Content uploaded by Tahir S Pillay
Author content
All content in this area was uploaded by Tahir S Pillay on May 14, 2015
Content may be subject to copyright.
Historical perspectives in clinical pathology: a history
of glucose measurement
Nareshni Moodley,
1
Unathi Ngxamngxa,
1
Magdalena J Turzyniecka,
1
Tahir S Pillay
2,3
1
Department of Chemical
Pathology & NHLS Inkosi
Albert Luthuli Central Hospital,
School of Laboratory Medicine
and Medical Sciences, College
of Health Sciences, University
of KwaZulu-Natal, Durban,
South Africa
2
Faculty of Health Sciences and
Steve Biko Academic Hospital,
Department of Chemical
Pathology and NHLS Tshwane
Academic Division, University
of Pretoria, Pretoria, South
Africa
3
Division of Chemical
Pathology, Department of
Clinical Laboratory Sciences,
University of Cape Town, Cape
Town, South Africa
Correspondence to
Professor Tahir Pillay,
Department of Chemical
Pathology, Faculty of Health
Sciences, Private Bag X323,
Arcadia 0007, South Africa;
tspillay@gmail.com
Received 25 September 2014
Revised 10 December 2014
Accepted 15 December 2014
To cite: Moodley N,
Ngxamngxa U,
Turzyniecka MJ, et al.J Clin
Pathol Published Online
First: [please include Day
Month Year] doi:10.1136/
jclinpath-2014-202672
ABSTRACT
This is the second in the series of historical articles
dealing with developments in clinical pathology. As one
of the most commonly measured analytes in pathology,
the assessment of glucose dates back to the time of the
ancient Egyptians. It was only in the 19th century that
advances in chemistry led to the identification of the
sugar in urine being glucose. The following century
witnessed the development of more chemical and
enzymatic methods which became incorporated into the
modern analysers and point-of-care instruments which
are as ubiquitous as the modern day cellphones.
Tracking the milestones in these developments shows
the striking paradigms and the many parallels in the
development of other clinical chemistry methods.
INTRODUCTION
Glucose is one of the most commonly measured
analytes in clinical pathology. It can be measured in
a number of different body fluids including blood
plasma, urine and cerebrospinal fluid (CSF). The
measurement of glucose has evolved from assessing
whether ants were attracted to urine to non-
invasive watches that measure glucose in a com-
pletely automated fashion.
1
In common with other
analytes, for example, human chorionic gonado-
trophin (HCG) (reviewed in a previous article),
detection was first achieved using organisms or
animals, progressing to manual chemical methods
and then eventually to the use of combination of
chemistry and electronics.
The first documentation of diabetes was carried
out by ancient Egyptians as early as 1500 BC.
Diabetes was initially thought to be a disease of the
kidneys and only in 1776 did British physiologist
Matthew Dobson (1713–1784) in his ‘Experiments
and Observations on the Urine in Diabetics’find
that diabetic urine was sweet as a result of its
sugars.
2
He then this associated with hypergly-
caemia and therefore concluded it was a multi-
system disease it is now known to be. Dobson
made these findings after drawing the blood of
patients with diabetes, allowing it to stand and then
tasting the serum and discovering it to be sweet.
The physiologist described the colour as ‘common
cheese whey’. He discovered that the substance in
urine was sugar after evaporating a number of
urine samples to dryness and noticing a white
residue remaining behind that smelled like brown
sugar; it could not be distinguished by taste from
sugar and fermented when yeast was added. In
1838, George Rees, a physician, isolated sugar
from the blood serum of a diabetic patient.
3–5
In the second century, Claudius Galen described
‘the clear fluid residue within the ventricles’, which
is CSF.
6
In 1672, Antonio Valsava drained clear
fluid from the lumbar sac of a dog. He likened this
fluid to synovial fluid. Formal analysis of CSF was
preceded by the lumbar puncture technique per-
fected by Quincke.
7
In 1893, Lichteim showed the
diagnostic value of CSF glucose in bacterial and TB
meningitis.
