatients once examined their HbA1c and 890（58.8%） didno less than 1 times exercises per week. 1400（92.9%） complied with doctor’s adviceon medication.We examined the influential factors of self-management after adjustment ofpotential confounding factors. The results showed that duration of diabetes（aOR=1.06,95%CI:1.03-1.09）; education level（aOR=2.02, 95%CI:1.74-2.35）; family history ofT2DM（aOR=I.64, 95%CI:1.19-2.27）; diagnosed in tertiary hospital（aOR=1.54,95%CI: 1.12-2.10）; educated by doctors（aOR=1.48, 95%CI: 1.04-2.10）; communicatedwith doctors（aOR=3.13, 95%CI:2.31-4.23） were associated significantly with masterof knowledge about type 2 diabetes and diabetic management. With respect to dietcontrol and regular exercises, those ageing and self-knowing diabetes or diabeticmanagement were more likely to control their diet（aOR=1.05, 95%CI:1.03-1.07; aOR=2.8, 95%CI:1.85-3.88, respectively） or did regular exercises（aOR=1.02,95%CI:1.01-1.04; aOR=1.78, 95%CI:1.34-2.37, respectively）. When it comes toblood monitoring, patients undergoing insulin therapy（aOR=1.92, 95%CI:1.42-2.60）,with higher education （aOR=1.27, 95%CI:1.12-1.45）, those believe that better glucosecontrol can prevent chronic complication from occurring （aOR=1.51,95%CI:1.03-2.22）, those have the knowledge of diabetes and diabeticmanagement（aOR=2.09, 95%CI:1.54-2.85） were more preferable to monitor theirblood glucose.2. To explore the influences of patients’ self-management to glycaemic controlamong type 2 diabetic mellitus（T2DM） outpatients in urban China, a cross-sectionalstudy was carried out in 1524 T2DM outpatients from 15 hospitals in 4 major cities ofChina. Questionnaire interview was used by trained surveyors to collect data ongeneral characters and self-management. HbA1c test was applied to measure bloodglucose in the centralized hospital in each city. Logistic regression was used toexamine the association between self-management components including diet control,knowledge, blood monitoring and regular exercises and level of HbA1c. A total of1511 T2DM outpatients completed the HbA1c test and mean HbA1c was 8.11±1.65%.Among the 1511 outpatients, 206（13.6%） had HbA1c＜6.5%, 408（27.0%） between6.5%-7.5%, 897（59.4%）＞7.5%. Diabetic duration was positively related with poorHbA1c control(χ²_trend=101.670, P=0.000). Multivariate analysis showed that patientswho were under diet control（OR=0.57, 95%CI: 0.38-0.83）, knowledgeable onDM（OR=0.74, 95%CI: 0.56-0.99）, compliance behavior（OR=0.63,95%CI:0.40-0.98）, having regular monitoring on blood glucose（≤4 times/month:OR=0.58, 95%CI: 0.40-0.83;＞4times/month: OR=0.68, 95%CI: 0.51-0.91） andexamining HbA1c（≥3 times per year: OR=0.33, 95%CI:0.23-0.48; 0-3 times per year:OR=0.57, 95%CI:0.43-0.74） were more likely to have a better glycaemic controlindicated by HbA1c.3. To determine the prevalence of the main chronic complications among urbanChinese T2DM outpatients, to examine the distribution of chronic complicationsamong different demographics strata, and to describe the relationship betweencomplications and age, as well duration since diagnosis, this cross-sectionalhospital-based study was carried out in 4 major Chinese cities of China. Of the 1,524study participants, 637（41.8%） were male, and the mean age of the subjects was63.3±10.2 years. At least one chronic complication was diagnosed in 792 individuals （52.0%） of the study participants; 509（33.4%） presented with macrovascularcomplications and 528（34.7%） with microvascular complications. The prevalence ofcardiovascular and cerebrovascular conditions, neuropathy, ocular lesions,nephropathy and foot disease were 30.1%, 14.8%, 17.8%, 10.7%, 10.7% and 0.8%,respectively. The prevalence of chronic complications varied between cities, andsignificantly increased with age and duration of diagnosed diabetes. The overallprevalence of complications among female patients was significantly higher than inmale patients（χ²=9.75, P=0.002）. The prevalence of complications also variedbetween patients from different regions of China（χ²=8.763, P=0.033）. Both theoverall prevalence of complications(χ²_trend=91.90, P=0.000) and the prevalence of allconsidered conditions increased with age（all P-values＜0.05）. After adjusting for age,the overall prevalence of complications significantly increased with disease duration（χ²=106.290, P=0.000） and all considered conditions increased with diabetic duration（all P-values＜0.05）.4. First, to evaluate the reliability, validity and responsiveness of diabetestreatment satisfaction questionnaire status version（DTSQs） among Chinese type 2diabetic mellitus（T2DM） outpatients, a cross-sectional study was carried out in 749T2DM outpatients from seven hospitals in Guangzhou and Shanghai, China. Thediabetes treatment satisfaction questionnaire status version was self-filled by patients,with helps from interviewers for those having difficulties in reading. Split-halfreliability correlation coefficient, Cronbachαand Spearman’ rho were used to test thereliability. Exploratory factor analysis（EFA） and confirmatory factor analysis（CFA）were employed separately to examine the construct validity of the scale. Theresponsiveness