Clinical Chemistry 45: 1821-1825, 1999;
(Clinical Chemistry. 1999;45:1821-1825.)
© 1999 American Association for Clinical Chemistry, Inc.
Newer Portable Glucose MetersAnalytical Improvement Compared with Previous Generation Devices?
Raimund Weitgassera,
Brigitta Gappmayer and
Maximilian Pichler
2nd Department of Medicine, St. Johanns Spital, Salzburg General Hospital, Muellner Hauptstrasse 48, A-5020 Salzburg, Austria.
a Author for correspondence: Fax 43-662-651674; e-mail R.Weitgasser{at}lkasbg.gv.at
 |
Abstract
|
|---|
Background: Newer glucose meters are easier to use, but direct
comparisons with older instruments are lacking. We wished to compare
analytical performances of four new and four previous generation
meters.
Methods: On average, 248 glucose measurements were performed with
two of each brand of meter on capillary blood samples from diabetic
patients attending our outpatient clinic. Two to three different lots
of strips were used. All measurements were performed by one experienced
technician, using blood from the same sample for the meters and the
comparison method (Beckman Analyzer 2). Results were evaluated by
analysis of clinical relevance using the percentage of values within a
maximum deviation of 5% from the reference value, by the method of
residuals, by error grid analysis, and by the CVs for measurements in
series.
Results: Altogether, 1987 blood glucose values were obtained with
meters compared with the reference values. By error grid analysis, the
newer devices gave more accurate results without significant
differences within the group (zone A, 9898.5%). Except for the One
Touch II (zone A, 98.5%), the other older devices were less exact
(zone A, 8792.5%), which was also true for all other evaluation
procedures.
Conclusions: New generation blood glucose meters are not only
smaller and more aesthetically appealing but are more accurate compared
with previous generation devices except the One Touch II. The
performance of the newer meters improved but did not meet the goals of
the latest American Diabetes Association recommendations in the hands
of an experienced operator.
 |
Introduction
|
|---|
Self-monitoring of blood glucose
(SMBG)1
is widely used because intensive insulin therapy has become a
standard treatment regimen in type 1 diabetic patients
(1)(2)(3) and recommendations for type 2 treatment are being
newly structured (4), with increases in SMBG expected.
Besides patients, nurses and technicians increasingly use portable
glucose meters for bedside glucose measurements in hospitals and for
in-home patient care. Because bloodless glucose measurement is still
restricted to research and is far from clinically routine
(5)(6), new, smaller, and fast-acting portable
glucose meters are being developed. The criteria for clinical
evaluation of glucose meters have been improved
(7)(8), and the performance of meters is
reported regularly (9)(10)(11)(12). To determine whether new
generation glucose meters perform as well as or better than previously
developed devices, we compared commonly used newer with older meters.
 |
Materials and Methods
|
|---|
Capillary blood samples were taken at room temperature
(~20 °C) from type 1 and type 2 diabetic patients attending our
outpatient clinic; these samples represented blood glucose values from
2.2 to 22.2 mmol/L to cover all clinically relevant ranges. Blood
glucose was measured with two devices from the following brands: for
the "old" meters, the Accutrend® [Europe
(Eu)] or Accu-Chek®
EasyTM [United States (US); Boehringer Mannheim,
Mannheim, Germany], the CompanionTM 2 (Eu
and US; MediSense, Birmingham, UK, and Cambridge, MA), the
Glucometer® 3 (Eu and US; Bayer Diagnostics,
Munich, Germany, and Miles Diagnostics Division, Elkhart, IN), and the
One Touch® II (Eu and US; LifeScan, Johnson &
Johnson, Milpitas, CA); for the "new" meters, the
GlucocardTM (Eu) or Glucometer
Elite® (US; Menarini Diagnostics, Florence,
Italy, and Bayer Diagnostics), the Glucometer
EspritTM (Eu) or Glucometer
DexTM (US; Bayer Diagnostics), the
GlucotouchTM (Eu) or
SureStep® (US; LifeScan, Johnson & Johnson), and
the Glucotrend® (Eu) or Accu-Chek
InstantTM (US; Roche Diagnostics, Mannheim,
Germany). Three meters use electronic sensor techniques (Companion 2,
Glucometer Esprit, and Glucocard), the other meters use reflectance
techniques. A total of 1987 measurements (on average, 248 measurements
for each brand) were performed using the same blood sample for
comparison with our glucose oxidase method (comparison method)
performed on a Beckman Analyzer 2 (Beckman). The two devices of each
brand were regularly exchanged with each other after every 510
measurements. Two to three different lots of strips were used for
measurements with each brand of meter. Calibration was performed daily
with strips or calibration solutions. All measurements were performed
according to the manufacturers' recommendations by the same
experienced technician. Measurements <2.2 and >22.2 mmol/L as well as
those that were indicated as "low" or "high" by the devices
were excluded from evaluation. We also excluded samples with hematocrit
values <30% and >60%.
