This is the text extract for Evaluation of blood glucose meters, browse documents here.

---

Meter Evaluations

Background:

Formal evaluation of meter performance was undertaken to assess the accuracy and precision of blood glucose meter systems and their strips. The systems studied are those which are funded by the Government in New Zealand (via a subsidy on the Pharmaceutical Schedule). Three separate studies were completed to assess the funded blood glucose meters. The following studies were completed by the Christchurch Diabetes Centre: Study 1: Accu-Chek Performa and Optium Xceed (5 second test strip): Completed Q2 20081 Study 2: Care Sens II, Care Sens POP, FreeStyle Lite: Completed Q2 2009 Study 3: On Call Advanced: Completed Q3 2009

Methods:

For each meter tested, simultaneously collected venous plasma and capillary blood samples were taken from at least 50 outpatients attending a diabetes centre. Capillary glucose was measured on at least two meters of each brand. Venous plasma glucose was measured by the laboratory hexokinase method. Method comparison was undertaken using the venous sample as reference standard, using Bland–Altman plots, Passing and Bablok regression analysis, and both Clarke and Consensus error grid analyses. The paper referenced below 1 provides full details of the methodology.

Results:

The key results from the meter evaluations are displayed visually on the following pages using both Clarke and Consensus error grid analyses. All meters performed satisfactorily with all results falling within the A and B zones (see below for clinical interpretation of these zones). The venous plasma sample (i.e. reference value) is displayed along the x axis. The corresponding mean capillary value is displayed along the y axis. The error grids assign a specific level of clinical risk to any possible capillary glucose error as follows: A - <20% deviation B - deviation that leads to no change in treatments C - overcorrection of an acceptable glucose level D - dangerous failure to detect and treat abnormal glucose levels E - erroneous treatment The zones are defined based on surveys of specialist medical practitioners. The only difference between the Clarke Diagram and the Consensus Diagram is in how the error zones are defined. Other test results which are not shown here, such as measurement of imprecision, showed no significant concerns.

Conclusions:

The authors of the study concluded that, in a controlled setting with testing undertaken by trained researchers, all meters performed to a standard that is acceptable for operational use.

1

Comparison of blood glucose meters in a New Zealand diabetes centre Florkowski et al. Ann Clin Biochem.2009; 46: 302-305 (http://acb.rsmjournals.com/cgi/content/full/46/4/302)

---

Error Grid Analysis

Measured Value - capillary glucose (mmol/L)

Accu-Chek Performa

Clarke grid

Measured Value - capillary glucose (mmol/L)

30

20

A 99% B 1% C 0% D 0% E 0%

30

20

A 99% B 1% C 0% D 0% E 0%

10 Roche Performa 0

0

5

10

15

20

25

30

Consensus grid 10 Roche Performa 0

Measured Value - capillary glucose (mmol/L)

Reference Value - plasma venous glucose (mmol/L)

30

0

5

10

15

20

25

30

20

A 99% B 1% C 0% D 0% E 0%

Reference Value - plasma venous glucose (mmol/L)

Measured Value - capillary glucose (mmol/L)

30

20

A 99% B 1% C 0% D 0% E 0%

10

Roche Performa

0

0

5

10

15

20

25

30

Reference Value - plasma venous glucose (mmol/L)

10

Roche Performa

Error Grid Analysis

Measured Value - capillary glucose (mmol/L)

CareSens II

0 0 5 10 15 Clarke grid

Measured Value - capillary glucose (mmol/L)

20

25

30

Reference Value - plasma venous glucose (mmol/L)

30

20

A 96% B 4% C 0% D 0% E 0%

30

20

A 96% B 4% C 0% D 0% E 0%

10 Care Sens II 0 0 5 10 15 20 25 30

Consensus grid 10 Care Sens II 0 0 5 10 15 20 25 30

Measured Value - capillary glucose (mmol/L)

Reference Value - plasma venous glucose (mmol/L)

30

20

A 94% B 6% C 0% D 0% E 0%

Reference Value - plasma venous glucose (mmol/L)

