唐氏综合征血清学产前筛查质量管理研究

发布时间：2018-06-17 08:54

本文选题：唐氏综合征 + 血清学筛查　；参考：《浙江大学》2014年博士论文

【摘要】：孕期血清学筛查、绒毛和羊水的细胞遗传学诊断及超声诊断仍然是目前唐氏综合征二级预防的主流干预措施。高检出率、低假阳性率、高卫生经济学效益的产前筛查方案一直是产前诊断领域专家追求的目标,通过近三十年的努力也确实卓有成效。新筛查标志物的开发及新筛查方案的建立可以获得更大的筛查效能,而完善的质量管理则是为了使某一筛查方案最大化地发挥其筛查效能,因此建立适宜的产前筛查质量控制体系是高效能筛查方案更好地向临床转化的基石。在我国虽然产前筛查工作开展已经13年,规模日益增大,但质量控制基本上仍停留在临床检验质量控制最基本的室内质控与室间质评阶段。以大样本筛查数据的统计学评价为基础,对所有这些质量控制过程进行分析,进而推动质量提高,在我国还是空白。唐氏综合征产前筛查的质量控制具有特殊性,产前筛查的结果通常以胎儿罹患某一先天异常的风险率为报告形式,其质量控制的核心是唐氏综合征风险评估的可靠性。目前各筛查机构常规进行的实验室室内质控与室间质评只涉及生化标志物检测的准确性,只是产前筛查质量控制的组成部分。完善的产前筛查质量控制应涉及影响风险计算可靠性的每一个重要环节及关键因素,包括孕妇年龄、体重、孕周等基本参数,实验室测定结果的准确性,风险计算软件及参比中位数的选择等。另一方面,产前筛查在早、中孕期进行而筛查假阳性率与异常胎儿的检出率等是在胎儿出生后才能统计评价的指标,存在时间跨度;孕妇的血清学筛查,高危孕妇进一步的产前诊断及最终的分娩并非均在一个医疗机构,同时血样的采集、转运、检测、软件应用可涉及几家不同的医疗机构,亦存在地域跨度。这使得产前筛查的质量评价和管理工作更加复杂和艰巨。对于每一个环节的评估更涉及到非常专业的统计学知识,选取哪些统计学指标,如何通过统计学指标的变化及时发现筛查的质量问题,分析原因,提出整改方法也是需要解决的问题。建立产前筛查质量管理软件系统并实现网络化,可以实时高效地对各级筛查机构的筛查质量进行评估,促进产前筛查管理体系的实际推广及长期应用。目的:为了在我国也能规范化地开展产前筛查质量管理,本研究以浙江省2010年～2012年的筛查病例作为数据基础,探讨质量评估的关键性统计学指标,重点评估环节及评估流程和筛查质量改善的方法,以期最终建立起适合我国国情的产前筛查诊断质量评价及管理体系,改善产前筛查效率,提升我国产前诊断的质量。方法:收集2010～2012年浙江省17家产前筛查机构的产前筛查数据,涉及筛查孕妇数1,227,388例。通过孕妇年龄、体重、孕周等基本参数的描述性分析,使用Q-Q图、年龄校正的筛查阳性率(screening positive rate,SPR)、95%置信区间内标志物中位数的倍数中值(MedianMoM)随筛查月份、孕周及体重的分布图及累积和图(CUSUM图)对17家筛查机构进行筛查质量评估。通过以上方法对17家筛查机构的筛查质量进行评估,发现筛查质量可提升的环节,选取具有代表性的部分,如标本质量、实验室检测、风险软件及参比中位数选择等环节进一步深入分析,寻找可能的原因及提出整改优化措施。评价血清标志物中位数的选择、分析软件使用等对筛查结果的影响。结果:本研究收集了 2010年、2011年及2012年346,195、417,347及463,846例唐氏综合征产前筛查数据。全省产前筛查孕妇年龄基本呈正态分布,不同筛查机构年龄分布不一致。全省及各筛查机构体重分布趋势相似,在体重高值部分有拖尾。筛查孕周集中在17周左右,孕周确定方式各筛查机构存在显著差异。使用LifeCycle软件的SPR高于使用2T软件的SPR,9家机构SPR各月份波动明显。各筛查机构SPR随月份波动明显,95%置信区间内标志物Median MoM值随筛查月份、孕周及体重变化超过10%或CUSUM图反映的正、负偏倚可以及时反映筛查机构的筛查质量。代表性问题举例分析:1、对标志物MedianMoM出现正、负偏倚的机构,采用2011年本地人群的参比中位数后,筛查标志物MedianMoM值更接近于1,筛查阳性率下降,但不影响检出率。2、以13号筛查机构为例,其2012年各月份21-三体SPR波动较明显,使用LifeCycle软件年龄校正筛查阳性率为4.98%。对同一时期该筛查机构由不同采血点采集的标本分开进行分析发现2012年筛查机构采血标本与采血点送检标本筛查阳性率不同,采血点送检标本SPR较高,两组标本甲胎蛋白(fetoprotein,AFP)相近且均负偏,游离绒毛膜促性腺激素β 亚基(free Beta human chorionic gonadotrophin,Free βhCG)的 Median MoM 不同,采血点送检标本Free βhCG正偏倚,高于筛查机构采血标本。两组标本未结合雌激素(unconjugaed estriol,uE3)的Median MoM不同,波动明显且均正偏倚,且采血点送检标本尤其明显。3、对筛查机构11更换检测仪器前后筛查数据进行比较,更换前后筛查标志物MoM值不同,更换仪器后AFP,Free βhCG的正偏倚更明显,未结合雌三醇(unconjugaedestriol,uE3)负偏倚明显。4、本研究共随访到213例阳性病例,对阳性病例进行分析,Lifecycle软件代替2T软件后多检出6例阳性病例,可提高检出率。结论:1、本研究建立了中国地区特色的产前筛查质量改进"DQASS"模式。产前筛查质量控制涉及孕妇年龄、体重、孕龄等基本参数,标本质量、实验室检测、风险分析软件及参比中位数选择等多个环节;2、年龄校正SPR、95%置信区间内标志物Median MoM值随筛查月份、孕周及体重的分布图及累积和图(CUSUM图)是用于筛查质量监测的有效统计学指标,基于这些评估指标及统计学方法建立产前筛查质量管理软件系统;3、定期使用本实验室的筛查数据更新中位数校正方程及体重校正方程,可以纠正系统误差,降低SPR及FPR。当筛查人群、仪器设备等发生改变时,需更新参比中位数,但在更新应用前应评价参比中位数的准确性及稳定性;4、标本质量影响筛查效能,对多采血点的筛查机构应定期对各采血点标本质量进行单独评价,将各采血点标本在同一个筛查机构进行风险计算时应注意采血机构与筛查机构标本质量同质化。5、风险计算软件的优化可以降低假阴性,提高检出率。
[Abstract]:Serological screening during pregnancy, cytogenetic diagnosis of chorionic and amniotic fluid and ultrasound diagnosis are still the mainstream interventions in the two level prevention of Down's syndrome. High detection rate, low false positive rate and high health economic benefits are always pursued by experts in the field of prenatal diagnosis. It is also true through the efforts of nearly thirty years. The development of new screening markers and the establishment of new screening programmes can achieve greater screening effectiveness, while perfect quality management is to maximize the screening effectiveness of a screening program. Therefore, the establishment of an appropriate quality control system for prenatal screening is the basis for the better effectiveness of the high efficiency screening program to the clinical transformation. Although the scale of prenatal screening has been carried out for 13 years in China, the scale is increasing, but the quality control remains basically the most basic indoor quality control and interventricular quality assessment stage. Based on the statistical evaluation of the large sample screening data, all these quality control processes are analyzed, and the