7
EVOLUTION AND DEVELOPMENT OF METHODS
FOR MEASURING GLUCOSE
It was only in the early 19th century that glucose was
identified as the sugar present in urine. In 1841 and
1848, respectively, Trommer and Fehling developed
qualitative tests to measure glucose in urine
(table 1).
89
They used the reducing properties of
glucose with alkaline cupric sulfate reagents to
produce coloured cuprous oxide. In Paris, a chemist
named Edme Jule Maumené produced the first test
strips using merino sheep wool containing stannous
chloride (SnCl
2
).
10
If a drop of urine containing
sugar was added to it and heat applied, the strip
would turn black. In 1908, Stanley Benedict devel-
oped a reagent based on this reducing property,
which contained copper sulfate, sodium citrate,
sodium carbonate and distilled water.
11
This method
was modified multiple times by Otto Folin and Hsien
Wu in 1919 who used a phenol reagent to react with
a weakly alkaline copper tartrate solution.
12
In 1940,
Nelson modified this method further with the use of
an arsenomolybdate reagent.
13
The disadvantage with the older screening tests
that detected sugars that reduced copper was that
they were non-specific and could react with other
substances besides glucose such as fructose, galact-
ose, uric acid, ascorbic acid, ketone bodies and sali-
cylates among other reducing substances in the
urine.
14
In 1959, a ortho-toluidine method was devel-
oped based on the reaction between the aldehyde
group of aldoses and aromatic amines such as o-
toluidine in hot glacial acetic acid resulting in the
formation of a blue-green product with an absorp-
tion maximum at a wavelength of 635 nm, which is
specific for aldoses but used an acid that is strong,
noxious and corrosive and that required incubation
at elevated temperatures.
15
Current methods used in the laboratory today
are enzymatic, which are usually more specific for
analytes than chemical methods generally. The
most frequently used method is the glucose hexoki-
nase method. This method is used to measure
glucose in urine, blood and CSF.
Glucose hexokinase method
In this method, hexokinase facilitates the phosphoryl-
ation of glucose by ATP in the presence of
Moodley N, et al.J Clin Pathol 2015;0:1–7. doi:10.1136/jclinpath-2014-202672 1
Review
JCP Online First, published on January 7, 2015 as 10.1136/jclinpath-2014-202672
Copyright Article author (or their employer) 2015. Produced by BMJ Publishing Group Ltd under licence.
group.bmj.com on January 7, 2015 - Published by http://jcp.bmj.com/Downloaded from
magnesium to form glucose-6-phosphate (figure 1). The glucose-
6-phosphate is oxidised by glucose-6-phosphate dehydrogenase
(G6PD) to 6-phosphogluconate in the presence of nicotinamide
adenine dinucleotide phosphate (NADP+) or nicotinamide
adenine dinucleotide (NAD+). Reduced nicotinamide adenine
dinucleotide (NADPH) is produced and measured spectrophoto-
metrically at an absorbance of 340 nm, as it is proportional to the
amount of glucose in the sample. The reference method for
glucose measurement was developed around 1976 by Neese,
Duncan and Bayse in the Centers for Disease Control and
Prevention and is based on the reaction referred to above where
serum or plasma is deproteinated with barium hydroxide and zinc
sulfate.
16
The clear supernatant is mixed and incubated at 25°C
with ATP, NAD+, hexokinase and G6PD until completion of the
reaction and NADH is measured. A more rapid method was devel-
oped in 1982 by Neese by applying reagent directly to serum or
plasma and using a specimen blank to correct for interferences.
