Investigators Did Not Follow Guidelines for Evaluating Point-of-Care Glucose Testing

In your January 2009 issue, Cook et al¹ compared glucose meter results for central venous catheter (CVC) blood and fingerstick to a laboratory method for CVC blood, and observed meter error that purportedly would lead to faulty treatment decisions. Although we agree with the “split-sample” design (ie, testing portions of the same blood sample with the meter and laboratory method), we believe the data were confounded by the authors’ failure to follow Clinical and Laboratory Standards Institute (CLSI, formerly the National Committee for Clinical Laboratory Standards) recommendations in several crucial aspects.² 

The authors state that the average time from blood sampling to laboratory analysis was less than 1 hour. However, CLSI recommends either blood analysis or centrifugation and erythrocyte removal within 5 minutes of collection, because longer delays can produce marked changes in glucose concentrations due to glycolysis.³ The bias and imprecision of the laboratory instrument (Olympus...