OBJECTIVE: To examine the effect of phlebotomy-induced hemolysis on serum insulin and C-peptide measurement by an immunochemiluminometric assay. METHODS: As part of a study designed to evaluate β-cell function in a group of adults with newly diagnosed type 2 diabetes, we tested insulin and C-peptide levels in 1,048 samples. In order to evaluate the effect of phlebotomy-induced hemolysis, we determined insulin and C-peptide levels simultaneously in hemolyzed and nonhemolyzed samples. RESULTS: Forty-seven (4.5%) of the 1,048 samples were affected by hemolysis. In 26 cases, we had paired hemolyzed and nonhemolyzed serum samples that allowed a simultaneous comparison. We found that all degrees of hemolysis led to a significant decrease in insulin level. In hemolyzed serum, the median (interquartile range) of the insulin was 5.6 (1.8 to 24.3) mIU/L, versus 21.3 (11.4 to 48.5) mIU/L in nonhemolyzed serum, representing a 25 to 98% loss. This phenomenon was not found for C-peptide levels. CONCLUSION: Clinicians have to be aware that even a mild degree of phlebotomy-induced hemolysis has a significant effect on serum insulin level determination, which can lead to misinterpretation of test results. This finding has important implications, especially in the evaluation of suspected cases of hyperinsulinemic hypoglycemia.
OBJECTIVE: To examine the effect of phlebotomy-induced hemolysis on serum insulin and C-peptide measurement by an immunochemiluminometric assay. METHODS: As part of a study designed to evaluate β-cell function in a group of adults with newly diagnosed type 2 diabetes, we tested insulin and C-peptide levels in 1,048 samples. In order to evaluate the effect of phlebotomy-induced hemolysis, we determined insulin and C-peptide levels simultaneously in hemolyzed and nonhemolyzed samples. RESULTS: Forty-seven (4.5%) of the 1,048 samples were affected by hemolysis. In 26 cases, we had paired hemolyzed and nonhemolyzed serum samples that allowed a simultaneous comparison. We found that all degrees of hemolysis led to a significant decrease in insulin level. In hemolyzed serum, the median (interquartile range) of the insulin was 5.6 (1.8 to 24.3) mIU/L, versus 21.3 (11.4 to 48.5) mIU/L in nonhemolyzed serum, representing a 25 to 98% loss. This phenomenon was not found for C-peptide levels. CONCLUSION: Clinicians have to be aware that even a mild degree of phlebotomy-induced hemolysis has a significant effect on serum insulin level determination, which can lead to misinterpretation of test results. This finding has important implications, especially in the evaluation of suspected cases of hyperinsulinemic hypoglycemia.
Authors: Ijeoma M Muo; Sandra D MacDonald; Ritu Madan; Sung-Jun Park; Ahmed M Gharib; Pedro E Martinez; Mary F Walter; Shanna B Yang; Justin A Rodante; Amber B Courville; Peter J Walter; Hongyi Cai; Michael Glicksman; Gioia M Guerrieri; Rivka R Ben-Dor; Ronald Ouwerkerk; Stephanie Mao; Jay H Chung Journal: Diabetes Metab Syndr Obes Date: 2019-05-27 Impact factor: 3.168
Authors: North Foulon; Elisha Goonatilleke; Michael J MacCoss; Michelle A Emrick; Andrew N Hoofnagle Journal: J Mass Spectrom Adv Clin Lab Date: 2022-06-10
Authors: Andrew P Demidowich; Jordan A Levine; Sheila M Brady; Cheryl D Johnson; Steven J Soldin; Jack A Yanovski Journal: Diabetes Obes Metab Date: 2020-04-20 Impact factor: 6.408