quality control is promoted. It is still blank in China. The quality control of prenatal screening for Down syndrome is special. The results of prenatal screening usually take the risk rate of a congenital anomaly as the report form. The core of the quality control is the reliability of the risk assessment of Down's syndrome. The evaluation of interventricular quality only involves the accuracy of biochemical markers, only part of the quality control of prenatal screening. The quality control of prenatal screening should involve every important link and key factors affecting the reliability of the risk calculation, including the basic parameters of pregnant women's age, weight, pregnancy weeks, and the accuracy of the laboratory test results, and the risk of the risk. On the other hand, prenatal screening for false positive rates and abnormal fetus detection rate in early and middle pregnancy is the index of statistical evaluation after birth, and there is a time span, serological screening for pregnant women, further prenatal diagnosis and final delivery of high-risk pregnant women are not all. In a medical institution, the collection, transport, testing and software application of blood samples can involve several different medical institutions and also have a geographical span. This makes the quality evaluation and management of prenatal screening more complex and arduous. How to find out the quality problems of screening, analyze the reasons, and put forward the rectification method is also a problem that needs to be solved through the change of statistical indicators. The quality management software system of prenatal screening and the realization of network can be used to evaluate the screening quality of screening institutions at all levels and efficiently, and promote the system of prenatal screening management. Objective: to promote the quality management of prenatal screening in our country. In order to standardize the quality management of prenatal screening in our country, this study takes the screening cases from 2010 to 2012 in Zhejiang as the data base, and discusses the key statistical indicators of quality assessment, the key assessment link, the evaluation flow and the methods of improving the quality of screening. The quality evaluation and management system of prenatal screening diagnosis suitable for the national conditions of our country, improve the efficiency of prenatal screening and improve the quality of pre - diagnosis in China. Methods: the data of prenatal screening for 2010~2012 years before 17 home production screening institutions in Zhejiang province were collected, involving 1227388 cases of pregnant women screening, through the basic parameters of pregnant women's age, weight, pregnancy weeks and so on. The Q-Q map, the Q-Q map, the age corrected screening positive rate (screening positive rate, SPR), the median median of the median of the 95% confidence interval (MedianMoM) with the screening month, the distribution map of pregnancy and weight, and the accumulation and graph (CUSUM chart) for screening the quality of the 17 screening machines. Through the above methods, 17 screening institutions To evaluate the quality of screening, find the link of screening quality, select representative parts, such as specimen quality, laboratory testing, risk software and reference median selection, to find out possible reasons and put forward improvement and optimization measures. Evaluation of the median of serum markers, analysis software Results: This study collected data of prenatal screening for 346195417347 and 463846 cases of Down's syndrome in 2010, 2011 and 2012. The age of prenatal screening of pregnant women was basically normal distribution, and the age distribution of different screening institutions was not consistent. The body weight distribution trend of the province and the screening institutions was similar, and the weight