Glucose meters have improved considerably since they were first introduced in 1960, but many questions are being asked about their accuracy and reliability in certain clinical situations. These questions have arisen because of the widespread use of these meters into clinical areas they have not been designed for such as critical care. The lack of understanding by some health professionals on factors that affect glucose results, such as sample type, glucose test strip methodologic limitations, calibration to recognized reference methods, and interferences, leads to misleading results that may affect patient care. Much debate continues on the quality specifications for glucose meters. Because there is an extensive use of these meters in different clinical scenarios, the setting of quality specifications will remain a challenge for regulatory and professional organizations. In this article, we have attempted to collect and provide relevant information addressing the limitations above. Pivotal to obtaining the best quality of results is education, particularly for diabetic patients monitoring their glucose. The International Federation of Clinical Chemistry and Laboratory Medicine through its Point-of-Care Testing Task Force and its Working Group on Glucose Point-of-Care Testing is actively working toward improving the quality of glucose results by improving education and working with the industry to improve strip performance and work toward the better standardization of strips.
In 1962, Clark and Lyons  developed the first glucose enzyme electrode that relied on a thin layer of glucose oxidase (GO) on an oxygen electrode with the sensor working by measuring the amount of oxygen the enzyme consumed. The first glucose meters were developed and marketed for self-monitoring of blood glucose for outpatients with diabetes. Glucose meters play an important role in the management and control of diabetes, particularly for those individuals requiring insulin. Over the years, meter usage has evolved to be used by individuals themselves for self-monitoring and by healthcare providers in a variety of clinical settings such as hospitals, emergency response units, nursing homes, physicians’ offices, and air ambulances. Glucose results derived from glucose meters are used by patients and healthcare professionals to make therapeutic decisions, so incorrect glucose results may have a negative impact on the patient outcomes.
The technology used by glucose meters has shown incremental improvements, such as ease of use, technical performance, and affordability, since the first handheld meter was commercially available in 1970s .
Over the years, their use, rightly or wrongly, has spread to hospitals and critical care units without confirming that they are “fit for purpose” for intended use, evidence of improved clinical benefit, and with limited understanding among healthcare professionals of their capability and limitations. Many healthcare professionals believe that the results from glucose meters parallel those produced by central laboratory analyzers and use them accordingly . The interchanging results between laboratory and glucose meters potentially lead to the misinterpretation of the patient’s glycemic status.
The glucose reference method is considered to be the isotope dilution technique using mass spectrometry . Point-of-care devices use direct reading biosensors that respond to active glucose molality, the amount of glucose per unit mass of water. The molality of glucose is identical in whole blood and plasma; however, when results are converted to a concentration so that they agree with laboratory methods, they become susceptible to water content, producing different results for plasma and whole blood.
This represents one of the key challenges of trying to develop a universal reference material that can be used to standardize all glucose meters. Even if the same sample type is used (e.g., whole blood or plasma), unacceptable differences have been obtained for both laboratory and point-of-care methods .
It has been recommended by the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) Scientific Division Working Group on Selective Electrodes and Point-of-Care Testing that a factor of 1.11 should be used to convert the concentration of capillary blood glucose to plasma glucose to reduce clinical misinterpretations when interchanging laboratory and point-of-care results [6, 7].
Blood glucose is monitored to manage diabetes treatment, including insulin dosage, use of oral antiglycemic agents, or measuring change in diet and/or lifestyle. The simulation models were developed by Boyd and Bruns to replicate the effects of glucose assay imprecision and bias on insulin dose to determine the desirable analytical performance for glucose monitoring . Using Monte Carlo simulation, they generated random true glucose values that were converted into observed glucose values. True glucose values across the range of 8.3–25.0mmol/L were arranged in increments coinciding with a sliding scale of insulin dose, which allowed the number of correct and incorrect insulin dosage decisions when increasing imprecision and bias were applied to the model.
With a total analytical error of 5%, 8%–23% incorrect insulin dosages were observed, which increased from 16% to 45% when imprecision increased to 10%. When the combination of imprecision and/or bias exceeded 10%, larger insulin dosage errors were found in >5% of occasions. To ensure correct insulin dosage in 95% of instances, bias and imprecision had to be <1%–2% depending on the average glucose and protocol used for insulin dosage. Although these dosage changes could not be directly linked with patient outcomes, they highlighted the importance of keeping imprecision and bias to a minimum and risks of using glucose methods that are imprecise.
