Abstract: Background　To improve the capabilities of general practitioners（GPs） working in primary care，one key approach is targeting patients' needs and problems，but there is no big data analysis of the gap between patients' needs and provision of services using the real-world data in China. Objective　To examine the capabilities of urban community health centers（CHCs） for outpatient service provision using a big data analysis of the real-world diagnostic data，offering a reference for improving GP's capabilities for providing first-contact services precisely targeting patients' needs. Methods　We collected outpatient diagnoses（n=2 897 700） by ICD-10 code from CHCs of Wuhou District in 2020，and analyzed the structure，diagnoses ranked from most to least，and degree of accuracy in diagnosing 4 types of chronic diseases using data structure analysis and data distribution analysis. Results　Of the diagnoses，99.31%（2 877 700/2 897 700）were assigned 912 types of six-digit diagnostic codes with a frequency greater than 30；87.31%（2 392 603/2 740 333） with assigned three-digit diagnostic codes were concentrated in the top 10. Among them，preventive diagnostic records are more concentrated，mainly for physical examination（including child health examination） and vaccination services. Most（11/13） CHCs' diagnostic and treatment capabilities had reached grade A standard defined in the Service Capacity Assessment for Community Health Centers（hereinafter referred to as SCACHCs），but the rank order of disease diagnoses was different from the diagnostic coding catalogue in the SCACHCs. The mean coefficients of variation of ICD type，common type，preventive type or clinical type across CHCs were all greater than 25%. The top 10 diagnoses with assigned three-digit diagnostic codes were highly consistent with the national essential public health services. Other common diagnoses included sleep disorders，acute respiratory diseases，digestive diseases and insomnia. Patients with chronic diseases such as hypertension，cardiovascular and cerebrovascular diseases，type 2 diabetes，（lipid） metabolic disorders，and sleep disorders were more likely to have higher number of CHCs visits（more than 5 times per year），and physical examination ranked last among the top 10 services ranked in terms of average number of annual use frequencies per capita. The majority of diagnostic records of common chronic diseases（more than 90%） were concentrated in the code bar of "not specifically specified" in this type of diagnosis. The number of CHCs visits for COPD（n=4 268） was lower than that of those for hypertension，type 2 diabetes or hyperlipidemia. Meanwhile，the number of COPD patients（n=1 675） encountered by the CHCs was less than that of patients with hypertension，type 2 diabetes and hyperlipidemia. Conclusion　Outpatient diagnoses in CHCs of Wuhou District were concentrated，especially in the field of preventive health care. Common diagnoses for diseases were highly consisted with the service profile and population of the national essential public health projects. Most of the CHCs' diagnostic and treatment capabilities had reached the standard of excellence，but there is still room for improvement in comprehensive diagnosis of patients，especially those with chronic diseases，with highlights on taking actions to improve capacities for comprehensive diagnosis of hypertension and related cardiovascular diseases，type 2 diabetes，acute respiratory diseases，digestive diseases，organic sleep disorders，non-organic sleep disorders or other diseases，and for management of diagnosis and treatment of such diseases. Heavy workload may result in the prevalence of diagnostic inertia in GPs，especially under the circumstances of encountering common chronic diseases. So it is suggested to take factors associated with GPs，family doctor teams，and institutional supply and security into account when systematically designing the path to improve CHCs' capabilities for outpatient service provision.

Key words: Community health services, Ambulatory care, Ambulatory care information systems, Real world research

摘要： 背景　以患者需求和问题为导向提升基层全科医生的接诊能力是一个重要的在岗能力建设方向，但目前国内尚无针对基层实际接诊环境下的患者需求和服务供给差距的大数据分析。目的　通过城市社区卫生服务机构门诊诊断的大数据分析，真实反映一个典型的城市区域的社区卫生服务机构门诊接诊能力特征，从而精准识别基层首诊能力需求，为提升基层首诊能力提供参考。方法　按照疾病和健康问题的国际统计分类（第10次修订本）（ICD-10）规则对武侯区社区卫生服务机构2020年门诊产生的2 897 689条真实世界ICD-10诊断数据进行分类汇总，分析其呈现的诊断结构、顺位和4种慢性病（高血压、2型糖尿病、脂代谢紊乱、慢性阻塞性肺疾病）的评估诊断深度。采用数据结构、分布等描述方法进行分析。结果　99.31%（2 877 653/2 897 689）的诊断记录集中在频次大于30的912种诊断编码（六位编码）中，87.31%（2 392 603/2 740 333）的诊断编码（三位编码）记录集中在前10顺位。其中预防类诊断记录相对更为集中，主要为体检（含儿童保健体检）和预防接种服务。13家社区卫生服务机构中，11家已达到“国家优质服务基层行”中诊疗能力A级标准，但疾病顺位与“国家优质服务基层行”能力标准中的诊断编码目录有所差异。机构间平均ICD-10诊断种数、常见诊断种数、预防类或临床类的种数的变异系数均大于25%。排名前10类（三位编码）的诊断与国家基本公共卫生服务范围高度一致，其他常见的诊断还包括睡眠障碍、急性呼吸道疾病、消化道疾病和失眠症等。高血压及心脑血管疾病、2型糖尿病、（脂）代谢紊乱、睡眠障碍等慢性病患者更容易有较高（>5次/年）的就诊频次，体检是频次较高前10位中人均年就诊频次最低的内容。常见慢性病的诊断记录高度（90%以上）集中于本类诊断中“未特指”类编码条，慢性阻塞性肺疾病患者的就诊人次（4 268人次）和人数（1 675人）低于高血压、2型糖尿病、高脂血症。结论　社区卫生服务机构门诊诊断聚集性明显，特别是在预防保健领域。常见诊断高度集中于国家基本公共卫生服务项目对应的服务内容和人群。大部分社区卫生服务机构的接诊能力范围已达到优秀标准，但对患者（特别是慢性病患者）进行综合的诊断性评估上还需改进，特别是高血压及相关心血管疾病、2型糖尿病、急性呼吸道疾病、消化道疾病、器质性或非器质性睡眠障碍等疾病综合诊断评估能力的建设和诊疗质量管理。全科医生在大量接诊负荷下，特别是在常见慢性病接诊上存在一定的“诊断惰性”，建议至少从全科医生、家庭医生团队、机构供给和保障三层系统地设计基层首诊能力的建设路径。

关键词: 社区卫生服务, 门诊医疗, 门诊医疗信息系统, 真实世界研究