Biomedical Engineering Reference
In-Depth Information
The results indicate that the linear ranges of glucose and lactate (oxidase
reactions) in whole blood correspond well with those of the same metabolites in
buffers. The sensor methods and the reference methods were in good correla-
tion for all analytes. The precision for the standards in buffers was always better
(about 2 - 3%) than that of the blood samples. This was ascribed in part to the
instability of the metabolite concentrations in blood, particularly that of lactate.
In addition, the blood viscosity and the nonspecific heat in the reaction can also
affect the final results.
The primary features of this method are its general principle, a uniform mea-
surement system, the use of untreated blood samples, minute sample volumes,
no fouling of the transducers, no electrochemical or optical interference, simple
procedures, rapid response, and low cost. According to the working principle,
other metabolites associated with enzyme reactions - in addition to glucose,
urea and lactate - can be analyzed in a similar way. Unlike electrochemical and
optical detection, where the potential or wavelength must be adapted to a speci-
fic analyte, no modification is needed for this measurement system other than
replacement of the enzyme matrix or column. Isolation of the thermal trans-
ducers from the reactants avoids fouling from blood samples, and facilitates the
stability and long-term operation of the sensors. The requirement of small
amounts of enzyme and sample, as well as the capability of more than 100 blood
assays per enzyme column, made the determination cheap and convenient, for
instance by using capillary blood taken from the finger. Recently cholesterol
determination was revisited, and a procedure was developed that allows estima-
tion of free and esterified LDL/HDL cholesterol [32].
A miniaturized thermal biosensor was evaluated [33] as part of a flow-injec-
tion analysis system for the determination of glucose in whole blood. Glucose
was determined by measuring the heat evolved when samples containing glu-
cose passed through a small column containing immobilized glucose oxidase
and catalase. Samples of whole blood (1
l) were measured directly, without any
pretreatment. The correlation between the response of the thermal biosensor
and other devices, i.e. the portable Reflolux S meter (Boehringer Mannheim,
Fig. 10a), the colorimetric Granutest 100 glucose test kit (Merck Diagnostica)
and the Ektachem (Kodak) instrument (Fig. 10b), was evaluated. The influence
of the hematocrit value and of possible interference was reported. The corre-
lation measurements showed that the thermal biosensor generally give lower
values than the reference methods when aqueous buffer standards were made
for calibration of the ET. Mean negative biases range from 0.53 to 1.16 mM.
Differences in sample treatment clearly complicated the comparisons and the
proper choice of reference method. There was no influence from substances
such as ascorbic acid (0.11 mM), uric acid (0.48 mM), urea (4.3 mM) and aceta-
minophen (0.17 mM), on the response to 5 mM glucose. The hematocrit value
did not influence the glucose determination, for hematocrit values between 13
and 53%.
m
Search WWH ::




Custom Search