Abstract: Background　Vigorously promoting the implementation of contracted family doctor services and the construction of the hierarchical medical system are important tasks in deepening the reform of the pharmaceutical and healthcare system，which is conducive to basically realizing the goal of universal health coverage.The implementation of contracted family doctor services is in full swing in primary medical institutions.And it is of great significance to scientifically and accurately evaluate the effectiveness of policies concerning such services.Objective　To assess the effect of the implementation of contracted family doctor services on residents' healthcare-seeking behaviors.Methods　In January 2019，495 residents were selected from 10 medical institutions in five districts of Tianjin by multistage random sampling，and received a questionnaire survey on their personal basic situation，prevalence of receiving contracted family doctor services，healthcare-seeking behaviors，and utilization of referral services.Propensity score matching was used to assess the impact of receiving contracted family doctor services on healthcare seeking behaviors by comparing such behaviors between contracted residents（treatment group） and those non-contracted（control group）.Results　Multivariate Logistic regression results showed that age and diabetes were associated with receiving contracted family doctor services（P<0.05）.Compared with those non-contracted，the probability of seeking healthcare in a hospital in the local district and using referral services in those contracted increased by 8.9%，and 14.5%，respectively （P<0.05）.Moreover，the probability of seeking healthcare in a hospital in the local district increased as much as 12.6% in contracted general populations in comparison with those non-contracted（P<0.05）.However，no significant differences were found in healthcare-seeking behaviors between contracted and non-contracted priority populations （P>0.05）.Conclusion　The probability of seeking healthcare in a hospital in the district is higher in contracted residents compared with those non-contracted，which may be related to increased utilization of referral services in the implementation of contracted family doctor services.The general population is more likely to change their behaviors after receiving contracted family doctor services rather than the priority population.

Key words: Contracted family doctor services, Residents' healthcare seeking behaviors, Propensity score matching

摘要： 背景　大力推进家庭医生签约服务，着力推进分级诊疗制度建设，是深化医药卫生体制改革的重要任务，有利于实现“人人享有基本医疗卫生服务”的目标。当前，家庭医生签约服务在全国基层卫生医疗机构开展得如火如荼，对家庭医生签约服务政策效应进行科学、准确的评估具有重要意义。目的　评估家庭医生签约服务对居民就诊行为的影响。方法　2019年1月，以多阶段随机抽样的方法，在天津市5个市辖区10个医疗机构抽取495例居民作为受访者，采用问卷调查的方式对受访者的个人基本情况、家庭医生签约服务情况、就诊行为、转诊服务利用情况等进行调查。以签约居民作为处理组，未签约居民作为控制组，利用倾向得分匹配方法估计平均处理效应。结果　多因素Logistic回归分析结果显示，年龄和糖尿病患病情况对居民是否享受家庭医生签约服务有影响（P<0.05）。签约居民在辖区内医院就诊率比未签约居民高8.9%，签约居民的转诊服务利用率比未签约居民高14.5%，享受家庭医生签约服务的普通人群在辖区内医院就诊率比未签约居民高12.6%（P<0.05），而重点人群的平均处理效应比较，差异无统计学意义（P>0.05）。结论　家庭医生签约居民在辖区内医院就诊率高于未签约居民，而这可能与家庭医生签约服务提高了签约居民的转诊服务利用水平有关。相比于重点人群，普通人群更容易由于享受家庭医生签约服务而改变就诊行为。

关键词: 家庭医生签约服务, 居民就诊行为, 倾向得分匹配