statistics
Values were analyzed for clinical relevance by determination of
the percentage of values within a maximum deviation of 5% from the
reference value, according to recommendations by the American Diabetes
Association (ADA) (1)(2). In addition, the error
grid analysis method of Clarke and co-workers
(7)(8) was used to assess accuracy. The error
grid defines the x-axis as the reference blood glucose and
the y-axis as the value generated by the glucose meter. The
graphic model describes clinically relevant deviations using
asymmetrically arranged areas for glucose ranges between 3.9 and 22.2
mmol/L. The agreement between glucose meter values and reference
glucose values is expressed by different zones and thus gives the
accuracy of the meters: zone A, clinically accurate; zone B, clinically
irrelevant deviation by >20% from the reference; zone C, unnecessary
overcorrection possible; zone D, "dangerous failure to detect and
treat" errors; and zone E, "erroneous treatment" danger. To
assess the overall deviation of the devices, we also calculated the
mean (SD) difference from the Beckman glucose oxidase results
(13). To determine within-run precision, the CVs for 10
measurements in series were calculated for three different clinically
relevant blood glucose ranges: 2.93.9 mmol/L, 9.110 mmol/L, and
1515.7 mmol/L.
 |
Results
|
|---|
The percentage of values within a maximum deviation of 5% from
the reference value as recommended by the most recent ADA conference on
SMBG is shown in Table 1
. The newer meters (Glucocard, Glucometer Esprit, Glucotouch,
and Glucotrend) were significantly better when compared with previous
generation meters (Accutrend, Companion 2, Glucometer 3, and One Touch
II). The only exception for the older devices was the One Touch II,
which performed at least as well as the Glucometer Esprit and the
Glucocard. None of the devices reached the ADA recommendations of 100%
of readings within a 5% deviation limit. The differences (mean ±
SD) between meter-generated results and the values measured with our
comparison method are shown in Table 2
. The Glucometer Esprit, Glucotouch, and Glucotrend slightly
overestimated (positive mean percentage) and the Glucocard slightly
underestimated (negative mean percentage) the "true" comparison
glucose values. For the previous generation meters, this method of
comparison gives much higher differences: overestimation of
>10% for the Glucometer 3 and underestimation of >10% for the
Companion 2.
The error grid analysis is shown in Fig. 1
. For all newer meters, 98% of values were within zone A and
100% were within zones A + B compared with the Accutrend, Companion 2,
and Glucometer 3, which gave a few values in "risk zones" C, D, and
E. Glucometer 3 had estimations in a low glucose range (<4 mmol/L) in
zone E and in a high glucose range (>13 mmol/L) in zone C, Accutrend
and Companion 2 gave values only in a high glucose range (>13 mmol/L)
in zone D. Again, One Touch II was an exception among the older
devices, performing similar to the new generation meters.

View larger version (36K):
[in this window]
[in a new window]
|
Figure 1. Error grid analysis for each blood glucose meter.
Readings of glucose meters are plotted against values from the
comparison method. The agreement of glucose meter values and reference
glucose values is expressed by different zones and thus gives the
accuracy of the meters: zone A, clinically accurate;
zone B, clinically irrelevant deviation by >20% from
the reference; zone C, unnecessary overcorrection
possible; zone D, dangerous failure to detect and treat
errors; zone E, erroneous treatment.
Panels a-d are older devices;
panels e-h are newer devices. a,
Accutrend; b, Companion 2; c, Glucometer
3; d, One Touch II; e, Glucocard;
f, Glucometer Esprit; g, Glucotouch;
h, Glucotrend.
|
|
The CVs for measurement in series, which were used to define precision
of the devices, are shown in Table 3
. The Glucometer Esprit performed the worst, especially with
respect to low glucose ranges, where only the old meter Companion 2
showed even more dispersion.
