| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
Integ, Inc., 2800 Patton Rd., St. Paul, MN 55113.
a Author for correspondence. Fax 651-639-9042; e-mail phil.stout{at}integ-inc.com
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Methods: Conventional electrochemical test strip technology
(Bayer Glucometer Elite®) was adapted to measure glucose
in small (0.5–2.0 µL) samples of ISF. Test strip glucose
measurements were performed on a commercial potentiostat and were
compared to various reference glucose methodologies (YSI 2300 analyzer,
microhexokinase procedure, Bayer Glucometer Elite).
Characterizations of the integrated ISF sampling-glucose test strip
design included accuracy and precision in various sample media (saline,
ISF surrogates, diabetic ISF samples), sample volume dependence, test
strip sterilization studies (electron beam, 
irradiation), and
diabetic ISF sampling and glucose measurements.
Results: Glucose measurements were free from significant media effects. Sample volume variations (0.6–3.2 µL) revealed only modest dependence of glucose measurement bias on sample volume (-1.5% per microliter). Sterilization treatments had only a minor impact on glucose response and test strip aging and no significant impact on interferent responses of the glucose test strips. Diabetic subject testing under minimum fasting conditions of at least 2 h with integrated ISF sampling and glucose measurement gave low ISF glucose measurement imprecision (CV, 4%) and mean glucose results that were indistinguishable from reference (microhexokinase) ISF glucose measurements and from capillary blood glucose measurements (Glucometer Elite).
Conclusions: Conventional single-use, electrochemical glucose test strip and ISF collection technologies can be readily integrated to provide real-time ISF sampling and glucose measurements for diabetic monitoring applications.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
To reduce the pain, mess, and inconvenience associated with glucose monitoring, many minimally and noninvasive glucose measurement systems are currently being developed. These technologies include near-infrared radiation transmission through, or reflectance from, body tissue; reverse iontophoretic or chemically enhanced fluid extraction from the skin; use of alternative body fluids; light scatter from body tissues; photoacoustic spectroscopy; Raman spectroscopy; and ocular fluid polarization changes. These techniques have been reviewed extensively (2)(3)(4)(5) and show varying degrees of promise.
In pursuit of a less invasive glucose measurement system, Integ has developed a proprietary method for extracting fluid from the dermis. This sampling method involves penetration by a cannula into the dermis and allows for the routine collection of ~1 µL of interstitial fluid (ISF), with a median collection time of between 4 and 5 s. Studies have suggested that glucose in ISF samples collected via this method closely reflects ambient glycemia (6) and that there is no clinical difference between glucose in the interstitial and venous compartments in subjects whose glucose concentrations are changing rapidly (7). These studies concluded that a correlation exists between ISF glucose and venous plasma glucose.
Many techniques have been used to sample and characterize fluid collected from the cutaneous and subcutaneous layers of skin. For example, blister fluid has been collected and studied as an analog for ISF. Studies conducted on human subjects concluded that blister fluid total protein is one-third that of serum and that there is a highly significant correlation between skin glucose and plasma glucose (8)(9)(10). Additional research utilizing suction effusion fluid and subcutaneous tissue fluid collected by the liquid paraffin cavity method and the wick method have shown that the total protein concentration in tissue fluid is approximately one-third that of plasma with an increased albumin/globulin ratio relative to plasma (11)(12)(13)(14)(15)(16). Work done with subcutaneous implanted glucose sensors has shown that ISF glucose correlates well with plasma glucose (17)(18)(19). Several studies with microdialysis probes have concluded that the glucose concentration in the subcutaneous tissue is equivalent to venous blood glucose under steady-state conditions (20)(21)(22)(23)(24). However, one of these studies also tested subjects after a very rapid increase in blood glucose and found a significant difference between ISF glucose and blood glucose concentrations (21). Research performed in our laboratory has shown that the ISF collected by the Integ sampling technology is consistent with the foregoing literature characterizations of ISF in that the total protein concentration is approximately one-third that of plasma and the albumin/globulin ratio is, on average, 1.85.
We reported previously on an earlier attempt to integrate our ISF sampling technology with mid-infrared measurement technology in a hand-held ISF glucose meter (25). In a correlation study of 445 subjects with type 1 and type 2 diabetes under 2-h fasting conditions, a standard error of prediction of 241 mg/L (24.1 mg/dL) and correlation coefficient of 0.95 was obtained using the infrared methodology when compared to venous plasma glucose.
