BACKGROUND: A spuriously elevated serum potassium value could possibly cause medical decision errors because it leads to masked hypokalemia or pseudohyperkalemia. The aim of this study was to develop a correction equation for falsely elevated potassium level caused by hemolysis. METHODS: A total of 988 samples with a hemolysis index (HI) value greater than the potassium alert HI value were recollected within two hours from initial collection. We divided 988 paired samples into 3 groups: hypokalemia, normal, and hyperkalemia. When samples were checked after recollection within 2 hours, 525 cases showed HI of 1. We analyzed the relationship between the delta of initial and recollected samples' HI values and the delta of initial and recollected samples' potassium levels, resulting in 5 different delta groups. RESULTS: The proportion of masked hypokalemia and pseudohyperkalemia was 17.6% (125/710) and 64.1% (139/217), respectively. The trend and distribution of potassium concentration for each of the 5 delta HI groups according to delta HI level showed an exponential curve. From this exponential curve function, a correction equation for estimation of true potassium concentration in hemolyzed specimens was calculated: measured potassium (-0.0561e 0.6578* delta HI + 0.0804). CONCLUSIONS: The clinical application of the correction equation for adjusting the hemolysis effect on potassium concentration could be useful for the detection of masked hypokalemia or pseudohyperkalemia.
BACKGROUND: A spuriously elevated serum potassium value could possibly cause medical decision errors because it leads to masked hypokalemia or pseudohyperkalemia. The aim of this study was to develop a correction equation for falsely elevated potassium level caused by hemolysis. METHODS: A total of 988 samples with a hemolysis index (HI) value greater than the potassium alert HI value were recollected within two hours from initial collection. We divided 988 paired samples into 3 groups: hypokalemia, normal, and hyperkalemia. When samples were checked after recollection within 2 hours, 525 cases showed HI of 1. We analyzed the relationship between the delta of initial and recollected samples' HI values and the delta of initial and recollected samples' potassium levels, resulting in 5 different delta groups. RESULTS: The proportion of masked hypokalemia and pseudohyperkalemia was 17.6% (125/710) and 64.1% (139/217), respectively. The trend and distribution of potassium concentration for each of the 5 delta HI groups according to delta HI level showed an exponential curve. From this exponential curve function, a correction equation for estimation of true potassium concentration in hemolyzed specimens was calculated: measured potassium (-0.0561e 0.6578* delta HI + 0.0804). CONCLUSIONS: The clinical application of the correction equation for adjusting the hemolysis effect on potassium concentration could be useful for the detection of masked hypokalemia or pseudohyperkalemia.