With the comprehensive implementation of the family doctor system in China, community health service resources have been widely utilized. However, there is a lack of research on the utilization tendency of primary outpatient medical resources for adult diabetic patients using objective medical visit data.

To understand the characteristics of outpatient visits for diabetic patients in the community, analyze the factors influencing frequent visits by patients enrolled in family doctor services, correctly identify the utilization tendency of primary outpatient medical resources by enrolled patients, and provide a basis for promoting the high-quality development of family doctor contract services.

In August 2023, outpatient visit records of adult diabetic patients from 13 community health service centers in Minhang District, Shanghai, for the year 2021 were retrieved. Personal information such as names and contact details were forcibly anonymized, while data on sex, age, enrollment in family doctor services, triggered diagnosis records, frequency of medical treatment, department of medical treatment, and Chinese medicine services during visits were collected. The characteristics of outpatient visits for adult diabetic patients in the community were analyzed, and multiple logistic regression and decision tree models were used to analyze the factors influencing frequent visits by enrolled patients.

There were 66 838 adult patients, resulting in 1 281 972 outpatient records. Among them, 48.96% (32 723) were male; the median age of the patients who visited the hospital was 71 (66, 77) years old; the median number of diagnosis types was 6 (3, 10) ; the median frequency of visits was 15 (9, 26), 10.28% (6 871) of the adult patients frequently visited the hospital for 362 068 times; 1-2 departments were the main ones (60.38%), and 9.33% (6 233) of the patients with≥5 departments involved; 39.53% (26 423) visited Chinese medicine services. During the year, Contracted patients accounted for 92.77% (62 005) of all the patients, while non-contracted patients only accounted for 7.23% (4 833). There were significant differences between contracted patients and non-contracted patients in sex, age, complexity of illness, frequency of medical treatment, department of medical treatment, Chinese medicine service (P<0.05). The results of multivariate Logistic regression analysis showed that sex, age, complexity of illness, department of medical treatment, Chinese medicine service were the influencing factors of frequent visits by contracted patients (P<0.05). The results of decision tree model analysis showed that the complexity of the disease was the most important factor affecting the frequent visits of contracted patients.

The signing rate of adult diabetic patients in community primary clinics in Minhang District of Shanghai was high, the effect of family doctor signing system was remarkable, the aging of adult diabetic patients in community primary clinics was prominent, the contracted patients' condition was more complicated, and the demand for Chinese medicine treatment was stronger. The frequent visits of contracted patients were affected by many factors, especially the complexity of their condition. Therefore, it is suggested that relevant departments identify frequent patients efficiently according to the patients' condition, so as to make more rational use of medical resources in primary clinics.

随着我国家庭医生制度构建工作的全面开展，社区卫生服务资源被广泛利用，然而较少有使用客观就诊数据对成年糖尿病患者基层门诊医疗资源的利用倾向进行研究。

了解糖尿病患者的社区门诊就诊特征，分析签约家庭医生服务患者频繁就诊的影响因素，正确识别签约患者对社区门诊医疗资源的利用倾向，为推进家庭医生签约服务高质量发展提供依据。

于2023年8月调取2021年上海市闵行区13家社区卫生服务中心的成年糖尿病患者门诊就诊记录，强制脱敏姓名、联系方式等信息，收集性别、年龄、签约家庭医生服务情况、触发的诊断记录、就诊频次、就诊科室数、就诊时接受中医类服务等信息。分析成年糖尿病患者社区门诊就诊特征，采用多因素Logistic回归和决策树模型分析签约患者频繁就诊的影响因素。

共获取符合研究要求的门诊记录1 281 972条，涉及糖尿病患者66 838例。其中，男32 723例（48.96%）；中位年龄为71（66，77）岁；中位诊断种数为6（3，10）种；中位就诊频次为15（9，26）次，10.28%（6 871例）频繁就诊的成年糖尿病患者的就诊频次为362 068次；就诊科室数以1~2个为主[40 354例（60.38%）]，就诊科室数≥5个者6 233例（9.33%）；就诊时接受中医类服务者26 423例（39.53%）。签约患者62 005例（92.77%），非签约患者4 833例（7.23%），签约患者与非签约患者的性别、年龄分组、病情复杂程度、频繁就诊情况、就诊科室数、就诊时接受中医类服务情况比较，差异有统计学意义（P<0.05）。多因素Logistic回归分析结果显示：性别、年龄分组、病情复杂程度、就诊科室数、就诊时接受中医类服务情况是签约患者频繁就诊的影响因素（P<0.05）。决策树分析结果显示：病情复杂程度是影响签约患者频繁就诊的最主要因素。

上海市闵行区社区门诊就诊成年糖尿病患者签约率较高，家庭医生签约制度效果明显。社区门诊就诊糖尿病患者老龄化凸显，签约患者病情更复杂、中医就诊需求更强烈。签约患者频繁就诊受多种因素影响，尤其是病情复杂程度，建议相关部门根据患者病情高效识别频繁就诊者，从而更合理利用社区门诊医疗资源。