2009年—2014年学龄前儿童龋齿患病率调查的Meta分析

Meta-analysis on prevalence of dental caries among preschool children in 2009-2014

目的：对学龄前儿童在不同年龄段和城乡之间的横断面进行龋齿患病率系统评价。方法计算机检索维普、万方、CNKI，检索时间从2010年—2014年，纳入所有调查不同年龄段和城乡之间的学龄前儿童龋齿患病率的文献，由2名评价者共同评价纳入研究质量，采用RevMan5.0软件对数据进行Meta分析，估计其OR值和95%CI。结果共纳入16个有关不同年龄段的儿童龋齿患病率的调查研究，包括3岁-4岁年龄组18527例，5岁-6岁组17697例；以及6个有关城乡地域差别的儿童龋齿患病率的调查研究，包括城市组8084例，乡村组5605例。按照调查来源进行分析，结果表明3岁-4岁年龄组和城市组的儿童龋齿患病率和5岁-6岁年龄组和乡村组的儿童龋齿患病率相比，差异均有统计学意义[OR合并=0.61（95% CI 0.58，0.63），OR合并=0.83（95% CI 0.77，0.89）]。结论研究结果显示，3岁-4岁年龄组和城市组的儿童龋齿患病率比5岁-6岁年龄组和乡村组的儿童龋齿患病率要低，研究结果的不一致是由于纳入研究的患者存在选择性偏倚、混杂性偏倚以及诊断性偏倚的中、高度可能性，影响了研究结果。

Objective At present,the results differs greatly from the prevalence of dental caries in preschool children between different ages and the urban and rural areas,so we Meta-analyze the prevalence of dental caries of preschool children between different ages and urban and rural areas.Methods We searthed VIP、Wangfang、CNKI （range from 2010 to 2014）. Including all the investigation on prevalence literature of dental caries in preschool children between different ages and the urban and rural areas.The quality of included case studies was evaluated independently by two reviewers. Rev Man 5.0 software was used for analyze and estimate its OR valute and 95%CI.ResultsWE identified 16 prevalence studies with the prevalence of dental caries in preschool children between different ages,including the 3-4 age group of 18527 cases,17697 cases of 5-6 age group,6 studies of the prevalence of dental caries in preschool children between the urban and rural areas,i ncluding the city group of 8084 cases,5605 cases of village group.The results of group analyses based on resource showed：statistical difference was founded between differ- ent ages and the urban and rural areas with OR valute=[0.61（95%CI 0.58,0.63）,OR合并=0.83（95%CI 0.77,0.89）].Conclusion The results show that the prevalence rate of children dental caries between 3-4 age group and urban group is lower than that of the 5-6 age group and rural groups,and the different conclusions of the 22 reports may be attributed to the high risk of selection,confounding and diagnose bias of included studies.