Because the use of glucose meters becomes more widespread, questions are being asked about their accuracy and reliability in certain clinical situations, particularly those related to critical care. These concerns relate to the effectiveness of glucose meters in patient management or if they are being overused or used in situations that may result in the failure to achieve desired outcomes, inappropriate management, and ultimately increased cost of care.
Glucose meters are subject to numerous interferences including both endogenous and exogenous patient factors, such as pregnancy, medications, environmental factors, and operational factors.
Glucose meters: professional recommendations
Glucose meters should be evaluated under objective quality specifications, which may be based on clinical needs, professional recommendations, and regulatory requirements. Currently, there is no consensus for the analytical performance of glucose meters. Despite the numerous attempts by regulatory and professional organizations such as the Food and Drug Administration (FDA), Centers for Disease Control and Prevention, and American Diabetes Association (ADA), no universally accepted standards for glucose testing exist. Having no universally accepted quality specifications for glucose meters makes it difficult to compare the many published studies available.
The ADA has recommended that glucose meters agree with laboratory methods to within ±15% at all concentrations. The future goal of the ADA is that this is reduced to ±5% across all glucose concentrations .
The International Standards Organisation (ISO) is revising its criteria (ISO 15197) that currently state that 95% of the individual glucose results shall fall within ±0.83mmol/L (15 mg/dL) of the results of the manufacturer’s measurement procedure at glucose concentrations <4.2mmol/L (<75 mg/dL) and within ±20% at glucose concentrations ≥4.2mmol/L (≥75 mg/dL) . Using ISO 15197 performance criteria, a patient with a glucose concentration of 6.1mmol/L could be reported by the meter from 5.0 to 7.2 mmol/L, which could make the patient diabetic or non-diabetic. According to the generally accepted guidelines, diabetes can be diagnosed in the presence of a fasting venous plasma glucose of ≥7.0mmol/L or if the 2h postglucose load (75 g) is ≥11.1mmol/L . Although capillary blood glucose testing is not recommended for diagnosis of diabetes, it is possible that such result variations may lead to inappropriate management justifying why meters should not be used to diagnose diabetes.
The FDA and DIN EN ISO 17511 require that the measurement results of laboratory methods must be traceable to a reference material. In addition, it is recommended that the test results of blood glucose monitoring systems should be traced to the definitive reference method and the primary reference material. The Directive 98/79/EC on in vitro diagnostic medical devices also requires the calibration of in vitro diagnostic medical devices to be traceable to a reference method and/or to a reference material of higher metrological order . In addition, the JCTLM highlights that a suitable reference system consists of not only applying the definitive reference method and the reference material but also operating an accredited reference measurement laboratory .
A blood glucose monitoring system is only able to comply with the highest requirements for analytical quality if the glucose values measured using that system are traceable to the highest standard through an unbroken chain of comparative measurements. Isotope diluted gas chromatography/mass spectrometry (ID-GC/MS) is the internationally recognized definitive reference method for determining the mass concentration of glucose. NIST 917, the primary reference material of the highest metrological order, is used to calibrate the primary reference method .
This is followed by the hexokinase method as the secondary reference method. This method is recognized as a standard method. The hexokinase reference method is calibrated and monitored using calibrators and controls, whereby the target values are assigned by the ID-GC/MS method. It follows that blood glucose monitoring systems calibrated to predicated devices, such as the YSI glucose analyzer, do not guarantee trueness of measurement results.
It is a reasonable assumption to make that the performance requirements of glucose meters will be dependent on the intended clinical use of the meter. To date, no meter can match the analytical performance of laboratory methods. It is therefore not unreasonable to state that if a glucose meter does not match laboratory performance, it should not be used for diagnosis, yet there is considerable reliance on these meters for the management of diabetes globally. It is therefore important that meters are used for the purpose for which they are intended. Although the industry has the responsibility to continue improving the performance of meters, scientists and diabetes health professionals must take up the responsibility of ensuring that glucose meters used in clinical environments are “fit for purpose”. Often, the purchase of glucose meters is left to procurement officers who do not have the clinical or scientific skills to determine if the meter will fit the clinical use it is intended for.
