军事知识服务平台
新生儿先天性甲状腺功能减退症筛查假阳性结果的发生特征及溯源分析
Occurrence characteristics and traceability analysis of false-positive results in newborn screening for congenital hypothyroidism
摘要 目的:分析新生儿先天性甲状腺功能减退症(CH)筛查假阳性结果发生特征,评估及其与试剂批号、质控失控、检测时段等因素的相关性,并构建假阳性结果的全流程溯源路径。方法:收集2021年1月1日至12月31日在河南省新生儿疾病筛查中心接受CH筛查的750 240例新生儿为研究对象,比较不同试剂批号与检测时段的假阳性发生率;通过标准差指数(SDI)分布、失控次数及失控率评价室内质控稳定性,及采血日龄、季节、不同试剂批号/检测时段与假阳性发生的关系;通过贝叶斯多因素回归模型量化质控失控、采血日龄、季节及试剂批号对假阳性风险的影响;基于Sankey图构建假阳性结果全流程溯源路径,采用2024年度独立筛查数据对质控异常与假阳性发生的关系进行时间外部验证。结果:不同采血日龄、季节、试剂批号/检测时段组间假阳性发生率差异均具有统计学意义(均 P<0.001)。共识别3次质控失控事件,其中低水平质控1次、高水平质控2次,室内质控分析显示,低、高水平质控失控率分别为0.12%(1/856)和0.23%(2/856),位于±3 SDI范围内的质控结果分别占99.88%(855/856)和99.77%(854/856)。失控批次及其前后各1批次内共筛查37 049例,假阳性1 011例,假阳性率为2.73%(1 011/37 049),高于非失控时段的1.01%(7 209/713 191)(χ 2=959.28, P<0.001)。贝叶斯回归分析表明,质控失控事件窗口假阳性风险升高( OR=2.35,95% CI 1.85~2.98),冬季检测与较高假阳性风险相关( OR=2.90,95% CI 2.30~3.65)。而采血日龄延后与假阳性风险降低相关。全流程溯源路径Sankey分析显示,假阳性主要来源于分析阶段(约占 46%),其中试剂批号差异和质控失控为最主要因素。2024年独立时间验证中,质控异常事件窗口假阳性率仍高于非异常窗口[2.50%(488/19 500)比 1.54%(7 788/504 999), P<0.001],与主研究结果一致。 结论:新生儿CH筛查假阳性与分析过程不稳定因素密切相关,尤其是质控异常;加强批次级动态质量监测有助于及时识别异常并降低假阳性风险。
Abstract:
Objective:To systematically characterize false-positive results in newborn screening for congenital hypothyroidism (CH), evaluate their associations with reagent lots, quality control (QC) failures, and testing periods, and establish a workflow-based tracing pathway for false-positive results.Methods:A total of 750 240 newborns screened for CH at the Henan Provincial Newborn Screening Center between January 1 and December 31, 2021 were included. False-positive rates were compared across reagent lots, testing periods, blood-sampling age groups, and seasons. Internal QC stability was evaluated using the distribution of standard deviation index (SDI) values, the number of out-of-control events, and out-of-control rates. Bayesian multivariable regression was performed to quantify the associations of QC failures, blood-sampling age, season, and reagent lot with the risk of false-positive results. A Sankey diagram was used to construct a workflow-based tracing pathway for false-positive results. An independent 2024 screening cohort was used for temporal external validation of the association between QC abnormalities and false-positive results.Results:False-positive rates differed significantly according to blood-sampling age, season, reagent lot, and testing period. Three QC out-of-control events were identified, including one at the low QC level and two at the high QC level. The out-of-control rates for the low-and high-level QC materials were 0.12%(1/856) and 0.23%(2/856), respectively, and 99.88%(855/856) and 99.77%(854/856) of QC measurements were within±3 SDI. A total of 37 049 newborns were screened within the QC out-of-control event windows, of whom 1, 011 had false-positive results. The false-positive rate was significantly higher within than outside these windows 2.73%(1 011/37 049) vs 1.01%(7 209/713 191)(χ 2=959.28, P<0.001). Bayesian regression showed that QC out-of-control event windows were associated with an increased risk of false-positive results ( OR=2.35, 95% CI 1.85-2.98), as was winter testing ( OR=2.90, 95% CI 2.30-3.65), whereas later blood sampling was associated with a lower risk. Workflow-based Sankey tracing showed that approximately 46% of false-positive results were attributable to the analytical phase, with reagent lot-related variation and QC failures being the major contributors. In the independent 2024 temporal validation cohort, the false-positive rate remained higher within QC abnormality event windows than outside these windows [2.50% (488/19 500)vs 1.54%(7 788/504 999), P<0.001], consistent with the primary cohort. Conclusion:False-positive results in newborn screening for CH are closely associated with instabilities in the analytical process, particularly quality control abnormalities. Strengthening dynamic, batch-level quality monitoring may facilitate the timely detection of abnormalities and reduce the risk of false-positive results.