17
Glucose oxidase method
The glucose oxidase method involves the oxidation of glucose
to gluconic acid and hydrogen peroxide (H
2
O
2
) catalysed by
glucose oxidase (figure 2). Peroxidase and a chromogenic
oxygen acceptor are then added, resulting in the generation of
H
2
O
2
. The H
2
O
2
then reacts with phenol and antipyrine (the
Trinder reaction) to generate a coloured product that is mea-
sured spectrophotometrically. Since the method measures the
beta-D-glucose form, it requires conversion from alpha to beta
forms, which can be achieved by the addition of mutarotase or
will occur spontaneously if incubated for long enough. The per-
oxidase step of this reaction is not completely specific for
glucose and can have falsely low results with substances such as
Table 1 The chronology of the development of methods for measuring glucose
1841/1848 Urine glucose Trommer/Fehling
1908 Urine glucose Benedict’s reagent
1928 Glucose oxidase discovered Muller
1945 Clinitest Crompton and Trenner (Miles)
1956 Clinistix Free & Free (Ames)
1964 Dextrostix Adams (Ames)
1975 Yellow Springs analyzer Clark’s biosensor
1976 Reference methods for glucose hexokinase Neese, Duncan and Bayse
Glucometer technology Instrument/method Company
First generation Reflectance meters
1966 Ames reflectance meter Clemens (Ames)
1968 Haemo-Glukotest Boehringer Mannheim
1972 Eyetone Kyoto-Daiichi
1974 Reflomat
1975 HGT 20–800 or Chemstrip
bG
Boehringer Mannheim
Dexstrostix Ames
1980 Dextrometer/Glucochek Lifescan (acquired by Johnson & Johnson)
1982 Reflochek Boehringer Mannheim
1984 Accuchek/Refloflux Boehringer Mannheim
1986 Glucometer M/Glucofacts Ames
1986 Glucoscan 2000 Lifescan
Second generation Modified sampling procedure with no wiping and reduced operator variation
1987 OneTouch Lifescan
1991 Reflolux Systema Boehringer Mannheim
1992 OneTouch IIa Johnson & Johnson
Third generation Electrode technology/biosensor glucose meters
1988 ExacTech Pen meter Medisense (acquired by Abbott)
1991 Hemocue B Hemocue
1992–1194 Accutrend, Accutrend Mini and Accutrend Alpha Boehringer Mannheim
1993 Glucometer Elite Bayer/Ames
1994 Companion II Medisense
1995–1998 Bayer acquires Ames, Abbott acquires Medisense, Roche acquires Boehringer Mannheim
1996 Accuchek Advantage Roche
1998 Medisense QID Medisense/Abbott
1999 Freestyle Therasense (acquired by Abbott)
Fourth generation Continuous subcutaneous monitoring
Figure 1 The hexokinase reaction for glucose.
2 Moodley N, et al.J Clin Pathol 2015;0:1–7. doi:10.1136/jclinpath-2014-202672
Review
group.bmj.com on January 7, 2015 - Published by http://jcp.bmj.com/Downloaded from
uric acid, ascorbic acid, bilirubin, haemoglobin, tetracycline and
glutathione, the effects of which can be diminished by potas-
sium ferrocyanide (for bilirubin) and a Somogyi filtrate for most
other interferents.
17
Leland Clark and Champ Lyons from the Children’s Hospital
in Cincinnatti initiated the idea of using a biosensor to measure
glucose in 1962. They used an oxygen electrode, inner oxygen
semipermeable membrane, a thin layer of glucose oxidase and
an outer dialysis membrane. The glucose oxidase was entrapped
over the oxygen electrode via the dialysis membrane. The
decrease in oxygen concentration was measured, as it was pro-
portional to the glucose concentration in the reaction shown in
figure 2. In 1967, Updike and Hicks stabilised glucose oxidase
by immobilising it in a polyacrylamide gel on an oxygen elec-
trode, which simplified the assay.
18 19
The Yellow Springs
Instrument Company analyser (Model 23A YSI analyzer) was
released in 1975 and was the first successful analyser sold using
Clark’s biosensor technology for the direct measurement of
glucose based on the amperometric detection (change in current
when oxidation or reduction takes place) of H
2
O
2
. The analyser
was costly owing to the platinum electrode used and was there-
fore mainly found in research laboratories.
In the 1980s, interferents such as ascorbic acid, uric acid,
certain drugs and the low solubility of oxygen in biological
fluids that resulted in an oxygen deficit due to changes in
oxygen tension were the main focus. These hurdles were finally
overcome with the second generation biosensors that replaced
oxygen with redox mediators. In these biosensors, electrons
were carried from the enzyme to the surface of the electrode
forming a reduced mediator instead of H
2
O
2
and the reoxida-
tion at the electrode formed the basis of signal detection.