It is also fair to assume that the reason professional societies have been challenged when attempting to set performance requirements of glucose meters is because they are used in a wide variety of clinical settings that have different performance requirements. It could be argued that devices should be categorized in relation to clinical use as having a minimum (clinically useable) and an optimal (guideline acceptable) performance requirement. A similar approach was suggested by Apple  for troponin testing and could equally be applied to glucose testing. Many devices may not achieve optimal requirement, but if they achieve minimum requirements, they could be safely used with care. Tables 1–3 summarize the basic information, interferences/operating restrictions, and important operational information that should be considered when selecting a new glucose meter.
|Test strip||Sample type||Sample volume, µL||Test principle||Chemistry||Test time, s||Test range||Individual strip package|
|Abbott Optium||Capillary||0.6||Electrochemical coulometric||GDH||20||20–500 mg/dL (1.1–27.8 mmol/L)||Yes|
|Accu-Chek Active||Capillary, Venous, Arterial, Neonate||2||Optical photometric||GDH||5||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|Accu-Chek Advantage/Sensor Comfort||Capillary, Venous, Arterial, Neonate||4||Electrochemical||GDH||5||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|Accu-Chek Comfort Curve||Capillary, Venous, Arterial||4||Electrochemical||GDH||26||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|Accu-Chek Go||Capillary, Venous, Arterial||1.5||Optical photometric||GDH||5||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|Accu-Chek Mobile Test Cassette||Capillary||0.3||Optical photometric||GDH-Mut PQQ||5||10–600 mg/dL (0.6–33.3 mmol/L)||Cassette|
|Accu-Chek Performa||Capillary, Venous, Arterial, Neonate||0.6||Electrochemical||GDH-Mut PQQ||5||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|Bayer Contour TS||Capillary||0.6||Electrochemical sensor||GDH||8||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|BGStarstrip||Capillary||0.5||Electrochemical||GOx||6||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|DiaCheck||Capillary||1||Electrochemical biosensor||GOx||5||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|EasyMate||Capillary||0.8||Electrode biosensor||N/A||6||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|FreeStyle Lite||Capillary, Venous||0.3||Electrochemical coulometric||GDH-FAD||5||20–500 mg/dL(1.1–27.8 mmol/L)||No|
|FreeStyle Papillon||Capillary, Venous||0.3||Electrochemical coulometric||GDH-PQQ||7||20–500 mg/dL(1.1–27.8 mmol/L)||No|
|Glucocard MX blood glucose test strips||Capillary||0.3||Electrochemical||FAD-GDH||5||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|GlucoMen LX Sensor||Capillary||0.3||Electrochemical||GOx||4||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|Glucosure AutoCode||Capillary||N/A||N/A||N/A||6||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|Match Okmeter||Capillary||0.7||Electrochemical biosensor||6||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|Nova||Capillary, Venous, Arterial, Neonate||1.3||Electrochemical biosensor||GOx||6||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|Nova Max Plus||Capillary||0.3||Electrochemical biosensor||GOx||5||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|On Call Chosen||Capillary||0.8||N/A||N/A||5||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|On Call Plus||Capillary||1||N/A||N/A||10||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|OneTouch Horizon||Capillary||1.5||Electrochemical biosensor||GOx||5||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|OneTouch Verio||Capillary, Venous||0.44||Electrochemical biosensor||FAD-GDH||5||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|SOLUS V2||Capillary||Amperometric||GOx||6||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|StatStrip Glucose Strip||Capillary, Venous, Arterial, Neonate||1.2||Electrochemical||GOx||6||10–600 mg/dL (0.6–33.3 mmol/L)||No|
|TRUEtest||Capillary, Venous||0.5||Electrochemical amperometric||GDH-PQQ||4||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|TRUEtrack||Capillary||1.0||Amperometric||GOx||10||20–600 mg/dL (1.1–33.3 mmol/L)||No|
|TysonBio AC100/200||Capillary||0.5||N/A||N/A||5||20–600 mg/dL (1.1-33.3 mmol/L)||No|
Information retrieved from package insert or company website.
