Glucose Measurement: Confounding Issues in Setting Targets for Inpatient Management: Response to Dungan et al.

We thank Dungan et al. (1) for their recent review of the potential factors that can adversely affect inpatient glucose measurements. Their review is timely and provides a much-needed overview of an important and complex subject, especially now that hospitals are using point-of-care devices to monitor intensive insulin protocols.

However, we feel that the information listed in Table 1 of their study is potentially misleading. First, the presence or absence of a chemical interference cannot always be predicted based on the type of enzyme used for glucose detection. Certain reducing compounds (e.g., acetaminophen) do not interfere with optical detection technologies and do not interfere with many electrochemical methods regardless of enzyme. Rather, it is the presence of electrodes along with the amount of applied voltage that most likely causes interference (2). Second, oxygen interference cannot be assigned to all glucose oxidase methods. In a recent survey, there were 2,078 laboratory-based glucose oxidase methods (42.3% of the total number of glucose instruments) (3), all of which are designed to ensure that oxygen in the test samples does not adversely affect glucose results. Tang et al. (4) concluded that varying oxygen partial pressures only had a minimal effect on results from an optical point-of-care method that utilizes glucose oxidase. Third, while we agree that clinicians should pay close attention to the patient blood sample, even laboratory-based methods would be affected by preanalytical factors that affect test sample glucose concentration (e.g., comparing arterial-venous blood samples drawn during peak postprandial glucose). Point-of-care devices cannot be held responsible for protocol-specific sources of error.

In conclusion, point-of-care glucose meters are easy to use, fast, and accurate when used according to instructions and within stated limitations. Glucose meters have been shown to provide a clinical benefit when used in conjunction with insulin protocols and a trained staff (5). However, glucose meter performance should be evaluated independently of which enzyme is employed, and clinicians should refer to specific product package inserts for stated limitations and known interferences.