Interference in A Glucose Dehydrogenase-Based Glucose Meter Revisited

The US Food and Drug Administration (FDA)1 has issued a second alert to warn of serious errors with certain blood glucose–monitoring strips that use glucose dehydrogenase pyrroloquinoline quinine (GDH-PQQ) methods (1). Various nonglucose sugars, including maltose, galactose, and xylose, increase the glucose concentrations reported by these methods. Nonglucose sugars are present in a variety of pharmaceutical preparations(1), including Extraneal (icodextrin) peritoneal dialysis solution(2) and various immunoglobulin preparations(3). At least 13 deaths associated with erroneous meter measurements of glucose in patients receiving such pharmaceutical preparations have been reported to the FDA. A 76-year-old man at our institution had 3 blood glucose readings that unexpectedly were >22.2 mmol/L (>400 mg/dL) on a point-of-care (POC) testing meter that used a GDH-PQQ method (Roche Accu-Chek Inform). By contrast, glucose measurements made with a hexokinase method (Abbott Architect) at about the same times were in the range of 7.2–11.2 mmol/L (129–202 mg/dL), suggesting the possibility of a positive interference in the POC method. The patient’s medication list contained none of the pharmaceutical preparations listed in the FDA alert as producing interference in the POC testing method and none that were known to contain nonglucose sugars. In a search for other sources of interfering compounds that might contribute to this patient’s high …