實驗室測量血中鉛含量一致性的評估：國人血中鉛濃度調查研究的經驗

Assessment of Test Consistency of Blood Lead Levels: The Experience from a Study on Blood Lead Levels in Taiwanese Population

孫建安(Chien-An Sun) ； 張玉坤(Yue-Cune Chang) ； 賴錦皇(Ching-Huang Lai) ； 吳德敏(Der-Min Wu) ； 劉紹興(Saou-Hsing Liou)

《中華公共衛生雜誌》 17卷2期 (1998/04) Pp. 103-110

https://doi.org/10.6288/CJPH1998-17-02-04

摘要

本研究藉由國內七家實驗室參與「國人血中鉛濃度調查研究」的機會，評估實驗室間與實驗室內測試低血鉛含量的一致性。測試樣本由本研究之中心實驗室所在之總醫院的血庫中隨機選取五名沒有職業性鉛暴露的捐血袋（每袋為250毫升），然後每位捐血者之捐血袋分裝成每管5毫升（共50管）的測試樣本，並冰存於4℃的冰箱中。研究人員於三次不同的時間點（每次間隔兩星期），從五名捐血者血袋所建構成的到試樣本群中，隨機選取每位捐血者之兩管測試樣本，以乾冰保存的方式送至受測實驗室。每個受到實驗室參考標準的樣本處理分析程序，以石墨爐原子吸收光譜儀來測試血液樣本中的鉛含量，並需於收到測試樣本兩個星期內完成測試工作。評估結果發現，各實驗室在測試同一時間送至實驗室之同一個案的第一管樣本時的變異係數（［標率差÷平均數］×100%）介於0.7%到58.9%之間，而在測試第二管樣本時的變異係數則為2.7%~102.6%。另外，以中心實驗室為比較基準，使用GEE(generalized estimating equation)分析模式發現各實驗室與中心實驗室之間亦存在有實驗室間的測試變異。這些結果顯示，血中鉛含量的測定存在有顯著性之實驗室間的測量差異。因此，在以血中鉛含量為生物標記的調查研究中應包含監測實驗室測試血中鉛品質的品管程序。本研究提供了評估實驗室測試血中鉛檢驗品質的模式，也提出了實證的數據來驗證實驗室測量血中鉛含量所面臨測試穩定性的實質困難。

關鍵字

血中鉛含量 ； 測試一致性

並列摘要

A proficiency study was done to assess the consistency of analyses for blood lead levels (BLLs) when performed by different laboratories. Specimens of blood were aliquoted into a series of test samples from five healthy blood donors and were sent to each laboratory on three separate occasions with two-week interval between each occasion. All test samples were handled identically during processing, storage, and retrieval, and were labeled in ways which would preclude their identification by the receiving laboratory. Blood lead concentrations were analyzed by seven participating laboratories, each used flameless atomic absorption spectrophotometry (FAAS) and followed the same protocols for BLL measurements. Statistical analysis of the data showed that the coefficients of variation (CVs) of laboratory performance on the first replicate of each subject were between 0.7% and 58.9% and the corresponding CVs for assays on the second replicate were in the range of 2.7% to 102.6%. In addition, results from the analysis based on generalized estimating equation (GEE) revealed a statistically significant variation in interlaboratory measurements of BLLs. These study results emphasized the need for rigorous evaluation and increased monitoring of laboratory performance in epidemiological studies using BLLs as biomarkers.

並列關鍵字

blood lead levels ； test consistency