Ferrocene is the most popular redox mediator since it does not
react with oxygen, is stable in both the oxidised and reduced
forms, is independent of pH, shows reversible electron transfer
kinetics, and has a rapid reaction. The 1980s also saw the intro-
duction of more optimal electrodes and membranes that
enhanced sensor performance.
The third generation biosensors used conducting organic salts
such as tetrathiafulvalene-tetracyanoquinodimethane, which
mediates the electrochemistry of glucose dehydrogenase-
pyrrole-quinolinequinone enzymes and glucose oxidase with
increased selectivity, but is limited by the number of enzymes
that have direct electron transfer at the electrode surface.
20
Glucose dehydrogenase method
In this method, glucose is oxidised to gluconolactone in the
presence of glucose dehydrogenase and the NADH formed is
measured spectrophotometrically (figure 3).
17
POC TESTING
Urine
During mediaeval times, various diseases were diagnosed by
examining urine samples for appearance, colour, sediment and
taste. Consequently in early history, diabetes monitoring was
dominated by urine tests.
5
Reference has been made to Jules Maumené who developed a
very simple reagent strip, which consisted of strips of sheep’s
wool containing stannous chloride which turned black when
urine containing sugar was added.
10
In 1883, George Oliver published bedside urine testing and
marked reagent papers for testing urine using the reduction of
alkaline indigo-carmine to detect sugar.
21
In 1908, Stanley
Benedict improved on the copper-reduction method by improv-
ing the copper reagent used.
11
For more than 50 years, the
Benedict method was the main method for monitoring glucose
in urine from diabetic patients.
5
Glucose oxidase was first discovered in bacteria by Muller in
1928
22
and today forms the basis for many of the detection
methods in glucose meters. In 1945, the Clinitest was intro-
duced by Crompton and Trenner at Miles Laboratories (later
part of Bayer from 1978).
23
This was a modified copper reagent
(alkaline copper sulfate; sodium citrate) tablet containing all the
required agents. The tablet reacted with a small quantity of
urine and glucose present, which could be oxidised, resulting in
a colour change to red cuprous oxide depending on the glucose
concentration. However, the test was subject to interference
from other susbtances.
23
In 1954, Eli Lilly and later Boehring
Mannheim marked the Glucotest/Testape roll which was based
on the glucose oxidase method.
24
In 1956, the Clinistix, a urine
reagent strip impregnated with glucose oxidase, peroxidase and
ortholidine, was introduced by Alfred and Helen Free
25
at the
Ames Corporation. In the presence of glucose, the ortholidine
was oxidised to a deep blue chromogen. In 1964, the Combur
test strips were launched by the company Boehringer
Mannheim, which later became part of Roche.
51726
Until the 1960s and 1970s, urine glucose was measured and
blood glucose levels were in turn inferred from these measure-
ments. The problem of generally high glucose levels being
detected as a result of the renal threshold resulted in urine
glucose tests being used to screen rather than monitor diabetes.
It was only in the early 1980s that the direct measuring of
blood glucose became widely accepted in the USA
27 28
after it
had become the widespread practice in Europe.
In 1956, Alfred and Helen Free showed the glucose oxidase
test method used by the Clinistix was very sensitive for the
detection of glucose. The enzyme glucose oxidase had been
identified by Muller about 30 years prior to this.
22
It was
demonstrated that there was better sensitivity of this method
when compared with other methods such as Benedict’s test
because the Benedict’s test was a test for total reducing sub-
stances rather than for glucose measurement.
25
An example of a modern day dry reagent test strip is the
Combur test strip by Roche. The test principle used is still that
of glucose oxidase. The H
2
O
2
formed by the glucose oxidase
oxidises the indicator tetramethylbenzidine, which results in the
yellow colour turning a blue-green colour. These test strips are
iodated, which results in the protection of glucose detection
even when an interferent such as ascorbic acid is present.
26
Figure 2 The glucose oxidase reaction for glucose.
Figure 3 The glucose dehydrogenase reaction for glucose.