Here, we report our efforts to integrate the ISF collection technique with an electrochemical biosensor measurement technology similar to measurement technologies typically used in conventional self-monitoring blood glucose meters. The objectives of these studies were (a) to demonstrate the feasibility of integrating ISF collection with conventional, single-use, electrochemical test strip glucose measurement technology and (b) to characterize the glucose measurement performance of an integrated ISF collection and glucose test strip device. Glucose measurement characterizations included studies of dependence on sample medium and sample volume, sterilization treatment, interferent response, and precision and accuracy.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
). This integration permits direct sampling of ISF from the
dermis and real-time sample transfer to the test strip read area for
subsequent ISF glucose measurement. All measurements made with
disposable integrated ISF collection-glucose test strips were performed
in single-use mode.
|
glucose test strip
The glucose test strip was a modified, electrochemical strip
(Glucometer Elite®; Bayer, Diagnostics
Division) that uses potentiostatically controlled working and
counter electrodes in combination with glucose oxidase enzyme and
ferricyanide electron transfer mediator reagents (26). The
glucose test strips were physically modified to allow coupling of the
test strip to the ISF collection adapter. To preserve glucose test
strip activity, the integrated ISF collection-test strip assemblies
were stored over desiccant and protected from light before use. Because
of the requirement for sterile ISF collection needles, test strips were
also subjected to sterilization treatments to model the glucose
measurement performance that could be expected from sterilized,
integrated ISF collection-test strip assemblies.
isf collection, sample detection, and glucose measurement
The integrated ISF collection-test strip assemblies required
sample volume of 0.5–1.0 µL to perform glucose measurements. In use,
the integrated ISF collection-test strip assembly was housed in a
hand-held ISF sampling device that provided rapid collection of ISF
from the forearm. The ISF sampling device was tethered to a custom
sample-detect circuit that detected placement of the sampler on the
arm. Additionally, the circuit detected ISF wetting of the test strip
as a decrease in resistance between the test strip electrodes.
Detection of sampler placement on the arm and ISF wetting of the test
strip enabled calculation of the time required for ISF collection with
each measurement. The sample-detect circuit provided an audible prompt
for the user to remove the sampler from their arm. The circuit then
connected the test strip electrode leads to a potentiostat control
circuit for subsequent measurement of sample glucose. Two potentiostat
control circuits were used. An unmodified Glucometer Elite meter was
used in selected studies to perform glucose measurements with the
integrated ISF collection-test strip assemblies. Glucometer Elite meter
readings were adjusted by an appropriate calibration response curve to
correct for the impact of the ISF collection adapter and the ISF
surrogate media on test strip glucose signals.
For all other studies, a commercial EG&G Princeton Applied Research model 273A potentiostat (EG&G Instruments) was used to perform the same test strip control functions as a conventional electrochemical glucometer circuit. The potentiostat was programmed with a 25-s delay to allow adequate glucose reaction time after sample detect. After the glucose reaction time was complete, the potentiostat was operated in a chronoamperometric mode for 10 s with the center working electrode of the test strip polarized at +0.50 V vs the outer counter electrode. During this polarization time, the potentiostat recorded the current flowing between the test strip electrodes. Current readings were made from the resulting current-time traces by averaging 0.5 s of current measurements centered around 5 s of polarization time (hereafter referred to as 5-s current readings). All glucose concentration predictions from EG&G potentiostat measurements were based on glucose calibration curves determined in physiologic saline media. Laboratory temperature was controlled (typically ±1.5 °C) so that compensation of glucose concentration predictions based on temperature was not necessary.
For ISF collection and glucose measurement studies, informed consent was obtained from diabetic subjects enrolled under a WIRB®-approved protocol for integrated ISF sampling and glucose measurement. Other studies required delivery of ISF surrogate samples to the integrated ISF collection-test strip assemblies. For some studies, surrogate sample was delivered directly to the test strip read area by a digital hand pipetter (Pipetman models P2 and P10; Rainin Instrument). In other studies, surrogate sample was delivered to the test strip through the ISF collection needle by a syringe pump (KdScientific model 200; KD Scientific) and Hamilton glass syringe (25-µL model 720RN; Hamilton) connected to the ISF collection needle via plastic tubing.
reagents and sample media
Nonsterile isotonic saline reagent (9.0 g/L NaCl) was from Fisher
Scientific. D-(+)-glucose, human albumin (fraction V),
human 
-globulins (Cohn fraction IV-1), Triton
X-100® surfactant
(t-octylphenoxypolyethoxyethanol), and sodium phosphate
(dibasic, anhydrous) were obtained from Sigma. Chemicals for
interference testing included acetaminophen
(Tylenol®), uric acid,
L-ascorbic acid (vitamin C),
L-(+)-lactic acid, acetylsalicylic acid
(aspirin), bilirubin (mixed isomers), cholesterol, and
triglycerides [triolein, C18:1, (cis)-9],
all obtained from Sigma. ProClin 150 preservative was from Supelco.