|Test strip||Temperature, °C||Humidity, %||Haematocrit, %||Interferences||Procedure limitations|
|Abbott Optium||15–40||10–90||20–70||Xylose, ascorbic acid, high level of bilirubin, triglycerides and haemolysis||Not to be used with serum or plasma, during IV infusion of high-dose ascorbic acid or during xylose absorption testing.|
|Accu-Chek Active||10–40||<85||25–55||Galactose >7 mg/dL, maltose >20 mg/dL, ascorbic acid >3 mg/dL, triglycerides >5000 mg/dL, bilirubin >20 mg/dL has not been tested||Not to be used during a xylose absorption test, in peritoneal dialysis with icodextrin. With impaired peripheral circulation capillary blood collection is not advised|
|Accu-Chek Advantage/Sens or Comfort||14–40||<85||20–65 when blood glucose <200 mg/dL; 20–55 when blood glucose >200 mg/dL||Galactose >10 mg/dL, bilirubin (unconjugated) >20 mg/dL, triglycerides >5000 mg/dL, Maltose >16 mg/dL, acetaminophen >8 mg/dL, uric acid >10 mg/dL (hypoglycaemic range) >12 mg/dL (euglycaemic range) >16 mg/dL (hyperglycaemic range)||Not to be used during a xylose absorption test, in peritoneal dialysis with icodextrin. In situations of decreased peripheral blood flow fingerstick testing may not be appropriate|
|Accu-Chek Comfort Curve||14–40||<85||20–65 for glucose <200 mg/dL; 20–55 for glucose >200 mg/dL||Galactose >10 mg/dL, bilirubin (unconjugated) >20 mg/dL, triglycerides >5000 mg/dL, Maltose >16 mg/dL, acetaminophen >8 mg/dL, Uric acid >10 mg/dL (hypoglycaemic range) >12 mg/dL (euglycaemic range) >16 mg/dL (hyperglycaemic range)||Not to be used during a xylose absorption test, in peritoneal dialysis with icodextrin, with certain types of IVIG therapies (Octagam 5%), with IV solutions containing maltose for hydration. In situations of decreased peripheral blood flow fingerstick testing may not be appropriate|
|Accu-Chek Go||10–40||<85||25–65||Galactose >7 mg/dL, maltose >20 mg/dL, ascorbic acid >3 mg/dL, triglycerides >5000 mg/dL, bilirubin >20 mg/dL has not been tested||Not to be used during a xylose absorption test, in peritoneal dialysis with icodextrin. In situations of decreased peripheral blood flow capillary blood from the fingertip is not advised.|
|Accu-Chek Mobile||10–40||<85||25–55||Galactose >0.56mmol/L (21.6 mg/dL), ascorbic acid >0.17mmol/L (6.6 mg/dL)||With impaired peripheral circulation capillary blood collection is not advised|
|Accu-Chek Performa||8–44||10–90||10–65||Galactose >15 mg/dL, ascorbic acid >3 mg/dL||With impaired peripheral circulation capillary blood collection is not advised|
|Bayer Contour TS||5–45||10–93||0–70||Uric and ascorbic acids, acetaminophen, icodextrin don’t interfere. High lipid concentrations may interfere||Not suitable in severe dehydration or hypotension as well as hyperosmolar states. Not suitable for neonates|
|BGStarstrip||1–40||0–90||20–60||Substances found in the body naturally (e.g., Uric acid up to 23.5 mg/dL, bilirubin) or from therapeutic treatment (e.g., cetaminophen) should not significantly affect the test results. Results may be overestimated with abnormally high concentrations of ascorbic acid (vitamin C) >2 mg/dL (>0.11 mmol/L)||Not to be used to test glucose levels in newborns (children under 4 weeks) and in venous blood|
|DiaCheck||10–40||<90||30–55||Haemolysis, icterus, no effect of ascorbic acid, paracetamol, uric acid, tolazamide, urea, lipids, and maltose in normal concentrations||Not suitable for newborn babies|
|EasyMate||14–40||<85||30–55||Uric acid, galactose, xylose, methyldopa, dopamine, ascorbic acid, acetaminophen||Low and high haematocrit, not to be used near cordless phones or electronic equipment|
|FreeStyle Lite||4–30||5–90||15–65||Samples with cholesterol >500 mg/dL and triglycerides >3000 mg/dL should be treated with caution||Not to be used during a xylose absorption test, severe dehydration or excessive water loss may cause false low results. In situations of decreased peripheral blood flow fingerstick testing may not be appropriate|
