Between-Lot Variation in External Quality Assessment of Glucose: Clinical Importance and Effect on Participant Performance Evaluation

Abstract

Background: External quality assessment schemes (EQAS) are conducted to evaluate user performance (participant assessment) and to assess different methods and instruments (method assessment). The quality of control materials is crucial to achieving these goals. Inconsistencies in between-lot variations detected by use of different control and sample materials may affect EQAS outcomes. Methods: For the Accu-Chek Sensor, Precision Xtra, Ascensia Elite, and HemoCue 201 glucometers, 3 different lots of glucose strips were used with each instrument. Method assessment results from analysis of capillary blood and 3 control materials were used to calculate between-lot differences. A simulation study was performed to evaluate the effect of between-lot variation on participant assessment results. Results: With the Precision Xtra, the results obtained with EQA control material mirrored those obtained with capillary blood, but for the other instruments, we found between-lot differences of as much as 1.3 mmol/L, which were substantially greater than those found with capillary blood and of clinical importance at decision limits. The simulation study showed an effect on participant assessment results related to the target values, with the percentage of poor results decreasing (38%, 10%, and 4%) with the use of common, method-specific, and lot-specific target values, respectively. Conclusions: Between-lot variation may influence participant EQA results for participant and method assessments. The clinical relevance of between-lot variation

External Quality Assessment of Point-of-Care Methods: Model For Combined Assessment of Method Bias and Single-Participant Performance by the Use of Native Patient Samples and Noncommutable Control Materials

“…It is known that noncommutable samples can give different lot-to-lot variation compared to commutable samples (28,29 ). This was also the case for some lot numbers in our study, for which significant differences were seen for the noncommutable samples and not for the native patient samples (Fig.…”

American Association Of Clinical Endocrinologists And American College Of Endocrinology 2016 Outpatient Glucose Monitoring Consensus Statement

“…Manufacturing defects and test-strip lot-to-lot variations directly impact accuracy and introduce bias (156,157). Bias is typically measured in the hypoglycemic range, target range, and hyperglycemic range.…”

The Importance of Reagent Lot Registration in External Quality Assurance/Proficiency Testing Schemes

“…However, one should take into account that the EQA materials also can be noncommutable between reagent lots in the same measurement procedure, as shown for INR in this study. Therefore in such cases, if large differences between reagent lots are present, lot-specific target values should be considered to avoid giving some participants wrong evaluations (7 ). As an example: for the high INR level with a target value of 4.50 INR in survey 2/2014, one of the reagent lots (lot E) had a median INR value of 3.85 INR, which was outside the lower limit of acceptable performance (3.92 INR), meaning that at least half of the participants using this lot were evaluated as "poor."…”

“…It is often taken for granted that noncommutable control materials are commutable within one measurement procedure, i.e., "commutable between lots," indicating that the differences between reagent lots found using the control material also will be valid for patient samples. This is, however, not always the case (7 ).…”

“…With the use of noncommutable material with peer group target values, the interlaboratory variation within one measurement procedure can be explained by lot-to-lot variation and participant performance. In EQA, focus has usually been on variation between measurement procedures and participant performance, but to our knowledge little has been published on lot-to-lot variation, although a few exceptions exist (6,7 ). One of the reasons could be that reagent lots are probably not commonly registered in EQA.…”