Do We Need the Replacement of YSI 2300? A View from the Clinical Laboratory

Abstract

In this issue of Journal of Diabetes Science and Technology, Baumstark et al. evaluated the analytical performance of a bench-top laboratory glucose analyzer (SUPER-GL) intended for replacement for the YSI2300-STAT analyzer, that served for several decades as a comparator method in clinical and analytical studies of blood glucose monitoring systems (BGMS). The authors concluded that the SUPER-GL’s overall performance is comparable to that of YSI2300-STAT, and has the potential to be a candidate comparator analyzer. However, the question is if we need to recommend as a “comparator method,” a specific device, that measure glucose using the same analytical method with most BGMS. In this analysis we present our point of view hoping to generate a discussion on the necessity for such a replacement.

Why We Want to Self-Monitoring Blood Glucose Levels?

Self-monitoring of blood glucose (SMBG) using a BGMS plays a central role in the optimal management of diabetes mellitus. It can serve people with diabetes in 2 ways: first, to increase patients’ awareness of his/her disease and the impact of lifestyle changes on controlling blood glucose levels and help them improve their self-care practices and second, to determine the appropriate insulin doses for those insulin-treated individuals.^1-3 Moreover, SMBG has many proven benefits, such as helping patients to achieve glycated hemoglobin A1c (HbA1c) targets, to minimize glucose variability, and to help them in the prediction of hypoglycemia.^4,5

Accurate knowledge of their measured blood glucose levels, can give people with diabetes more confidence in adjusting their insulin doses. Several studies, including modeling analyses, have demonstrated that errors in measuring blood glucose can lead to inaccuracies in insulin dosing.^6-9

How We Measure Glucose Concentrations in Blood?

There are different ways to measure glucose concentrations in blood. Nowadays, glucose is almost always determined using an enzymatic assay based on one of the following enzymes: glucose oxidase, hexokinase or glycose dehydrogonase. These enzymatic reactions are linked to chromogenic reactions that are measured photometrically (Laboratory methods) or to reactions that show changes in electron flux and are measured electrochemically (POCT and BGMS). All 3 enzymes are highly specific for glucose, thus any measurement discrepancies with enzymatic methods are usually not due to interference from other carbohydrates, but rather to the measurements and measuring devices.

Which Samples Are Used to Measure Glucose and How Results Are Reported?

Glucose concentration can be determined using either whole blood, plasma or serum samples. The last 2 are preferred by the labs, because glucose concentrations from whole blood are usually 15% lower compared to plasma, due to the additional water content in the blood cells. ¹⁰ Standard laboratory methods require serum or plasma and venous blood collection. At the point-of-care (POCT) however, plasma is less convenient and whole venous or capillary blood is used. BGSM patients use capillary blood with the finger t being the most preferred sampling site.

In order to avoid clinical misinterpretations, since 2005, the International Federation of Clinical Chemistry (IFCC) recommends that all glucose-measuring devices, should report the glucose concentration in plasma values. Whole blood values should be multiplied by a fixed factor, 1.11, to obtain plasma equivalent values. Directly reading glucose biosensors, measuring molality, must also recalculate measurements to report in plasma glucose. ¹¹

What is the Reference Method to Measure Plasma Glucose?

Measurement of glucose in human serum by, have been reported since the late 1970s. In 1982 NIST described 2 ID GC-MS methods for the measurement of glucose in serum (or plasma). The NIST method were laborious, time consuming and required extensive sample pre-analytical preparation before the sample is injected to GC-MS. ¹² However, this method, is considered by Joint Committee for Traceability in Laboratory Medicine (JCTLM), the reference method for quantification of glucose in biological matrices. In 2010, NIST modified and simplified this method by reducing the pre-analytical preparation time. ¹³ So far, all isotope dilution mass spectrometric methods (IDMS) methods that have been developed, for the analysis of glucose in human serum (or plasma), were performed on GC-MS systems.^14-16 Recently, IDMS methods, based on liquid chromatography tandem mass spectrometry (LC-MS/MS), have been developed requiring less laborious sample preparation and are listed in JCTLM as reference methods for glucose determination.^17,18 The main purpose of the NIST reference method was to assign values to standard reference material SRM 965b, which is intended primarily for use in evaluating the accuracy of secondary procedures for the determination of human serum glucose and for the validation of secondary reference materials. That means that manufacturers could use the SRM 965b to calibrate their own “reference methods.” These in turn will be used to evaluate the analytical bias of BGMS.

What Are the Regulatory Requirements for BGMS Accuracy Today?

Requirements for BGMS are currently regulated, by the international standard ISO 15197:20131 (harmonized in the European Union as EN ISO 15197:20152).^19,20 On the other hand, in US, FDA regulates the specifications of all BGMS intended for patient use, with a separate guidance since 2016 which was recently updated in 2020. ²¹ The FDA believes that the criteria set forth in the ISO 15197 are insufficient and that the standard does not adequately address the performance of BGMS in the hypoglycemia range and across test strip lots.