|FreeStyle Papillon||15–65||5–90||15–65||Galactose ≥13 mg/dL, maltose ≥20 mg/dL, lactose ≥10 mg/dL. Samples with cholesterol >500 mg/dL and triglycerides >3000 mg/dL should be treated with caution||Not to be used during a xylose absorption test or in peritoneal dialysis with icodextrin. In situations of decreased peripheral blood flow fingerstick testing may not be appropriate|
|Glucocard MX||10–40||20–80||30–50||Icodextrin metabolites (maltose, maltotriose and maltotetraose) do not affect test results||Not available|
|GlucoMen LX Sensor||5–45||10–90||25–60||Elevated levels of acetaminophen, ascorbic acid, tolazamide, uric acid, bilirubin, ephedrine and methyldopa. Cholesterol up to 503 mg/dL (13 mmol/L) and triglycerides up to 333 mg/dL (8.6 mmol/L) do not affect results||Results may be falsely low in severe dehydration, severe hypertension, in shock or in a hyperosmolar state|
|Glucosure||10–40||<85||30–55||Not available||Not intended for use in neonates|
|Match Okmeter||10–40||10–85||20–60||Ascorbic and uric acid||Not intended for newborns|
|Nova||15–40||10–90||20–65||Not to be used with serum or plasma|
|Nova Max Plus||14–40||10–90||25–60||Acetaminophen, tolazamide, uric acid, bilirubin, ephedrine, methyldopa||Not intended for use in neonates or critically ill patients. Can’t be used to diagnose patients with DKA|
|On Call Chosen||5–45||10–90||N/A||Not available||Not available|
|On Call Plus||5–45||20–90||30–55||High levels of vitamin C, acetaminophen, uric acid, L-dopa, tolazamide||Not available|
|OneTouch Horizon||12–42||10–90||30–55||Abnormally high concentrations of acetaminophen, salicylates, uric acid, ascorbic acid and other reducing substances. Oxygen therapy and severely dehydrated, shocked or hyperosmolar patients may yield false low results. Cholesterol >700 mg/dL and triglycerides >3000 mg/dL are not recommended for testing||Not for testing newborns, specific to D-glucose and does not react with other sugars present|
|OneTouch Verio||6–44||10–90||20–60||Abnormally high concentrations of acetaminophen, uric acid, ascorbic acid and other reducing substances. Cholesterol >700 mg/dL l (18.1 mmol/L) and triglycerides >1311 mg/dL (33.9 mmol/L) are not recommended||Not for testing newborns, not to be used within 24h of a xylose absorption test, severe dehydration may cause false low results|
|SOLUS V2||10–40||20–80||32–56||Uric and ascorbic acids, acetaminophen, lipid levels ibuprofen, high||Affected by severe dehydration, not suitable for neonates or critically ill patients|
|StatStrip||15–40||10–90||20–65 (in-line measurement)||The StatStrip exhibits no interference from the following substances at known therapeutic levels: Acetaminophen, ascorbic acid, dopamine, ephedra, D(+)Galactose, ibuprofen, L-Dopa, Metyl-Dopa, salicylate, tetracycline, tolazamide and tolbudamide. No interference is observed from the following at or above the upper clinical normal range: bilirubin, cholesterol, creatinine, triglycerides and uric acid. No interference is observed from the following at normal therapeutic levels found in renal dialysis: D(+) maltose monohydrate, D(+)Maltotetraose and D(+) Maltotetriose. The StatStrip exhibits no interference in blood specimens with varying oxygen content||Fluoride or EDTA should not be used as preservatives for venous specimens|
|TRUEbalance||30–55||Abnormally high doses of acetaminophen||Not for testing blood flow in newborns or for use in patients with reduced peripheral blood flow|
|TRUEtest||15–30||10–90||20–60||Galactose >10 mg/dL, maltose >12.5 mg/dL, maltotriose >20 mg/dL, maltotetraose >10 mg/dL, uric acid ≥9 mg/dL at glucose levels <240 mg/dL||Peritoneal dialysis patients using icodextrin should not use these strips, or during a xylose infusion test. Capillary blood glucose levels in critically ill patients with reduced peripheral blood flow may not reflect the true physiological state|
|TRUEtrack||10–40||10–90||30–55||Abnormally high doses of acetaminophen||Not for testing blood flow in newborns or for use in patients with reduced peripheral blood flow|
|TysonBio AC100/200||10–40||10–90||35–55||Not available||Not suitable for newborn babies, affected by dehydration, hypotension, shock and hyperosmolar states|
Information retrieved from package insert or company website.
