儿童龋病风险评估模型构建及儿童龋病预防策略研究

Dental caries is one of the chronic diseases that affect seriously children’s overall health, dental health and life quality in China. Incidence of early dental caries was 71.9% but caries filling rate was only 4.1% in 5-year children group. The dental caries incidence rate was constantly increased in recent decade, which indicated the faultiness and inefficiency of dental preventive strategy. New path of dental caries controlling strategy should be explored immediately. We found there was no caries risk assessment model constructed according to our domestic child patient characteristic in previous studies reviewing, therefore there was no preventive dental strategy on individual level based on the caries risk assessment. Study deeply on this topic may lead us to expore new path for preventive dental care for children. In this study, caries risk assessment model would be constructed based on the result of children’s dental caries risk factors research. Logistic Regression Model and Cox Regression Model would be used seperately under a retrospective cohort study design. Optimal model would be chosen on the consideration of clinical opinion as well. This model would be tested in community and made more compatible with our nation’s situation. Children’s dental caries preventive strategy would be developed on Social Marketing Theory. The preventive dental strategy would be used and moniterd in one year in target population by an app. This app would be developing by our research group. Effectiveness evaluation of dental care strategy would be tested by Difference-in-differences Method. The result of this study would enlighten a new thought for preventive dental care police optimization and improvement.

龋病是严重影响我国儿童口腔健康、全身健康和生命质量的慢性病之一。目前我国5岁组儿童龋病发病率高达71.9%，而充填率仅为4.1%。现行防控策略不完善导致儿童龋病发病率持续上升，亟需发现新途径以实现综合防控。课题组在前期文献梳理和口腔公共卫生策略研究的基础上发现，并无符合我国国情特征性的龋风险评估模型，故也无基于评估的个体龋病防控手段，对此问题进行深入研究将有助于发现儿童龋病防控新途径。本课题拟对目标儿童群体开展回顾性队列研究，全面收集儿童患龋影响因素，分别用Logistic回归和Cox回归模型建模，结合临床实际择优构建符合我国国情的儿童龋风险评估模型并进行群体验证。运用社会营销理论开发针对不同龋风险患儿的干预策略，以手机移动端程序的创新方式对儿童家长进行一年干预及监测；最后利用双重差分模型对干预措施的效果进行评价确认该干预方式的有效性。本研究结果将为优化和完善我国儿童龋病防控提供新的思路。

龋病在我国俗称蛀牙，是牙齿在以细菌为主的多种因素影响下，无机物脱矿、有机物分解，牙体硬组织缺损的慢性、进行性、破坏性疾病，是目前世界范围和我国患病率最高的口腔感染性疾病。我国低龄儿童口腔龋病发病率高但就诊率极低，未治疗龋病已经严重影响我国儿童身体素质。由于儿童生命早期的口腔健康及口腔卫生行为习惯，会对其成年后的全身健康以及生活质量产生重要影响。要达成健康中国提出的龋病防控目标，关键是开展低龄儿童口腔健康管理，这也是对青少年组、成人组的龋病防控的基础手段。由于在儿童群体中普遍存在龋病风险差异，故通过龋病风险评估来识别高龋风险人群是开展精准预防的基础。.本研究内容包括4个部分，（1）探索低龄儿童龋（early childhood caries, ECC）发病的危险因素和保护因素；（2）构建ECC预测模型；（3）根据龋风险评估结果开发干预策略并进行现场社区干预；（4）对社区干预效果进行评价。.研究结果显示，影响我国成都地区儿童ECC患病率的主要因素有儿童年龄、身高、体重、家人蛀牙情况、刷牙时是否是用含氟牙膏等12个因素。采用机器学习中的随机森林模型对数据进行拟合后，ECC风险预测模型的ROC指数可达到0.8832，精准度达到0.8795。本研究进一步在成都市3所社区卫生服务中心和3所幼儿园进行了为期一年的ECC风险管理干预对照实验。对干预组开展现场宣教、一对一行为指导、微信群等综合干预方式干预，对照组仅采用现场集体宣教等方式。双重差分法结果显示干预组的龋失补指数情况比对照组降低了9.8%（P<0.05），说明我们的干预对于预防新发龋坏是有效果的。.本研究开发了国内第一个基于非生物学因素的低龄儿童龋风险评估模型，该模型为基层医疗机构中开展儿童龋风险管理提供了基础工具。由于该模型不包括生物学指标，数据获取相对简便易行，更适合推广使用。同时该模型包括了行为学因素，可以发现儿童生命早期中的家长行为，有利于有的放矢的开展儿童口腔健康管理，尤其是降低缺乏口腔医疗资源地区的低龄儿童患龋发病率。