- capillary glucose (mmol/L)

30

20

A 94% B 6% C 0% D 0% E 0%

10

Care Sens II

0

0

5

10

15

20

25

30

Reference Value - plasma venous glucose (mmol/L)

---

Error Grid Analysis

Measured Value - capillary glucose (mmol/L)

CareSens POP

Clarke grid

Measured Value - capillary glucose (mmol/L)

30

20

A 98% B 2% C 0% D 0% E 0%

30

20

A 98% B 2% C 0% D 0% E 0%

10 Care Sens POP 0 0 5 10 15 20 25 30

Consensus grid 10 Care Sens POP 0 0 5 10 15 20 25 30

Measured Value - capillary glucose (mmol/L)

Reference Value - plasma venous glucose (mmol/L)

30

20

A 96% B 4% C 0% D 0% E 0%

Reference Value - plasma venous glucose (mmol/L)

Measured Value - capillary glucose (mmol/L)

30

20

A 96% B 4% C 0% D 0% E 0%

10

Care Sens POP

0

0

5

10

15

20

25

30

Reference Value - plasma venous glucose (mmol/L)

10

Care Sens POP

Error Grid Analysis

Measured Value - capillary glucose (mmol/L)

FreeStyle Lite

0 0 5 10 15 Clarke grid

Measured Value - capillary glucose (mmol/L)

20

25

30

Reference Value - plasma venous glucose (mmol/L)

30

20

A 98% B 2% C 0% D 0% E 0%

30

20

A 98% B 2% C 0% D 0% E 0%

10 Freestyle 0 0 5 10 15 20 25 30

Consensus grid 10 Freestyle 0 0 5 10 15 20 25 30

Measured Value - capillary glucose (mmol/L)

Reference Value - plasma venous glucose (mmol/L)

30

20

A 100% B 0% C 0% D 0% E 0%

Reference Value - plasma venous glucose (mmol/L)

- capillary glucose (mmol/L)

30

20

A 100% B 0% C 0% D 0% E 0%

10

Freestyle

0

0

5

10

15

20

25

30

Reference Value - plasma venous glucose (mmol/L)

---

Error Grid Analysis

Measured Value - capillary glucose (mmol/L)

On Call Advanced

Clarke grid

Measured Value - capillary glucose (mmol/L)

30

20

A 98% B 2% C 0% D 0% E 0%

30

20

A 98% B 2% C 0% D 0% E 0%

10 On Call Advanced 0 0 5 10 15 20 25 30

Consensus grid 10 On Call Advanced 0 0 5 10 15 20 25 30

Measured Value - capillary glucose (mmol/L)

Reference Value - plasma venous glucose (mmol/L)

30

20

A 100% B 0% C 0% D 0% E 0%

Reference Value - plasma venous glucose (mmol/L)

Measured Value - capillary glucose (mmol/L)

30

20

A 100% B 0% C 0% D 0% E 0%

10 On Call Advanced 0

0

5

10

15

20

25

30

Reference Value - plasma venous glucose (mmol/L)

10 On Call Advanced 0

Error Grid Analysis

Measured Value - capillary glucose (mmol/L)

Optium Xceed

0 5 10 15 20 Clarke grid

Measured Value - capillary glucose (mmol/L)

25

30

Reference Value - plasma venous glucose (mmol/L)

30

20

A 98% B 2% C 0% D 0% E 0%

30

20

A 98% B 2% C 0% D 0% E 0%

10 Optium Xceed 0 0 5 10 15 20 25 30

Consensus grid 10 Optium Xceed 0 0 5 10 15 20 25 30

Measured Value - capillary glucose (mmol/L)

Reference Value - plasma venous glucose (mmol/L)

30

20

A 100% B 0% C 0% D 0% E 0%

Reference Value - plasma venous glucose (mmol/L)

- capillary glucose (mmol/L)

30

20

A 100% B 0% C 0% D 0% E 0%

10 Optium Xceed 0

0

5

10

15

20

25

30

Reference Value - plasma venous glucose (mmol/L)