|Test strip||Calibration reference||Accuracy||QC recommendations|
|Abbott Optium||YSI glucose analyzer||Capillary results <75 mg/dL 100% within ±15 mg/dL, >75 mg/dL|
97% within ±20%. Total 97% within ±20%. Venous results
99% within ±20%. Arterial 97% within ±20%, Neonatal 96% ±20%
|MediSense or optium control solutions when you question a result and with each new strip lot. In multi-user setting, daily QC|
|Accu-Chek Active||Capillary de-proteinised blood using the Hexokinase method, traceable to a NIST standard confirmed by ID/MSMS||Mean systematic deviation max 4%. Glucose <75 mg/dL 100% within ±15 mg/dL, >75 mg/dL 99% within ±20%||With every new strip lot, when in doubt about the result, when you suspect the product may have been damaged; frequency depends on the number of glucose tests performed per day, week, month|
|Accu-Chek Advantage/Sensor Comfort||Capillary de-proteinised blood using the Hexokinase method, traceable to a NIST standard confirmed by ID/MSMS||97.5% of samples are within minimum acceptable performance criteria. Glucose <75 mg/dL 96.7% within ±15 mg/dL, glucose ≥75 mg/dL 97.6% within ±20%||With every new strip lot, when in doubt about the result, when you suspect the product may have been damaged; frequency depends on the number of glucose tests performed per day, week, month|
|Accu-Chek Comfort Curve||Capillary de-proteinised blood using the Hexokinase method, traceable to a NIST standard confirmed by ID/MSMS||97.5% of samples are within minimum acceptable performance criteria. Glucose <75 mg/dL 96.7% within ±15 mg/dL, glucose ≥75 mg/dL 95.9% within ±15%||Each time you open a new container of strips, if you left the strip container open and think it may have been damaged, if the strips were stored in extreme temperature or humidity, if you dropped your meter, your result doesn’t match the way you feel, if you want to check you are testing correctly|
|Accu-Chek Go||Mean systematic deviation max 4%. 100% of samples are within minimum acceptable performance criteria (±20%)||With every new strip lot, when in doubt about the result, when you suspect the product may have been damaged; frequency depends on the number of glucose tests performed per day, week, month|
|Accu-Chek Mobile||Mean systematic deviation max 4%. 100% of samples are within minimum acceptable performance criteria (±20%)|
|Accu-Chek Performa||99.5% of samples within minimum acceptable performance criteria. Glucose <75 mg/dL 100% within ±15 mg/dL, glucose ≥75 mg/dL 99.4% within ±20%|
|Bayer Contour TS||Referenced to plasma||Exceeds goals set by ISO15197||Requires no coding, controls provided, colour coded|
|BGStarstrip||YSI glucose analyser||Glucose <75 mg/dL 97% within ±10 mg/dL, 100% within ±15 mg/dL. Glucose=75 mg/dL 99% within ±20%, 98% within ±15% and 91% within ±10%||BGStar control solution to be used at each first meter use and in case of questionable results|
|DiaCheck||Referenced to plasma and calibrated to YSI 2300||<162 mg/dL (4.2 mmol/L), 97.5% results within 32 mg/dL (0.83 mmol/L)||Control solutions not provided with pack. USB port, memory function|
|EasyMate||Referenced to plasma||Not available||Control solutions provided, can analyse cholesterol and haemoglobin|
|FreeStyle Lite||YSI glucose analyser||Glucose <75 mg/dL 100% within ±15 mg/dL, glucose ≥75 mg/dL 99% within ±20%||When you question your results and want to confirm your meter and strips are working correctly|
|FreeStyle Papillon||YSI Model 2300 glucose analyser||Not available||Check meter and test strip performance regularly using FreeStyle control solution|
|Glucocard MX||YSI 2300||Glucose concentrations <75 mg/dL 100% capillary results ±10 mg/dL. Glucose concentrations >75 mg/dL 98% capillary results within ±20%||Perform a control test if you suspect the meter or test strips are not working properly, you dropped or damaged the meter, your test results do not match how you feel, you want to check the performance of the meter and test strips before a blood glucose test|
|GlucoMen LX Sensor||YSI Model 2300 glucose analyser||Glucose <162 mg/dL (4.2 mmol/L) 100% within ±32 mg/dL (0.83 mmol/L), glucose >162 mg/dL (4.2 mmol/L) 97% within ±20%||GlucoMen LX control solution ensures meter and strips are working correctly. Do a control before using the meter for the first time, each time you open a new vial of strips, if you drop your meter, when|
|glucose results are not consistent with how you feel or if you think the results are not accurate|
|GlucoSure AutoCode||Calibrated to display a plasma result||Not available||Contrex plus 111 glucose control solution|
|Match Okmeter||Referenced to plasma||32 mg/dL (0.83 mmol/L) at glucose conc <162 mg/dL (4.2 mmol/L) and within 20% glucose concentration >162 mg/dL (4.2 mmol/L)||Not available|
|Nova||Calibrated to display a plasma result||Not available||As required by your institutions quality control policy or local regulatory requirements|
|Nova Max Plus||Calibrated against plasma||6% or 11.6 mg/dL (0.3 mmol/L)||Nova max control solution. display screen with memory storage|
|On Call Chosen||Plasma equivalent, calibrated to YSI||Not available||Control solutions, test reminder, memory and insufficient sample indicator|
|On Call Plus||Calibrated to display a plasma result||Not available||On call plus glucose control solutions. Repeat unexpected results|
|OneTouch Horizon||YSI model 2300 glucose analyser||Not available||Do a control solution test (OneTouch Ultra control solution), refer to instructions for use|
