評估人類血清與品管血清的檢測結果差異－以人類甲狀腺刺激賀爾蒙（hTSH）為例

Assessment of the Variation between Human Sera and Control Sera Samples - In Human Thyroid Stimulating Hormone (hTSH) Determination

王安美(An-Mei Wang) ； 吳明哲(Ming-Che Wu) ； 羅青雯(Ching-Wen Lo)

《核子醫學暨分子影像雜誌》 30卷1期 (2017/03) Pp. 13-19

https://doi.org/10.6332/ANMMI.3001.002

摘要

背景：現今醫學實驗室必須執行內部品管並且參加外部品管評估計劃（亦稱能力試驗），以確保實驗室的檢測品質。但倘若實驗室得到的品管不合格結果是源自於內外部品管提供機構，所提供的品管物質與試劑方法之間有干擾存，實驗室應要有驗證方法以確認實驗室的檢測品質。材料與方法：本研究以一般人類檢體匯集成三個不同濃度的人類血清檢體（serum pools），並以內部品管物質（lyphocheck control material）做比對驗證基礎，分析比較四種不同試劑方法的精密度（precision）、p值（p-value）、Z-評分（Z-Score），以客觀的統計數據驗證。結果：由數據中發現品管物質檢體和人類血清檢體之間存在著檢測結果的差異，且不同的試劑方法其差異性和差異程度亦不相同。結論：實驗室要正確的使用品管物質檢體，並且辨認其檢測結果的統計及詮釋方式是否正確，以及建立實驗室的驗證方法，方能判斷實驗室的檢測品質是否符合要求以避免啟動不必要的檢討措施造成額外的檢測成本增加和後續的醫療風險。

關鍵字

品管物質 ； 干擾 ； p值 ； 驗證 ； Z-評分

並列摘要

Background: Modern regulations require medical laboratories to perform internal quality control and also participate in external quality control assessment programs (also known as Proficiency Testing) in order to ensure the quality of laboratory tests. When QC results fail due to interference between internal and external quality control materials and reagents however, laboratories must then establish verification methods to verify the quality of laboratory tests. Materials and Methods: In this study, human serum samples were collected and divided into serum pools of three different concentrations and compared with internal control materials (Lyphocheck control material) as the verification basis. The Precision, p-value, and Z-Score were analyzed with four different reagent methods by objective statistics. Results: We found that there are variations between test results of quality control materials and human serum pools and also that the degree of variations depends on different reagent methods. Conclusion: Medical laboratories should correctly use quality control materials and identify statistics and interpret assessment results with the use of verification methods in order to judge whether or not the test quality meets quality requirements and avoid initiating unnecessary review processes which cause additional laboratory testing and consequent medical risks.

並列關鍵字

control material ； interference ； p-value ； verification ； z-Score