郑建东,郑庆斯.Monte Carlo方法在碘盐监测抽样方法中的应用[J].Chinese journal of Epidemiology,2002,23(4):262-264 |
Monte Carlo方法在碘盐监测抽样方法中的应用 |
Application of Monte Carlo sampling method in iodized salt monitoring |
Received:February 25, 2002 Revised:June 05, 2012 |
DOI: |
KeyWord: 碘缺乏病 Monte Carlo方法 碘盐监测 |
English Key Word: Iodine deficiency disorders Monte Carlo method Iodized salt monitoring |
FundProject:联合国儿童基金会资助项目 (O135YH10112) |
|
Hits: 3365 |
Download times: 1196 |
Abstract: |
目的 探索既能反映省级水平的碘缺乏病状况 ,又能发现地区级水平的非碘盐问题地区的新监测方案。方法 对甘肃省与福建省碘缺乏病综合干预试点研究的基线调查结果 ,运用MonteCarlo方法进行模拟抽样试验 ,分县级与地级两个层次进行。结果 在县级模拟中 ,一个县随机抽取 7至 8个村或 4至 5个乡 ,可较为准确地判定出全省的非碘盐问题县 ,真阳性率在 80 %左右 ,而对于非问题县的误判率较低 ,假阳性率在 20 %左右。在地级模拟中 ,每个地区选取 3个县 ,可对全省的非碘盐问题地区较为准确地判定 ,真阳性率达到 100 %。结论 建议全国碘缺乏病流行病学调查中碘盐的抽样方案覆盖各省所有地区 ,每个地区随机抽取 3个县 ,每个县随机抽取 4个乡 ,每乡随机抽取 2个村 ,每村随机采取 10户盐样 ,所获得资料经人口加权计算省级水平的碘盐覆盖率。 |
English Abstract: |
Objective To explore a sampling method which could reflect iodine deficiency disorders (IDD) status at provincial level and discover risk areas with non iodized salt problem. Method Baseline data of Iodized salt from Gansu and Fujian provinces were analyzed with Monte Carlo method both at county and prefecture levels respectively. True positive rate and false positive rate were also calculated. Results With data from 7 8 villages or 4 5 townships counties at risk could be discovered. The true positive rate was around 80% and false positive rate was around 20%. At prefecture level, when randomly selecting and checking 3 counties,the samples would satisfy the discovery of all the risk areas with non iodized salt problem. Conclusions We suggested that the sampling method of iodized salt investigation in national IDD surveillance as follows: to randomly choose 3 counties at each prefecture, 4 townships at each county, 2 villages at each township and 10 salt samples by household survey. The coverage rate of iodized salt in a province could be calculated by post weighted method with population number. |
View Fulltext
Html FullText
View/Add Comment Download reader |
Close |
|
|
|