Physiologic glucose concentrations were prepared in various media by
adding microliter volumes of a 1 mol/L glucose stock solution
prepared in deionized water (MilliQ Plus deionizer; Millipore). Sample
media included physiologic saline, synthetic ISF surrogates, and human
plasma- and serum-based ISF surrogates. Synthetic ISF surrogates were
prepared by dissolving human albumin and -globulins in a ratio of 60
g/L albumin to 40 g/L -glob ulins in isotonic saline containing 6
mg/L ProClin 150 preservative to achieve total measured protein
concentrations of 11 and 31 g/L (1.1 and 3.1 g/dL). Human plasma-based
surrogates were prepared from human plasma isolated by centrifugation
of whole blood drawn into 7-mL gray-stoppered
Vacutainer® Tubes (14 mg of potassium oxalate
and 17.5 mg of sodium fluoride; Becton Dickinson). Plasma aliquots were
pooled and diluted 1:2 (500 µL + 500 µL) with isotonic saline to
give 1:2 diluted plasma-based ISF surrogate. We assumed the total
protein concentrations of 1:2 diluted plasma-based ISF surrogate to be
35 g/L (3.5 g/dL). Human serum-based surrogates were prepared from
human serum isolated by clotting and centrifugation of whole blood
drawn into 7-mL red-stoppered Vacutainer Tubes. Serum aliquots
were pooled and diluted 1:3 (333 µL + 667 µL) with 60 mmol/L
phosphate-buffered isotonic saline, pH 7.4, to give 1:3 diluted
serum-based ISF surrogate. Buffered diluent was used for serum-based
ISF surrogates to buffer the surrogate against addition of acidic
interferent test compounds (see below). We assumed the total protein
concentration of 1:3 diluted serum-based ISF surrogate to be 23 g/L
(2.3 g/dL). Assumed protein concentrations were used for plasma- and
serum-based ISF surrogates because synthetic surrogate studies
demonstrated negligible impact of protein concentration variations
[0–31 g/L (0–3.1 g/dL)] on ISF glucose measurements (see
Results and Discussion). Saline glucose calibrators and
synthetic and human plasma- and serum-based ISF surrogate solutions
were stored at -70 °C.
glucose reference measurements
Reference glucose concentration measurements were made in ISF
surrogate samples using a YSI model 2300 Stat Plus glucose analyzer
(Yellow Springs Instruments). The YSI glucose measurements differed
from direct, undiluted test strip measurements as a function of the
protein content of the sample because of the sample dilution used in
the YSI 2300. Accordingly, ISF surrogate reference glucose measurements
were corrected for YSI sample dilution errors by dividing the YSI
result by the water content of the sample {1 - [protein
concentration (g/dL)]/100)} (27). For synthetic ISF
surrogates, total protein concentrations measured by electrophoresis
were used for YSI dilution corrections. For plasma- and serum-based ISF
surrogates, YSI dilution corrections assumed 35 g/L (3.5 g/dL) protein
for 1:2 diluted plasma surrogates and 23 g/L (2.3 g/dL) protein for 1:3
diluted serum surrogates.
Reference glucose concentrations were measured in diabetic ISF samples by a an enzymatic hexokinase method modified for use with microliter sample volumes (hereafter called microhexokinase procedure). For this procedure, ISF samples of ~1.5 µL were collected into 2-µL heparin-containing capillary tubes (Drummond Scientific) and were stored at -70 °C. Modifications to the standard single-reagent, manual hexokinase procedure (kit no. 16-50; Sigma Diagnostics) were necessary because of the small ISF sample size, the range of volumes collected, and the spectral sensitivity requirements at low glucose concentrations. The standard procedure described by Sigma requires a 10.0-µL sample volume and 1.00-mL reagent volume. The procedure was modified to accommodate volumes of between 0.50 and 2.00 µL and a reagent volume of 75.0 µL. To account for the variable sample volumes, each ISF sample was weighed using an MT5 microbalance (Mettler-Toledo), and a gravimetric correction was applied. Absorbances were measured with a Spectronic Genesys 5 spectrophotometer (Milton Roy).
