Potential medical impact of unrecognized invitro hypokalemia due to hemolysis: a case series.

Abstract

The destruction of red cells during blood collection or with the processing of the sample continues to occur at a high rate, especially among emergency department (ED) patients. This can produce pre-analytical laboratory errors, particularly for potassium. We determined the incidence of hemolyzed samples and discuss the potential medical impact for hypokalemic patients who potassium level is artificially normal (pseudoeukalemia). Potassium results were obtained for a 6-month period. Using a measured hemolysis index (HI), hemolysis was present in 3.1 % for all potassium ordered (n=94,783) and 7.5 % for ED orders (n=22,770). Most of these samples were reported as having high normal result or were hyperkalemic. There were 22 hemolytic samples with a potassium of <3.5 mmol/L, and 57 hemolytic samples with a potassium in lower limit of normal (3.5-3.8 mmol/L). From this group, we examined the medical histories of 8 selected patients whose initially normal potassium levels were subsequently confirmed to have a potassium values that were below, at, or just above the lower limit of normal due to hemolysis. The primary complaint for these patients were: necrotizing soft tissue infection, pancreatitis, volume overload from heart failure with reduced ejection fraction, hypertension treated with hydrochlorothiazide, and presence of a short bowel syndrome. A subsequent non-hemolyzed sample was collected demonstrating hypokalemia in all of these patients. Within these cases, there was a potential for harm had hemolysis detection not been performed. We demonstrate the medical importance of detecting hemolysis for patients who have pseudoeukalemia. This is relevant because the HI cannot be obtained when electrolytes are tested using whole blood samples, and a normal potassium may lead to inappropriate patient management.