of the scale was appraised by the multiple analysis of covariance（MANCOVA）. The split-half correlation coefficient was 0.807 and the Cronbach awas 0.717 of the six items of treatment satisfaction. The three factors extracted byEFA explained 67.66% of the overall variance. The relativeχ² was 4.95 and DK was0.18 in goodness of test by CFA. The vector of treatment satisfaction, perceivedhyperglycemia and hypoglycemia were different in males versus females, in poorversus good blood glucose control group. The DTSQs showed good reliability andvalidity among T2DM outpatients and was applicable to Chinese T2DM patients.Then, DTSQ was used to evaluate the treatment satisfaction of the 1524 diabeticoutpatients. The mean score for treatment satisfaction was 23.49±6.33, the medianfor perceived hyperglycemia and hypoglycemia were 3（P25-P75:1-4） and 1（P25-P75:0-2） respectively. Multivariate analysis indicated that the mean score fortreatment satisfaction showed an increase in males（P=0.024）, in outpatients havingHbA1c＜7.5%（P=0.000）, in outpatients communicating with doctors（P=0.000）; andoutpatients with higher treatment satisfaction（aOR=1.08, 95%CI=1.05-1.12） weremore likely to comply with doctor’s advice.5. To evaluate the quality of life of diabetic outpatients and to explore the impactof chronic complications of T2DM on quality of life among T2DM outpatients inurban China, a cross-sectional study was carried out in 1524 T2DM outpatientsrecruited from 15 hospitals in 4 major cities of China. Questionnaire interviews wereused to collect data on general characters and complications of T2DM. SF-36questionnaire（version 1） was used and self-completed by patients under the help frominterviewers for those having difficulties in reading.The mean scores of role-physical and Social Functioning subscale were higherthan 80, of general health subscale were less than 50, of other subscales were in 60-80.The scores of most subscales were lower than the norm scores in China or America,and the scores of PCS and MCS were less than 50. The results indicated that thequality of life was lower in diabetic outpatients than in norm population.Complication was a significant predictor leading to poorer SF-36 subscale scores（decreased by 4.68-16.06 scores） and two summary scores（PCS, MCS）. The scores ofrole-physical subscale（decreased by 22.45%） and role-emotional subscale（decreasedby 16.28%） decreased much more than the other subscales in patients withcomplication. Overall, QOL was reduced by 11.02% in patients with complication.The QOL had a gradual decrease with the increasing number of complications. Theaverage scores descended from 2.82 to 10.33 in the eight subscales with the increaseof one complication. Knowing diabetic knowledge increased SF-36 subscale scoresand two summary scores（PCS, MCS） by 1.03-11.26%. Diet control increased scoresof vitality subscale, mental health subscale and MCS by 2.22-3.06%. Regularexercises increased scores of physical functioning, role-physical, general health,vitality, social functioning, role-emotional subscales and PCS, MCS by 4.11-20.05%.Self-monitoring blood glucose increased scores of vitality, mental health androle-physical subscales by 5.03%, 4.27% and 2.86% respectively. The eight subscalescores of SF-36 were positively related with treatment satisfaction scores（coefficientsbeing 0.091-0.289 respectively）; negatively related with perceived hyperglycemicscores（coefficients being 0.056-0.206, respectively） and hypoglycemic scores （coefficients being 0.060-0.122, respectively）.6. To examine the health utilization of diabetic patients and economic burden ofT2DM, a cross-sectional study was carried out in 1524 T2DM outpatients and 516T2DM inpatients recruited from 20 hospitals in 4 major cities of China. Questionnaireinterviews were used to collect data on general characters, health seeking behaviorsand cost of illness during the past 1 month, 6 months. The mean times of visitingoutpatient department（OPD） during the past 1 month were 1.15±1.34 in outpatientsand 1.42±1.97 in inpatients. And during the past 6 months, the average times ofvisiting OPD were 7.33±7.21 in outpatients and 8.52±11.81 in inpatients respectively.The mean hospitalization times during the past 1 year were 0.19±0.48 in outpatientsand 0.56±1.13 in inpatients. Chronic complication significantly led to increasedutilization of OPD services in the past 1（Z=-3.578, P=0.000）, 6（Z=-2.472, P=0.000）months and to increased utilization of hospitalization services（Z=-9.296, P=0.000） inthe past 1 year. The annual total cost was estimated to be 5992 CNY in median（9963CNY in mean） per outpatient and 8355 CNY in median（23231 CNY in mean） perinpatient. The annual total cost per patient was estimated to be 6110 CNY in median（12523 CNY in mean） with direct medical cost accounting for 77.7%, directnon-medical cost 17.2% and indirect cost 5.1%. The annual total costs per patient（Z=-9.270, P=0.000）、direct medical costs per patient（Z=-8.317, P=0.000） were bothdifferent between patients with and without complications. The annual total costs anddirect medical costs increased with number increasing of complications.更多还原