南臺灣某區域醫院假性低糖化血色素值之原因分析

An Analysis of Factors Associated with Falsely Low Hemoglobin A1c Values in a Community Hospital in Southern Taiwan

目的：糖化血色素（hemoglobin A1c）除了是監測血糖的重要指標，自2010年起，更被美國糖尿病學會建議作為診斷糖尿病的標準。然而，糖化血色素的測定有其干擾因子，臨床醫師若未能察覺，會因誤判而影響病患的診斷及處置。本研究旨在分析及探討糖化血色素值≦4%的原因，以了解造成假性低糖化血色素的干擾因子。方法：於2010年1月至2011年12月期間，本院以離子交換高效能液相層析法（ion-exchange high-performance liquid chromatography）測定所有送檢糖化血色素的樣本，針對糖化血色素值≦4%的個案，執行血色素電泳、全血球計數及glucose-6-phosphate dehydrogenase（G6PD）的檢驗，並進行回溯性病歷分析。結果：在64653件的檢體中，計有27例個案的糖化血色素值≦4%，依據可能干擾糖化血色素測定的原因分成四類：（1）8例變異血色素，全屬Hb J；（2）9例溶血性貧血，包括4例G6PD缺乏症及4例使用治療C型肝炎的藥物ribavirin；（3）7例上消化道出血，包括5例肝硬化；（4）3例原因待確認的貧血。相較於無變異血色素的個案，具變異血色素的個案有較低的糖化血色素值（p=0.008）及較高的血色素（p=0.002）。結論：在南台灣，變異血色素（Hb J）、急性出血及溶血性貧血是造成假性低糖化血色素的主要干擾因子，G6PD缺乏症及抗病毒藥物ribavirin則是引發溶血反應而造成假性低糖化血色素的兩個最常見原因。作者建議以離子交換高效能液相層析法測定糖化血色素時，臨床醫師若遇見疑似假性低糖化血色素個案，除了檢視有否異常的層析圖，以便發現變異血色素之外，應常規性測定G6PD和全血球計數以篩檢可能的干擾因子。

Purpose: Hemoglobin A1c (A1C) is an invaluable indicator of glycemic control, and, since 2010, the A1C test has been recommended as one of the diagnostic criteria for diabetes. There are some factors that interfere with the measurement of A1C. If clinicians are not aware of these limitations, they may be misled in their diagnosis or manage diabetes inappropriately. This study explored the factors associated with A1C values≦4% and identified the causes of falsely low A1C values. Methods: Between January 2010 and December 2011, we used an ion-exchange high-performance liquid chromatogram (HPLC) to measure A1C in our hospital. All samples with A1C values≦4% were submitted for hemoglobin electrophoresis, complete blood count (CBC) and glucose-6-phosphate dehydrogenase (G6PD) levels. Medical records were reviewed to identify conditions that might be associated with these low A1C values. Results: Of the 64,653 samples analyzed, 27 had A1C values≦4%. The interfering factors could be divided into four categories: 1) eight cases with hemoglobin variant J (Hb J); 2) nine with hemolytic anemia, including 4 with G6PD deficiency and 4 taking ribavirin for hepatitis C; 3) seven with upper gastrointestinal bleeding, including 5 with cirrhosis of the liver; 4) 3 with anemia of undetermined cause. The Hb variant group had lower A1C values (p=0.008) and the non-variant group had lower hemoglobin levels (p=0.002). Conclusions: Hemoglobin variant J, acute blood loss and hemolytic anemia were the major causes of falsely low A1C when the ion-exchange HPLC method of testing was used. G6PD deficiency and anti-viral drug ribavirin were the two most common causes of hemolysis. In addition to visual inspection of the chromatogram to detect abnormalities, CBC and G6PD should be determined in order to identify the reasons for falsely low A1C values.