目的 运用双重差分（difference-in-differences,DID）模型精准评估寄生虫病防治措施对四川省凉山州彝族欠发达农村地区儿童健康相关行为及其土源性线虫感染状况的改善情况。 方法 于2016年在四川省凉山州采用多阶段分层随机整群抽样方法选取彝族农村儿童,采用问卷调查法收集目标儿童基本特征、健康相关行为的信息,并采集3岁及以上儿童粪样,用改良加藤厚涂片法（Kato-Katz法,一粪三检）检测儿童土源性线虫感染情况。研究对象以村为单位分为干预组和对照组,在实施综合干预措施1年之后,对追踪儿童采用同样的方法收集相关信息。运用DID模型对儿童健康相关行为和土源性线虫感染情况进行分析。 结果 有效评估儿童共162名,平均年龄为7.7岁,其中干预组99名、对照组63名。DID模型结果显示,实施健康教育和驱虫治疗为主的干预措施后,干预组儿童与对照组相比,在家饭前洗手行为的净效应（DID 值 = 37.5%,P < 0.01）、在校饭前洗手行为净效应（DID值 = 24.4%,P < 0.01）、在家便后洗手行为的净效应（DID值 = 45.3%,P < 0.01）、在校便后洗手行为的净效应（DID 值 = 35.9%,P < 0.01）、刷牙洗脸行为净效应（DID 值 = 29.6%,P < 0.01）,均有统计学意义;儿童蛔虫感染率的净效应（DID值 = -37.5%,P < 0.01）具有统计学意义;其他不良行为（吃未洗蔬果、喝生水、光脚下地、玩泥巴、吮手指）和儿童总感染率、钩虫、鞭虫感染率的净效应,均不具有统计学意义（P > 0.05）。 结论 在彝族欠发达农村地区实施寄生虫病综合防治措施后,儿童洗手行为有所改善,蛔虫感染率有减少。运用DID模型,科学评估了干预措施改善儿童健康相关行为和降低土源性线虫感染率的真实效果。

Objective The difference-in-differences (DID) estimation was used to accurately evaluate the effect of anti-parasitic disease intervention on children’s health-related behaviors and soil-transmitted nematode infection. Methods Children in Yi villages in Liangshan Yi Autonomous Prefecture, Sichuan Province, were selected based on a multistage stratified random cluster sampling method for the investigation in 2016. The demographic information of the participated children and their health-related behaviors were collected by a issued questionnaire. Fecal samples were collected from participated children with age ≥ 3 and the eggs of soil-transmitted nematodes were examined using Kato-Katz method under microscope. The children were randomly divided into the intervention and control group in the same village. A comprehensive intervention measures, including health education and anthelmintic treatment, were performed in the intervention group. One year after the intervention, the same method was used to collect the relevant information from the follow-upped children. Using the DID model, the accurate effect of the intervention on children’s health-related behaviors and the control of soil-transmitted nematode infection was analyzed. Results Total 162 children were recruited in this study with an average age of 7.7 years old. They were divided into the intervention group (99) and the control group (63). After the intervention, the DID analysis results demonstrated that the sanitation and hygiene behaviors have been significantly improved. These behaviors included the hand washing before meal at home (DID = 37.5%, P < 0.05), hand washing before meal in school (DID = 24.4%, P < 0.01), hand washing after toilet at home (DID = 45.3%, P < 0.01), hand washing after toilet in school (DID = 35.9%, P < 0.01) and brushing teeth and washing face (DID = 29.6%, P < 0.01). The positive effect was also reflected by the reduced infection rate of Ascaris lumbricoides (DID = -37.5%, P < 0.01). However, other DID values for some high-risk behaviors (eating unwashed fruits and vegetables, drinking un-boiled water, walking in barefoot, playing with mud on the ground and sucking fingers) and infection rates of hookworm and Trichuris trichiura were not significantly improved (P > 0.05). Conclusion After the interventions, some hygiene behaviors, as well as A. lumbricoides infection rate of school-age children has significantly improved. The DID estimation demonstrated the positive effect of interventions on the health-related behaviors and A. lumbricoides infection in children.