 |
Discussion
|
|---|
In this study, we compared four new generation portable glucose
meters with four previous generation meters with respect to accuracy
and precision. According to recommendations in previous studies
(7)(8)(14), we chose brands of
glucose meters that are used frequently in clinical and outpatient
care. For our evaluation, we did not use statistical methods, such as
linear regression analysis, that have only limited value for clinical
evaluation of glucose meters. We chose the error-grid analysis as one
of the clinically most relevant approaches
(12)(14). With the newer glucose meters, 100%
of measurements were within error grid zone A (accurate zone) and zone
B (clinically irrelevant deviation). For the older meters, this was
true only for the One Touch II; the other meters of this generation
gave a few values in zones C, D, and E, making clinically incorrect
decisions based on the measured values at least possible. Because
error-grid analysis represents only one analytical view, we added a set
of analyses according to recommendations of the ADA. According to these
recommendations, the accuracy of measurements is expressed by the
percentage of deviations from the reference value. A previously
recommended deviation of within 10% for 100% of measurements was
recently replaced by recommendations for a target variability of <5%
(2). Although only 4956% of the values measured with the
new meters met the new ADA criteria, improvement was significant: the
older meters reached these target values in
25% of
measurements. The reason for this improved performance by the
newer glucose meters is probably attributable to both technical
improvements in the devices and the reduced blood volumes necessary for
measurement. The newer meters need only 35 µL compared with 1050
µL for previous systems, which makes mistakes in application of blood
drops to test strips unlikely. Strips such as those used for the
Glucocard take up only a limited amount of blood for measurement. These
advantages are combined with fast measurement within 2060 s; memory
function for up to 300 measurements; and smaller, more aesthetically
appealing devices. However, only the Glucotouch is still equipped with
test strips for visible control, which may help detect meter
dysfunction.
We found no substantial difference with respect to the technical
equipment of the tested meters using either reflectance or electronic
sensor technique. Whether additional new techniques will help meet the
goals of the latest ADA criteria is thus questionable. To justify these
stringent criteria, one must also be aware of a broad variability in
the skill of users. SMBG, meanwhile, is performed by so many patients
and healthcare personnel in various settings that user errors may
impair results in daily practice despite improvement in analytical
performance. As implemented by the Diabetes Control and Complications
Trial (3) and the United Kingdom Prospective Diabetes Study
(15), current standards for diabetes care (4)
include increasing the frequency of SMBG by an increasing number of
intensively treated type 1 and type 2 diabetic patients. Therefore, in
addition to improvements in technical accuracy, appropriate training of
the patients and healthcare personnel using glucose meters is the a
mainstay of well-established SMBG. In addition to regular calibration
and maintenance of meters, frequent comparison of function with values
obtained by a reference laboratory method seems advisable.
In summary, the performance of the newer portable glucose meters
when clinically assessed under laboratory conditions was substantially
improved compared with all previous generation devices except for the
One Touch II. These results may be extrapolated but have yet to be
demonstrated in daily use by patients and healthcare personnel. Until
reliable noninvasive blood glucose measurement methods are available
for everyday clinical use, further improvement of currently available
devices to meet the ADA standards is necessary.
 |
Acknowledgments
|
|---|
Blood glucose meters and strips were kindly provided by the
manufacturers.
 |
Footnotes
|
|---|
1 Nonstandard abbreviations: SMBG, self-monitoring of blood glucose; Eu, Europe; US, United States; and ADA, American Diabetes Association. 
 |
References
|
|---|
-
American Diabetes Association. Consensus statement
on self-monitoring of blood glucose. Diabetes Care 1987;10:959..
-
American Diabetes Association. Self-monitoring of blood
glucose consensus statement. Diabetes Care 1996;19(Suppl 1):S626..
-
The Diabetes Control and Complications Trial Research
Group. The effect of intensive treatment of diabetes on the development
and progression of long-term complications in insulin-dependent
diabetes mellitus. N Engl J Med 1993;329:97786..
-
American Diabetes Association. Standards of medical care
for patients with diabetes mellitus. Diabetes Care 1999;22(Suppl
1):S3241..
-
Klonoff DC. Noninvasive blood glucose monitoring. Diabetes Care 1997;20:433-437.
[Abstract]
-
Khalil OS. Spectroscopic and clinical aspects of noninvasive glucose measurements. Clin Chem 1999;45:165-177.
[Abstract/Free Full Text]
-
Clarke WL, Cox D, Gonder-Frederick LA, Carter W, Pohl SL. Evaluating clinical accuracy of systems for self-monitoring of blood glucose. Diabetes Care 1987;10:622-628.
[Abstract]
-
Cox DJ, Gonder-Frederick LA, Kovatchev BP, Julian DM, Clarke WL. Understanding error grid analysis. Diabetes Care 1997;20:911-912.
[ISI][Medline]
[Order article via Infotrieve]
-
Chan JC, Wong RY, Cheung CK, Lam P, Chow CC, Yeung VT, et al. Accuracy, precision and user-acceptability of self blood glucose monitoring machines. Diabetes Res Clin Pract 1997;36:91-104.
