基于捕获-再捕获方法的海南省2015 - 2017年死因监测漏报情况调查

目的 进一步了解海南省死因监测数据的真实情况，为科学制定公共卫生策略和措施提供依据。方法 采用捕获 - 再捕获方法开展调查，共分为三步:第一步，在海南省22个市、县（区）各选取满足以下条件的3个乡镇（街道）:人口数量和社会经济水平接近本县(区)各乡镇(街道)的平均水平，从3个乡镇（街道）中随机选择一个乡镇（街道）作为调查点，点内所有住户作为调查对象，在《人口死亡信息登记管理系统》中查询调查点死亡时间在2015 - 2017年的一批随机样本（“捕获”），将其做标记。第二步，通过入户调查和查阅地民政和公安部门收集常住居民死亡信息，收集死亡时间在2015 - 2017年的随机样本（“再捕获”），与“捕获”样本进行比对，发现漏报病例后，收集基本信息、死亡信息和生前诊断信息等。第三步，计算漏报率，剔除粗死亡率低于5‰的调查点。结果 共计12个乡镇的116个自然村纳入分析，共收集到死亡个案5 021例，死因监测系统查询到个案4 583例，两种方式中重复个案3 496例。全省2015 - 2017年的漏报率估计为27.44%、22.70%和24.46%，平均漏报率为24.97%（95%CI = 23.74%～26.24%）。漏报率高的前3位地区是海口市琼山区（49.21%）、龙华区（41.55%）、昌江县（39.29%）；漏报率低的3个地区为白沙县（0.56%）、五指山市（4.62%）、万宁市（9.63%）。调查点间漏报率的差异有统计学意义（χ2 = 721.25 ，P＜0.001）。年份、城市与农村、性别、年龄组间漏报率差异无统计学意义（都P＞0.05）。结论 海南省死因监测系统中漏报率较高，各市、区（县）都存在漏报现象，但漏报率高低存在地区差异。建议强化死因监测系统的培训和督查力度，将死因监测质量与绩效和奖惩机制挂钩，同时开展死因监测系统评估。

Analysis of the investigation results of missed cause of death monitoring in Hainan province from 2015 to 2017

Objective To understand the real situation of the cause of death monitoring data in Hainan province, so as to provide a basis for the scientific formulation of public health strategies and measures. Methods The capture-mark-recapture method was used to carry out the survey, which is divided into three steps. For the first step, among 22 cities and counties in Hainan province(area), select three villages and towns(street) meeting the following conditions: population and social economy level close to the county(district) the average level of each villages and towns(street), and then from three villages and towns(street), randomly select villages and towns(street) as the infested, all households within the site as investigation object, in the "population death registration management information system" to query the infested the time of death from2015 to 2017, a group of random sample( "capture"), to be marked. The second step is to collect death information of permanent residents through household investigation and consulting local civil affairs and public security departments, collect random samples( "re-capture") of the time of death from 2015 to 2017, compare them with "capture" samples, and collect basic information, death information and pre-life diagnosis information after cases are found to be unreported. The third step is to calculate the missing FRR, but the investigation points with crude mortality rate lower than 5‰ after the missing FRR are not included in the statistical analysis. Results A total of 116 natural villages in 12 townships were included in the analysis. A total of 5021 cases of death were collected, 4583 cases were found by the cause of death monitoring system, and3496 cases were repeated in the two methods. The FRR of the whole province is estimated to be 27%, 23% and 24%, with an average FRR of 25%(95%CI=24-26%) in 2015-2017. The three areas with high negative rate were Qiongshan district(49%), Longhua district(42%) and Changjiang county(39%). Three areas with low negative rate were Baisha county(0.56%),Wuzhishan city(4.6%) and Wanning city(9.6%). There was a statistically significant difference in sites(χ2=721.25, P<0.001).There was no statistical difference in missing rate among year, city and country, gender and age group(all P >0.05).Conclusion In Hainan province, the alarm failure rate is relatively high, which exists in all cities and districts(counties), but there are regional differences. It is suggested to strengthen the training and supervision of cause-of-death monitoring system,link the quality of cause-of-death monitoring with performance reward and punishment mechanism, and carry out the evaluation of cause-of-death monitoring system.