|OneTouch Verio||The OneTouch® Verio™ System has demonstrated accuracy within ± 27 mg/dL (0.7 mmol/L) of the laboratory method for glucose concentrations lower than 170 mg/dL (4.4 mmol/L), and within ±15% of the laboratory method for glucose concentrations 170 mg/dL (4.4 mmol/L) or higher||OneTouch Verio control solutions. Test when you open a new bottle of strips, if you suspect the meter or strips are not working properly, if you have had repeated unexpected results, if you drop or damage your meter|
|Solus V2||Not available||Control solutions, memory for 500 tests, USB port for computers, audible commands|
|StatStrip||Plasma glucose hexokinase method||Glucose concentrations <75 mg/dL 100% adult ICU arterial and venous results ±15 mg/dL. Glucose concentrations >75 mg/dL 100% adult ICU arterial and venous results within ±20%||Not available|
|TRUEbalance||Calibrated to display a plasma result, YSI Model 23A||Not available||Automatic self-test performed each time a strip is inserted into the meter, TRUEcontrol solutions check testing technique and system performance. Perform a control if your result is unusually high or low or if it doesn’t match the way you feel|
|TRUEtest||YSI model 2300 glucose analyser||Glucose <75 mg/dL 100% within ±15 mg/dL, glucose ≥75 mg/dL 98–99.7% within ±15 mg/dL||Automatic self-test performed each time a strip is inserted into the meter, TRUEcontrol solutions check testing technique and system performance. Perform a control if your result is unusually high or low or if it doesn’t match the way you feel|
|TRUEtrack||YSI model 23A glucose analyser||TRUEtrack® System meets the ISO criteria for accuracy with 96.5% of results within the limits||Automatic self-test performed each time a strip is inserted into the meter, TRUEcontrol solutions check testing technique and system performance. Perform a control if your result is unusually high or low or if it doesn’t match the way you feel|
|TysonBio AC100/200||Not Available||Control solutions provided, voice function|
Information retrieved from package insert or company website. Illustrates the different of US and European IVD industries as US-manufacturers reference their glucose POCT meter to the YSI glucose analyzer from Life Sciences, rather than to an endorsed reference method.
Glucose meters: interferences
The type of interferences a glucose meter may exhibit is affected by the test strip technology and the detection method employed by the glucose meters.
Currently, the main enzymes used by glucose test strips are glucose oxidase (GOD) or glucose dehydrogenase (GDH). Both enzymes are coupled to a cofactor, such as flavin adenine dinucleotide (FAD), nicotinamide adenine dinucleotide (NAD), or pyrrolo-quinoline quinine (PQQ).
The glucose test strips have the respective enzyme dehydrated on the strip, and when blood is added to the strip, it rehydrates and reacts with the enzyme producing a product that can be detected by the meter. If the enzyme used is GO, glucose is oxidized to gluconic acid and hydrogen peroxide, whereas strips with GDH-PQQ enzyme oxidize glucose to gluconolactone and converts NAD to NADH [10, 16, 17]. Glucose concentration is detected either colorimetrically or amperometrically by the glucose meter.
To better understand why it is important to ensure that the meter being used is fit for purpose, one needs to understand what factors may influence the quality of glucose measurement obtained in different clinical situations. If these potential interferences are not taken into account, misleading results can be obtained leading to incorrect medications being administered, potentially resulting in hypoglycemia, coma, or even death. Table 2 outlines the potential interferences that need to be considered when selecting glucose devices for use in general medicine (hospital general medicine wards, general practice, and community health), intensive care unit, pediatric ward, and patient self-monitoring. When selecting a device for a particular clinical situation, it is important to consider what interferences will be common in the targeted patient population. This would assist in selecting the most appropriate device that best suits the intended population and clinical use.
Educating the user
The performance of glucose meters cannot be discussed without mentioning the importance of education for users of devices. How a meter is used has a significant influence on the accuracy of the result produced by the blood glucose meters. Patients often are not given appropriate education on how to use the system (meter, strip, and capillary collection) and teach themselves. This can result in a poor understanding of correct procedures when using meters such as washing hands, not using correct coding chip or following the manufacturer’s instructions, using expired test strips, and exposing strips to light and moisture [9, 10, 17, 18]. The studies exploring the effects of not correctly coding have shown that approximately 16% of patients in a typical endocrinology practice have miscoded their meters, leading to average errors of −37% to +29% . A study looking at the effect of washing hands before taking a capillary sample for glucose testing showed that if hands were not washed with water, erroneous glucose results were obtained. This effect was not nullified if hand washing was substituted with an alcohol swab. Also worrying was that a careful check of all instruction manuals for glucose devices in Japan failed to find recommendations for hand washing with water before testing for glucose .