[ISI][Medline]
[Order article via Infotrieve]
-
Weitgasser R, Gappmayer B. New portable devices for blood glucose measurement: comparison of MediSense Card, Glucometer Elite, Hypocount Supreme and Omnican control with the reference glucose oxidase method. Schweiz Med Wochenschr 1997;127:12-17.
[Medline]
[Order article via Infotrieve]
-
Brunner GA, Ellmerer M, Sendlhofer G, Wutte A, Trajanoski Z, Schaupp L, et al. Validation of home blood glucose meters with respect to clinical and analytical approaches. Diabetes Care 1998;21:585-590.
[Abstract]
-
Poirier J-Y, Le Prieur N, Campion L, Guilhem I, Allannic H, Maugendre D. Clinical and statistical evaluation of self-monitoring blood glucose meters. Diabetes Care 1998;21:1919-1924.
[Abstract]
-
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;i:307-310.
-
Dedrick RF, Davis WK. What do statistics really tell us about the quality of the data from self-monitoring of blood glucose?. Diabet Med 1989;6:267-273.
[Medline]
[Order article via Infotrieve]
-
UKPDS Group. Intensive blood-glucose control with
sulfonylureas or insulin compared with conventional treatment and the
risk of complications in patients with type diabetes (UKPDS 33). Lancet
1998;352:83753..
The following articles in journals at HighWire Press have cited this article:

|
 |

|
 |
 
A. G. Niehoff, T. W. van Haeften, N. C. Onland-Moret, C. C. Elbers, C. Wijmenga, and Y. T. van der Schouw
C-Reactive Protein Is Independently Associated With Glucose but Not With Insulin Resistance in Healthy Men
Diabetes Care,
June 1, 2007;
30(6):
1627 - 1629.
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
K. Dungan, J. Chapman, S. S. Braithwaite, and J. Buse
Glucose Measurement: Confounding Issues in Setting Targets for Inpatient Management
Diabetes Care,
February 1, 2007;
30(2):
403 - 409.
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
W. L. Clarke, S. Anderson, L. Farhy, M. Breton, L. Gonder-Frederick, D. Cox, and B. Kovatchev
Evaluating the Clinical Accuracy of Two Continuous Glucose Sensors Using Continuous Glucose-Error Grid Analysis
Diabetes Care,
October 1, 2005;
28(10):
2412 - 2417.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
A. D M Stork, H. Kemperman, D W. Erkelens, and T. F Veneman
Comparison of the accuracy of the hemocue glucose analyzer with the Yellow Springs Instrument glucose oxidase analyzer, particularly in hypoglycemia
Eur. J. Endocrinol.,
August 1, 2005;
153(2):
275 - 281.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
D. C. Klonoff
The Need for Separate Performance Goals for Glucose Sensors in the Hypoglycemic, Normoglycemic, and Hyperglycemic Ranges
Diabetes Care,
March 1, 2004;
27(3):
834 - 836.
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
P. Bohme, M. Floriot, M.-A. Sirveaux, D. Durain, O. Ziegler, P. Drouin, and B. Guerci
Evolution of Analytical Performance in Portable Glucose Meters in the Last Decade
Diabetes Care,
April 1, 2003;
26(4):
1170 - 1175.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
R. Amin, K. Ross, C. L. Acerini, J. A. Edge, J. Warner, and D. B. Dunger
Hypoglycemia Prevalence in Prepubertal Children With Type 1 Diabetes on Standard Insulin Regimen: Use of Continuous Glucose Monitoring System
Diabetes Care,
March 1, 2003;
26(3):
662 - 667.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
S. Skeie, G. Thue, K. Nerhus, and S. Sandberg
Instruments for Self-Monitoring of Blood Glucose: Comparisons of Testing Quality Achieved by Patients and a Technician
Clin. Chem.,
July 1, 2002;
48(7):
994 - 1003.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
D. B. Sacks, D. E. Bruns, D. E. Goldstein, N. K. Maclaren, J. M. McDonald, and M. Parrott
Guidelines and Recommendations for Laboratory Analysis in the Diagnosis and Management of Diabetes Mellitus
Clin. Chem.,
March 1, 2002;
48(3):
436 - 472.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
J. C. Boyd and D. E. Bruns
Quality Specifications for Glucose Meters: Assessment by Simulation Modeling of Errors in Insulin Dose
Clin. Chem.,
February 1, 2001;
47(2):
209 - 214.
[Abstract]
[Full Text]
[PDF]
|
 |
|