glucose test strip sterilization and interferent studies
The impact of sterilization on the glucose response, aging, and
selectivity of the electrochemical glucose test strips was modeled by
exposing test strips in their original packaging to -irradiation
(Steris-Isomedix) and electron-beam (Titan Scan Systems) sterilization
treatments at 20 and 40 kGy doses. The 20-kGy dose represents the
radiation dose required to sterilize the needle assembly shown in Fig. 1
. Test strip aging time points were 0, 1, and 4 months of real-time
aging (ambient storage temperature, 20–25 °C) and accelerated
equivalent aging of 6 months [storage for 36 days at 45 °C in an
ESPEC environmental chamber (model LHU-112; ESPEC)]. Test strips were
stored unopened in their original packaging for all sterilization and
aging treatments (0% relative humidity was assumed for the package
headspace).
The glucose responses of sterilized and nonsterilized test strips were determined for each experimental treatment by measuring 10 replicates at each of two glucose concentrations (5 and 15 mmol/L) in synthetic ISF surrogate medium [31 g/L (3.1 g/dL) total protein; 5-µL sample volume]. Aside from the sterilization and aging treatments, the test strips used in this study were unmodified (i.e., test strips were not combined with ISF collection needle adapters). All glucose measurements were performed on a single unmodified Glucometer Elite meter according to the manufacturer's instructions.
The effect of the sterilization treatment on test strip glucose selectivity was determined by measuring the response of glucose test strips to various endogenous and exogenous compounds that have potential to interfere with electrochemical glucose measurements (28)(29)(30). The sample medium used for interference testing was 1:3 diluted human serum-based ISF surrogate containing 4.7 mmol/L glucose. Serum-based ISF surrogate was used for interference studies to avoid contaminating the sample medium with the anticoagulants and antiglycolytic preservatives present in plasma collection tubes. The glucose concentration for the interference studies was based on NCCLS testing guidelines (31) and was chosen to represent the lower acceptable limit for controlling blood glucose as established by a survey of physicians (32) Interferent concentrations were chosen based on NCCLS testing guidelines (31). Most interferent test samples were prepared by adding interferent stock solution to serum-based ISF surrogate. Because of solubility limitations, triglyceride and cholesterol interferent test samples were prepared as emulsions by direct dispersion into ISF surrogate containing 5 g/L Triton X-100 surfactant. ISF surrogate pH extremes (pH 6.8 and 8.8) were prepared by adding microliter amounts of concentrated HCl or NaOH. Appropriate precautions were taken to avoid exposure of interferent test solutions to air and light.
Interferent responses were measured on nonsterilized and sterilized
(electron beam, 20-kGy dose) glucose test strips. Unmodified,
nonsterilized control test strips were tested for interferent
responses, using 5.0-µL sample volumes on an unmodified Glucometer
Elite meter. Sterilized test strips were integrated with ISF collection
needle adapters (Fig. 1
). These were subsequently tested for
interferent responses using 2.0-µL sample volumes delivered by pipet
to the test strip read area followed by measurement of
chronoamperometric current-time response curves using the EG&G
potentiostat configuration outlined above. Seven replicates were tested
for each interferent test solution on both nonsterilized and sterilized
test strips, and glucose measurement biases were calculated relative to
a control sample to which blank, buffered saline had been added.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
) was
measured in physiologic saline medium. Glucose samples (0.5 µL)
covering the concentration range 1.9–29 mmol/L (35–529 mg/dL; six
concentrations; five replicates per concentration) were pipetted into
the ISF collection-test strip assemblies, and the glucose current vs
time response curves were recorded using the EG&G potentiostat. Glucose
current readings were plotted against the corresponding YSI reference
glucose concentrations to produce the glucose calibration response
curve for saline medium (Fig. 2
). The saline calibration response was linear from 1.9 to 29
mmol/L with a correlation coefficient of 0.998, an
Sy|x of 0.29 µA, and an intercept statistically
indistinguishable from zero [95% confidence interval (CI) =
-0.35 to 0.04 µA; Table 1
].
|
|
Glucose measurements were also performed in 0.5-µL samples of
synthetic and human plasma-based ISF surrogates [1.9–32.6 mmol/L
glucose (35–586 mg/dL); six concentrations; five replicates per
concentration]. Glucose current measurements were made in each of the
ISF surrogate media, and corresponding medium-specific glucose
calibration curves and regression statistics were generated (Table 1
).