Moodley N, et al.J Clin Pathol 2015;0:1–7. doi:10.1136/jclinpath-2014-202672 3
Review
group.bmj.com on January 7, 2015 - Published by http://jcp.bmj.com/Downloaded from
Although the urine tests discussed above are rapid and inex-
pensive, there were numerous problems associated with them
such as the strip having a high imprecision at low urine glucose
concentrations; the misleading falsely high glucose concentra-
tion in dilute urine specimens; false positives obtained when
urine is contaminated by H
2
O
2
or a strong oxidising agent such
as hypochlorite; and false negative resulting in the presence of
large quantities of reducing substances such as ketones, ascorbic
acid and salicylates.
17
Another problem highlighted regarding the use of these urine
point-of-care (POC) tests was the role of human error in inter-
preting them, as they needed to be visualised. Different indivi-
duals would visualise them differently and, consequently, report
different results.
28
Cerebrospinal fluid
Following on from the observation of Kohn using Clinistix,
29
attempts were made to measure glucose in CSF and in a later
study, glucose was measured in CSF using Clinistix.
30
In a study
from 1974, CSF glucose was determined using a urine dipstick
and Benedict’s method. The results showed that when used in
combination, the tests could screen for low concentrations of
glucose with a high sensitivity and a low specificity. In a study
by Moosa in 1995 where reagent strips were used to measure
glucose in CSF, there was a high specificity in the determination
of low CSF glucose. In this application, the author was able to
distinguish normal from infected CSF.
31
Blood
In 1957, glucose in blood was measured for the first time by a
POC device when Joachim Kohn discovered that Clinistix could
also be used to determine approximate levels of blood
glucose.
29
Further improvements came when in 1964 an
employee of Ames Corporation, Ernest C Adams, developed
Dextrostix designed specifically for blood.
23
This was based on
a blue colour change that occurred after a drop of blood was
applied for a minute and then wiped off. The colour change
was compared with a colour chart to estimate blood glucose
levels.
3
It was based on the glucose oxidase–peroxidase reaction
and the first layer of the strip was semipermeable, which pre-
vented the egress of red blood cells. It required 50–100 μLof
blood to measure whole blood glucose. However, the method
and interpretation were operator variable and interpretation was
affected by ambient light and sample size, which left consider-
able room for error. Boehring Mannheim, a German company,
developed an improved product called ChemstripbG which had
a beige and blue colour strip, aiding in ease of interpretation,
and it used a cotton wool bud to wipe off the blood.
5
The development of the first glucometer in 1966 by Anton
(Tom) Clemens
32
at the Ames Corporation was a milestone in
POC glucose monitoring as it helped reduced interpretive vari-
ation and initiated the growth of POC machines in the 1980s.
The glucometer (called the Ames Reflectance Meter) measured
glucose with reflectance photometry, using the Dextrostix, and
displayed the approximate glucose reading as either 0–4, 4–10
and 10–55 mmol/L via a moving pointer. The glucometer had a
designated calibration step. It was, however, an expensive and
large device because of its lead acid battery weighing 1.2 kg and
gave falsely high results in the lower range of blood glucose
values and lower results in the higher range as compared with
laboratory values.
25
The Ames Corporation then approached a
Japanese Company Arkray Inc (formerly known as
Kyoto-Daiichi until 2000) to make a better, more convenient
glucometer.
33
This was developed in 1972 and was called the
Eyetone (Dextrometer in Japan),
33
which was more precise as a
result of sphere measurement to detect all the reflected light.
Eyetone used Dextrostix strips and gave readings with a single
analogue scale and had calibration strips.
33
Instead of a battery,
it connected to the main power outlet, which improved the
weight, price and ease of use. It performed acceptably when
compared with laboratory results but was to be used carefully in
patients with neonatal hypoglycaemia. Arkray also introduced
plastic film test strips called Glucoscot that improved colour
development and allowed removal of sample from the test strip
with cotton wool balls instead of rinsing with water, which
improved precision and prevented contamination of the
sample.
533
In 1968, the Haemo-Glukotest was developed and
later improved in 1979—this was regarded as the gold standard
of accuracy for purely visual blood glucose determination.
In 1974, Boehringer Mannheim introduced Reflomat which
was a strip that required less sample (20–30 μL) that was wiped
away with a cotton wool ball and correlated well with labora-
tory values. The year 1980 brought with it the first glucometer
with digital display called the Dextrometer and the
Glucochek.