was high. The value part has trailing. The screening of pregnancy weeks is concentrated for about 17 weeks. There are significant differences in the method of screening for pregnancy. The SPR using LifeCycle software is higher than that of SPR using 2T software. The fluctuation of SPR in each month is obvious in 9 institutions. The Median MoM value of the marker in the 95% confidence interval varies with the month of screening, pregnancy week and in the 95% confidence interval. The negative bias can reflect the screening quality of screening institutions in time. The negative bias can reflect the screening quality in time. Representative problems are analyzed with examples: 1, the positive and negative bias of the marker MedianMoM, with the median of the local population in 2011, the screening marker MedianMoM is closer to 1, the positive rate of screening is decreased, but no shadow is found. The detection rate of.2 was taken as an example. The 21- trisomy SPR fluctuation in 2012 was more obvious. The positive rate of the LifeCycle software age correction screening was 4.98%. on the samples collected at the same period from different blood collecting points in the same period, and found that the screening machine and blood samples were screened in 2012. The sex rate was different, the sample of blood sampling was higher in SPR. The two groups were similar and negative to alpha fetoprotein (fetoprotein, AFP), and the Median MoM of free chorio gonadotropin beta subunit (free Beta human chorionic gonadotrophin, Free beta hCG) was different. The Median MoM which was not combined with unconjugaed estriol (uE3) was different, and the fluctuation was obvious and positive bias, and the blood sampling inspection specimens were especially.3. The screening data were compared before and after the replacement of the screening apparatus 11. The MoM value of the screening markers before and after the replacement was different. After the replacement of the instrument, the positive bias of the Free beta hCG was more obvious and uncombined. The negative bias of female three alcohol (unconjugaedestriol, uE3) was obvious.4. This study was followed up to 213 positive cases. The positive cases were analyzed. 6 positive cases were detected by Lifecycle software instead of 2T software. Conclusion: 1. This study established the quality improvement "DQASS" model of prenatal screening in China. Quantity control involves the basic parameters of pregnant women's age, weight, gestational age, specimen quality, laboratory testing, risk analysis software and selection of reference median. 2, age correction SPR, 95% confidence interval marker Median MoM value with screening month, gestational and weight distribution and accumulation and map (CUSUM chart) are used for screening quality monitoring Effective statistical indicators, based on these evaluation indicators and statistical methods to establish the quality management software system for prenatal screening. 3, regular use of the laboratory screening data to update the median correction equation and weight correction equation, can correct the system error, reduce SPR and FPR. when screening people, equipment and other changes, need to be updated. The median of the reference ratio, but the accuracy and stability of the median of the reference ratio should be evaluated before the renewal of the application. 4, the quality of the specimen affects the screening effectiveness. The screening institutions for the multiple blood sampling should evaluate the quality of the samples on a regular basis, and should pay attention to the blood collecting mechanism and screening in the same screening organization for the risk calculation. Check the quality of the specimen is homogeneous.5, the optimization of risk calculation software can reduce false negative and improve the detection rate.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：R714.5

【相似文献】