All media exhibited calibration intercept 95% CIs that were
statistically indistinguishable from zero. Comparison of calibration
slope 95% CIs revealed slight statistical differences between the
sample media, with the protein-containing ISF surrogate media
consistently exhibiting slightly lower slopes than the saline media
[largest slope difference, 7.5%, saline vs 31 g/L (3.1 g/dL)
synthetic surrogate].
To determine the clinical significance of the small slope differences
observed between the glucose response curves in different media,
additional measurements were made in each sample medium at 5 and 20
mmol/L (90 and 360 mg/dL) glucose. Glucose current readings in each
medium were used to make glucose concentration predictions from the
saline calibration curve of Fig. 2
. At 5 mmol/L glucose (Table 2
), the mean relative glucose prediction biases were similar
between the different ISF surrogate media (range, -1.9% to 2.0%).
The saline medium exhibits a slightly higher mean relative glucose
prediction bias (8.2%), suggesting slight positive curvature in the
saline glucose response curve near 5 mmol/L. With exclusion of a single
data point from the 31 g/L (3.1 g/dL) synthetic ISF surrogate data set,
all media exhibited CVs of 5% or less at 5 mmol/L glucose. Similar
results were observed for 20 mmol/L glucose (Table 3
), with mean relative glucose prediction biases ranging from
1.6% to 3.6% and glucose prediction CVs ranging from 1.6% to 3.6%.
For all media, the 5 and 20 mmol/L glucose prediction biases were well
within the Clarke error grid A zone (±20% bias) designated as
clinically accurate (33). Thus, medium-dependent differences
in glucose response appear to be clinically insignificant in terms of
glucose prediction accuracy and precision.
|
|
sample volume and glucose response
Physiologic and sampling variables produced a range of ISF sample
volumes during ISF collection and measurement studies. The effects of
sample volume variations on glucose prediction bias were studied by
using a syringe pump to deliver synthetic ISF surrogate samples [31
g/L (3.1 g/dL) protein, 5 or 15 mmol/L glucose] through the needle of
the ISF collection-test strip assembly. Variations in sample volume
delivered to the test strip read area were produced by varying syringe
pump sample delivery rates (2, 8, or 20 µL/min) and by introducing
variable delays (0, 1.5, or 3 s) before terminating sample
delivery following sample detect. These operating variables were chosen
to produce delivered sample volumes that ranged from 0.6 to 3.2 µL.
Delivered sample volumes were recorded from the syringe pump after each
measurement cycle. Glucose measurements were made using an unmodified
Glucometer Elite meter connected to the electrode leads of the
integrated ISF collection-test strip assembly. Glucose measurement
biases were calculated vs the appropriate YSI reference glucose
measurements. Over the range of sample volumes studied (0.6–3.2 µL),
the dependence of glucose prediction biases on sample volume was slight
and on the order of -1.5% bias per microliter of increased sample
volume (Fig. 3
). ISF collection and sampling experience with the integrated
ISF collection-test strip design suggested 0.9–2.5 µL as the
worst-case range of sample volumes that can be expected from diabetic
ISF sampling. Over this range of sample volumes, the impact of
volume variations on glucose measurement bias is expected to be
clinically insignificant.
|
sterilization and glucose response
Current commercial home use glucose monitoring systems do not
require sterilization of glucose test strips. However, integration of
glucose measurement with biological fluid sampling needles may require
sterilization of glucose test strips. Fig. 4
shows the effect of sterilization treatments on the glucose
response and aging of the glucose test strips when tested with 5 mmol/L
(90 mg/dL) glucose samples. The initial impact of sterilization
treatments is seen from the day 1 data in Fig. 4
. Both -irradiation
and electron-beam sterilization treatments initially increased glucose
responses compared with the control by 10–30%, depending on the
sterilization dose and method. At a given dose, sterilization-induced
biases were higher for irradiation than for electron-beam
sterilization. The more pronounced effect of sterilization by
irradiation may be attributable to the greater oxidative reactivity
expected from irradiation for a given sterilization dose
(34). Sterilization-induced biases also increased with
sterilization dose by both methods.