This paper attempts to highlight that when selecting a glucose meter one needs to take into account what the clinical use of that meter will be and this will determine desirable analytical requirements and what potential interfering factors need to be considered. It is important to always remember that whether the glucose meter is for home use or for use in an intensive care unit, education is paramount in obtaining the best results possible. While education can address many pre-analytical issues that affect the quality of results, standardization and strip performance will require significant collaboration from both the manufacturers and clinical laboratory scientists to achieve the desired patient outcomes. The IFCC through its Point-of-Care Task Force and Working Group on Glucose Point-of-Care Testing comprised of both scientists and industry representatives are actively working toward these issues.
Conflict of interest statement
Authors’ conflict of interest disclosure: The authors stated that there are no conflicts of interest regarding the publication of this article.
Research funding: None declared.
Employment or leadership: Dr. George Koumantakis is representative of the manufacturers (Roche Diagnostics) among the WG and his presence did not influence the content of this paper.
Honorarium: None declared.
1. Clark LC Jr., Lyons C. Electrode systems for continuous monitoring in cardiovascular surgery. Ann NY Acad Sci 1962;102:29–45.Search in Google Scholar
2. Lunt H, Florkowski C, Bignall M, Budgen C. Capillary glucose meter accuracy and sources of error in the ambulatory setting. NZMJ 2010;123:1310.Search in Google Scholar
3. Dimeski G, Jones BW, Tilley V, Greenslade MN, Russell AW. Glucose meters: evaluation of the new formulation measuring strips from Roche (Accu-Chek) and Abbott (MediSense). Ann Clin Biochem 2010;47:358–65.10.1258/acb.2010.009291Search in Google Scholar PubMed
6. D’Orazio P, Burnett RW, Fogh-Andersen N, Jacobs E, Kuwa K, Külpman WR, etal. Approved IFCC recommendation on reporting results for blood glucose (abbreviated). Clin Chem 2005;51:1573–6.10.1373/clinchem.2005.051979Search in Google Scholar PubMed
7. D’Orazio P, Burnett RW, Fogh-Andersen N, Jacobs E, Kuwa K, Külpmann WR, et al. Approved IFCC recommendation on reporting results for blood glucose. Clin Chem Lab Med 2006;44:1486–90.10.1515/CCLM.2006.275Search in Google Scholar PubMed
9. Tonyushkina K, Nichols JH. Glucose meters: a review of technical challenges to obtaining accurate results. J Diabetes Sci Technol 2009;3:971–80.10.1177/193229680900300446Search in Google Scholar PubMed PubMed Central
10. Kristensen GB, Sandberg S. Self-monitoring of blood glucose with a focus on analytical quality: an overview. Clin Chem Lab Med 2010;48:2–10.Search in Google Scholar
11. World Health Organization. Definition and diagnosis of diabetes mellitus and intermediate hyperglycaemia: report of a WHO/IDF consultation. Geneva (Switzerland): WHO, 2006.Search in Google Scholar
12. Directive 98/79/EC of the European Parliament and of the Council of 27 October 1998 on in vitro diagnostic medical devices OJ L 331 of 7 December 1998. http://ec.europa.eu/enterprise/policies/european-standards/harmonised-standards/iv-diagnostic-medical-devices/index_en.htm. Accessed 10 September 2012.Search in Google Scholar
16. Pitkins AD, Rice MJ. Challenges to glycemic measurement in the perioperative and critically ill patient: a review. J Diabetes Sci Technol 2009;3:1270–81.10.1177/193229680900300606Search in Google Scholar PubMed PubMed Central
17. Dungan K, Chapman J, Braithwaite SS, Buse J. Glucose measurement: confounding issues in setting targets for inpatient management. Diabetes Care 2007;30:403–9.10.2337/dc06-1679Search in Google Scholar PubMed
18. Ginsberg GH. Factors affecting blood glucose monitoring: sources of errors in measurement. J Diabetes Sci Technol 2009;3:903–13.10.1177/193229680900300438Search in Google Scholar PubMed PubMed Central
19. Hirose T, Tomoya M, Yoshio Y, Kawamori R, Hirotaka W. Glucose monitoring after fruit peeling: pseudohyperglycaemia. Diabetes Care 2011;34:596–7.10.2337/dc10-1705Search in Google Scholar PubMed PubMed Central
©2013 by Walter de Gruyter Berlin Boston