Sten Westgard, MS
Is it time for a tighter quality requirement for glucose meters?
Fresh on the heels of Dr. George Klee's review of setting performance specifications, as well as the recent FDA public meeting on glucose meter quality, Dr. Klee and Dr. Brad Karon of the Mayo Clinic and Dr. James C. Boyd of the University of Virginia recently published a study that used simulation modeling to determine performance criteria for glucose meters:
"Glucose Meter Performance Criteria for Tight Glycemic Control Estimated by Simulation Modeling", Brad S Karon, James C. Boy, and George G. Klee, Clinical Chemistry 56:7; 1091-1097 (2010)
In this study, the authors used "29,920 glucose values from patients on TGC at 1 institution to represent the expected distribution of glucose values during TGC." Next, they "used 2 different simulation models to relate glucose meter analytical performance to insulin dosing error" using the following total allowable error (TEa) criteria: 10%, 15%, and 20%.
Just to quickly review, currently there are different recommendations for the quality required for glucose tests. ISO 15197 recommends a 20% total allowable error (TEa) for POC glucose devices. CLIA sets a 10% goal for glucose methods in the central laboratory. Both POC testing and laboratory quality testing devices are often used interchangeably in clinical settings, e.g., for tight glycemic control.The study authors judged those TEa quality requirements on the basis of their ability to minimize large (> 3 category) insulin dosing errors. What is a 3-category insulin dosing error? It's an error that causes the wrong dose of insulin to be administered, by more than one category of units. As the authors describe it:
"In our model, we assume that a single inaccurate glucose measurement leads to hypoglycemia via administration of excess insulin over a short time period. For this type of dosing error, glucose meter imprecision appears to be the major factor in determining error rate."
While the simulations showed that a 20% requirement would generate 6.1% 2-category insulin dosing errors and 0.3% >3-category insulin dosing errors. If the quality requirement were tightened to 15%, those chances of dosing errors would decrease to 2% and 0.02%, respectively. If the quality requirement was cut in half, to 10%, the simulation estimated no >3 category insulin dosing error, and only a 0.2% chance of a 2-category insulin dosing error.
After the simulation and evaluation of different quality requirements, Karon, Klee, and Boyd settled on 15% TEa as the better quality requirement:
The outcome of the recent public meeting on glucose meter quality held by the FDA earlier this year seems to be heading toward a same conclusion. A 20% quality requirement is too big and allows too many significant errors."The simulations suggest that 20% TEa would allow large insulin dosing errors that might lead to hypoglycemia. At 15% total error, both simulation models predict that these large (>3-category) insulin dosing errors occur very infrequently. Both simulation models also predict that further reducing total error to 10% reduces the frequency of 2-category insulin dosing errors, although additional studies will be necessary to determine the clinical impact of such errors."
In Dr. Klee's earlier editorial on setting quality specifications, he noted that using simulation studies and patient data was the best technique. The advantage of this approach is that it fits at the top of the hierarchy in the "Stockholm consensus methodology" for setting quality requirements, i.e., establishing requirements for a specific clinical situation, thus the outcome of such studies should care significant weight in the laboratory community.
Comments