|
Inspection of the 1-, 4-, and 6-month (accelerated) stability time
points in Fig. 4
reveals that sterilization-induced biases did not
change dramatically over the course of the study. Sterilization-induced
biases were smaller but exhibited similar patterns when tested at 15
mmol/L glucose, and sterilization and aging treatments had no
appreciable impact on the precision of glucose measurements at either
glucose concentration (data not shown). These results suggest that the
impact of sterilization treatment on test strip glucose response may
readily be accounted for by appropriate sterilization process controls
and strip calibration strategies. In addition, Table 4
shows that 20-kGy electron-beam sterilization treatment has no
significant effect on the interferent response of the electrochemical
glucose test strip. The results of these studies suggest that
sterilized integration of glucose measurement and ISF sampling
technology may be readily achieved with conventional electrochemical
glucose test strip designs.
|
diabetic isf glucose measurement
Integrated ISF sampling-glucose measurement was studied in
diabetic subjects after a minimum of 2 h fasting. The first
of these studies focused on the accuracy of real-time electrochemical
measurements of glucose in ISF samples collected using the integrated
ISF collection-test strip assembly shown in Fig. 1
. In this study (Fig. 5
), pairs of capillary blood Glucometer Elite measurements were
alternated with sets of four ISF glucose measurements. The first set of
four ISF measurements was made real-time on integrated, sterilized
test-strip ISF collection adapter assemblies using the EG&G
potentiostat. Reference microhexokinase glucose measurements were
performed on the second set of ISF samples (~1.5 µL of ISF
collected into each of four capillary tubes). The data shown in Fig. 5
are representative of results obtained from diabetic subjects and
illustrate the good agreement between the two ISF glucose measurement
methods as well as between ISF and capillary blood glucose
measurements.
|
A second experiment focused on ISF glucose measurement precision during
real-time integrated ISF sampling and measurement in diabetic subjects.
For this experiment (Fig. 6
), capillary blood glucose measurements on a Glucometer Elite
were alternated with two ISF glucose measurements performed on the EG&G
potentiostat for periods of 40–60 min for each subject. As can be seen
from the representative subject results shown in Fig. 6
, the ISF
glucose measurements showed good agreement with capillary blood glucose
measurements. Moreover, the precision of real-time, integrated ISF
sampling and electrochemical glucose measurements compared favorably
with the capillary blood glucose measurement precision. The observed
ISF precision reveals both good glucose measurement precision and low
site-to-site variation of glucose in ISF samples. This minimal
site-to-site sampling variability in ISF glucose concentrations is
consistent with more detailed studies of site-to-site ISF sampling
glucose variability conducted in our laboratories and communicated
elsewhere in this issue (35).
|
In conclusion, this work has demonstrated that conventional single-use, electrochemical glucose test strip technology can be readily integrated with an ISF collection device to provide real-time ISF sampling and glucose measurements. This integration was accomplished with low ISF sample volume requirements and negligible sample volume dependence. In addition, the integrated ISF collection-glucose test strip assembly exhibited clinically insignificant dependence on sample medium and protein content and was only minimally affected by sterilization treatments. Lastly, diabetic studies of real-time ISF collection and glucose measurement gave good precision and accuracy. Together, these findings highlight the potential clinical utility of real-time, integrated ISF collection and electrochemical glucose measurements for diabetic glucose monitoring applications. Because ISF can be obtained in a substantially painless and blood-free operation, it is reasonable to expect that integration of accurate and precise glucose measurements with convenient ISF sampling would encourage aggressive monitoring of ambient glycemia among patients with diabetes mellitus. Indeed, subjects with diabetes who have participated in these tests have expressed a strong preference for the Integ non-fingerstick sampling methodology. Accordingly, we believe that integrated ISF sampling and glucose measurement technology will lead to improved glycemic control among patients with diabetes and ultimately to reductions in the long-term complications that result from diabetes (1).
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  S. Schmidt, R. Banks, V. Kumar, K. H. Rand, and H. Derendorf Clinical Microdialysis in Skin and Soft Tissues: An Update J. Clin. Pharmacol., March 1, 2008; 48(3): 351 - 364. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, ISF, real-time electrochemical measurement; 
, ISF,
microhexokinase measurement; 
, capillary blood, Glucometer Elite,
sample 1;
,
capillary blood, Glucometer Elite, sample 2. Mean glucose results: 6.38
mmol/L for integrated ISF sampling + test strip (95% CI, 5.68–7.09
mmol/L); 5.99 mmol/L for microhexokinase (95% CI, 5.38–6.61 mmol/L);
6.50 mmol/L for capillary blood (95% CI, 6.28–6.71 mmol/L).