Studies Clinical Studies
I.Glucose meters Principles

1) Past technology

First Generation: Optical Biosensor → chemical reactants on the strip will interact with glucose in the blood and the color showed will be detected by a meter, (e,g, Superb and Evolve)

2) Actual technology

Second Generation: Electrochemical Biosensor → the applied blood on the strip will be drawn in and dissolve in reaction zone, where enzyme, chemical mediators, and stabilizers are located Biochemical reactions will convert oxidation/reduction potential into mediator_ Then the meter will measure the electiical currency, which depends on the quantity of glucose level found in sample and result will be displayed on the LCD monitor atter calculation. (Most current available meters).

II.BGM Facts

1) Accuracy In the medical community a home blood glucose meter's test result is considered clinically accurate if it falls within +/-20% of the EGA analysis of an accepted reference result, usually a lab test What that means is, the only way to be certain that a meter is accurate is to compare the result (using the same blood sample) with that of a lab test.
2) Bias
Low concentration Bias < 75 (mg/dL) I strip test result value - standard result value
High concentration Bias (%) = (strip test result value - standard result value)/ standard result value)x100%

e.g. TB100 test result =120mg/mg/dL
Bias: (120/100)/1o0=20%

3) International Standard :ISO 15197:2003
ISO 15197:2003 specifies requirements for in vitro glucose monitoring systems that measures glucose concentrations in capillary blood samples and procedures for the verification and the validation of performance by the intended users. These systems are intended for self-testing by lay persons for management of diabetes mellitus.

For 95% of the individual glucose result shall fall within:
± 15 mg/dL at BG < 75mg/dL
± 20% at BG ≥ 75mg/dL
Both FDA and CE follow this regulation
e.g.Table below shows example of different distribution of fingertip YSI with 128 pateints
*Blood Glucose Concentration < 75mg/dL

Finger Difference within ± 5mg/dL Difference within ± 10mg/dL Difference within ± 15mg/dL
Lay user 7 / 12 (58%) 12 / 12 (100%) 12 / 12 (100%)
Technician 8 / 19 (42%) 18 / 19 (95%) 18 /19 (95%)
*Blood Glucose Concentration < 75mg/dL

Finger Difference within 5% Difference within 10% Difference within 15% Difference within 20%
Lay user 49 / 100 (49%) 81 / 100 (81%) 94 / 100 (94%) 98 / 100 (98%)
Technician 55 / 105 (52%) 86 / 105 (82%) 99 / 105 (94%) 105 / 105 (100%)

4) EGA (Error)
First develped in 1987
Measures the difference in results between meter and clinical value.
This method of comparing glucose determinations allows for the separation of pairs of data into zones that suggest different levels of clinical concern and urgency of intervention.
(a) Z one A - Clinical accurate. The values that deviate from the reference by < 20% or values that are in the hypoglycemic range (< 3.89 mmol/l when the reference value is also < 3.89 mmol/l.
(b) Zone B - Acceptable results. i.e., values that deviate by > 20% but would lead to no benign or no treatment.
(c) Zone C Over - correction of glucose level, deviating from the reference method by more than 20% and would lead to unnecessary corrective treatment
(d) Zone D Dangerous failure to detect and treat. Potentially dangerous failure to detect and treat blood glucose levels outside of desired target range.
(e) Zone E Erroneous treatment. Resulting in erroneous treatment.

Total Points A B C D E A+B
124 124 0 0 0 0 124
% 100% 0% 0% 0% 0% 100%
A A-(+/-15-20%) A+(+/- 15%)
points % points % points %
124 100% 3% 2.4% 121 97.6%

5) Precision To be considered precise, a given meter must be able to consistently produce the same results. it the result is close after doing 20 tests then it is precise. A meter can be precise but not accurate but cannot be accurate without being precise.

6) Refer to meter's test value as (y), and standard value as (x). Every point is composed of (X, Y), after conducting linear regression, we obtain a straight line Y=Ax+b, that is equals to a decision coefficient R2 value. When the meter value equals to standard value, then this can be expressed as Y=R, R2=1.0000

R2 refers to the relation between these points (X, Y) and the regression line. Maximum value is 1, most manufacturers results are around 0.97 while TysonBio is around 0.98-0.99.

Whole blood= plasma + blood cells. After centrifugation, the upper layer is plasma, lower layer are blood cells.HT (Hematocrit) is the proportion of blood volume that is occupied by red blood cells. lt is normally about 48% for men and 38% for women.

The glucose concentiation in the plasma is about 1.10 - 1.12 times of that of the whole blood. Results of A and B, depends on set criteria

Assumptions: refers to the whole blood concentration of blood glucose as standard (taking samples of whole blood from the finger tip and performing standardize test). The meter regression results on the standard instrument are assumed to be y = x, meaning that Meter test value equals to the concentration of whole blood
2013 Ashish Healthcare all rights reserved.BADRI Computers