Regarding the analytical accuracy of a BGMS, ISO 15197:2013 specifies the following criteria:

• (1) compared to a traceable laboratory method, at least 95% of BGMS results have to be within ±15 mg/dl at glucose concentrations <100 mg/dl and within ±15% at ≥100 mg/dl;

• (2) in a consensus error grid analysis at least 99% of results have to be within zones A and B

For the comparison method ISO requires a “measurement procedure that conforms to the traceability requirements of ISO 17511” and it is up to the lab to verify “trueness and precision.”

On the other hand, FDA requires the following:

• (1) 95% of glucose values must be within ±15% of the comparator method results, across the entire measuring range of the device, and 99% of all SMBG results must be within ±20% of the comparator method results across the entire measuring range of the device and

• (2) In order to meet the clinical needs of the user population, BGMS should minimally be able to measure blood glucose accurately between 50 and 400 mg/dL

The term “comparator method” here refers to a laboratory based method, which measures blood glucose that has been validated for precision and accuracy and which is traceable to a higher order reference material and/or reference method. This traceability chain should include as few steps as possible to reduce bias.

While the required performance characteristics for the BGMS are described in detail, it is our belief that the analytical performance specifications for this “comparator method” are not addressed adequately in any standard. A “comparator method,” in order to fulfill its mission, should be able to prove that is capable of very narrow coefficient variations (CVs) in precision and bias. The impact of reference measurement procedures bias on the accuracy of BGMS, has been presented excellently in several publications.^22-24

Laboratories performing the “comparator method” should participate in accuracy based external quality assessment schemes (EQAS). In such a scheme the EQAS samples distributed to labs must have values assigned by one of the JCTLM recognized reference methods. We recognize the difficulties of creating such a program using whole blood or even worse capillary blood, but since BGMS should report glucose as plasma glucose, we don’t see why the comparator method should not be a plasma based method. In any case the NIST SRM 965b is a set of plasma pools and the reference method they maintain is also based on plasma glucose measurements.

The need for specific accuracy and precision characteristics for the “comparator method” has been emphasized previously and needs to be addressed universally by the Diabetes and Laboratory communities. ²⁵

Can we avoid the Clinical Chemistry Lab by creating a “comparator method” that will have true reference method characteristics and at the same time will not require core lab facilities and procedures? In our opinion, this is difficult if we set the specifications of the comparator method up to a true reference method. One can argue that mass spectrometric methods are not the fastest methods in giving results but newer methods are becoming faster every day. ²⁶ And as standardized methods for sample collection and preparation and storage are already published we believe that a laboratory based true reference methods for our comparisons it is possible.^27,28

How Do We Establish the Accuracy of a New Blood Glucose Monitoring System (BGMS) Today?

Currently, there is considerable discussion about the accuracy of blood glucose meters and the standards used to determine their accuracy.

It is extremely important to establish the accuracy of glucose meters throughout the range of their intended use and under conditions of daily use. It is well known that BGMS inaccuracy affect not only insulin doses, resulting in hyperglycemia or hypoglycemia, but also the diagnosis and treatment of hypoglycemia events and even more increases patient anxiety and lack of confidence in their meters, resulting in the deregulation of diabetes.

Currently the most widely used method for determining accuracy of a new blood BGMS is by comparing its performance to a bench-top POCT analyzer, that measures glucose using a method based on enzymatic glucose reactions similar to those used by BGMS themselves.

Although these measurement procedures could be applied on several stand alone instruments, the YSI 2300 Stat Plus Glucose Lactate Analyzer (YSI2300) has been widely used as a comparator “reference system” in clinical and analytical studies, to determine the accuracy of BGMS.^29-34 This device has also been used by manufacturers of BGMS as a comparator instrument for the calibration of lots of many glucose test strips.

This analyzer is no longer produced by its manufacturer and from July 2021 the company will stop supporting it. This initiated a discussion on its replacement.^35,36

We also do not believe that the “comparator method” should be based on the same reaction principle as the BGMS, as it will suffer the same limitations as the devices it will be called to evaluate.

Do We Need a Replacement of YSI2300 as a Sole Comparator Method?

It is our opinion that we do not, for several reasons.

The use as a “comparator method” to determine the accuracy of the BGMS, of a device that measures glucose using the same enzymatic reactions with the BGMS under evaluation is fraught with all possible sources of error the BGMS have. Therefore, such a device is not necessarily the best method of comparison.

The result achieved with this device may vary significantly between testing sites (laboratories).

It may create a monopoly for a specific manufacturer and does not give the manufacturer or the evaluation laboratory the freedom to choose a real reference method. The chosen device might not be available everywhere around the world. At some point of time we will need to replace it as better methods will become available.

In conclusion: we believe that it is important that the performance of each meter to be evaluated in comparison with a laboratory based “true reference method” and that detailed analytical performance specification should be published not only for BGMS but also for reference methods.