34 35
The Dextrometer was both battery and mains
operated using Dextrostix and a known standard for calibration.
It correlated well with the laboratory but was thought to be
inferior to Eyetone.
In 1981, the Glucochek (name was later changed to
Glucoscan) machine was also marketed by Lifescan (later
acquired by Johnson & Johnson) and it was run via battery and
was digital. This represented the first generation of digital gluc-
ometers. However, the batteries did not last very long and the
timer was inaccurate. In 1983, Lifescan improved the
Glucoscan’s short battery life and imprecise timer. The 1980s
also saw the introduction of the first enzyme electrode strips that
used electrochemistry and commercial screen printed strips for
self-monitoring of blood glucose.
520
Ames took a leap in tech-
nology with the digital Glucometer I for Dextrostix that was
portable and light, battery operated and had new features like a
countdown timer, audio signals when the results were ready,
when results were >22 mmol/l and when the battery was low.
The results correlated well over a wide range with the glucose
hexokinase method and was acceptable in terms of precision.
36
Boehringer Mannheim re-entered the arena in 1982 with
Reflocheck with its own Reflotest strips, which were wiped off
with cotton wool and had barcode calibration. Reflotest was
cheaper and gave acceptable results within the error parameters
required by organisations such as the American Diabetes
Association (ADA).
37 38
In 1984, Boehringer Mannheim made
the first Accu-chek meter (marketed as Reflolux in Europe),
which visually read strips like BM Test-Glycemie 20-800R,
required smaller amounts of blood and produced a more stable
colour with Accu-Chek II. Good performance followed soon
after in 1986.
5
The first 2-pad reagent strip (called Glucostix) and data man-
agement system (called Glucofacts) were created by Ames in
1986.
36
Glucostix was used in the Glucometer 2 which was
controlled with buttons, could be programmed for preset cali-
bration, was easy to use and correlated well with the laboratory.
Glucofacts was linked to Glucometer M (that stored over 300
results, the date and the time) and eventually to a smaller meter
Glucometer GX in 1990. Glucoscan 2000 by Lifescan was also
released in 1986 and had a more reliable meter, but was not
easy to use and did not correlate well with the laboratory. With
some of the meters, overloading or underloading could poten-
tially cause problems and generate readings incongruous with
the clinical context.
39
4 Moodley N, et al.J Clin Pathol 2015;0:1–7. doi:10.1136/jclinpath-2014-202672
Review
group.bmj.com on January 7, 2015 - Published by http://jcp.bmj.com/Downloaded from
In 1987, the ADA lowered the acceptable difference of hand-
held meters from laboratory values to 15%
40
and the first
second generation glucometer was introduced, called
‘OneTouch’. This used an improved sampling protocol where
the reagent strip was preinserted and did not require washing,
wiping or blotting and sample results were produced after 45 s
which markedly reduced operator error. However, a single small
study demonstrated a 30% unacceptable variation from labora-
tory values.
41
In the same year, the first blood glucose biosensor
system called ExacTech by MediSense was also released. This
had an enzyme electrode strip originally developed in Cranford
and Oxford universities. It contained glucose oxidase and ferro-
cene that was reoxidised at the electrode, generating a current
that was read by a sensor (amperometric measurement).
ExacTech was sold in the form of a slim pen or credit card and
performed acceptably when compared with the laboratory.
42
The release of the ExacTech was a landmark in bedside monitor-
ing devices.
510
Clarke et al
41
developed an evaluation statistical
tool called ‘error grid analysis’, which was later use by
Koschinsky et al
43
and helped tell if the accuracy of a gluc-
ometer was clinically significant and was within decision limits.
5
In the early 1990s, a company called Terumo Co developed
Mediace, a glucose oxidase colorimetric system.
33
From 1995 to 1998, there were changes in ownership with
Bayer, Abbott and Roche purchasing Ames, MediSense and
Boehringer Mannheim, respectively.
5
Boehringer Mannheim produced Reflolux Systema in 1991,
which was a reflectance meter that used the BM 1-44 Glycaemie
strips and that needed 20 μL blood, produced results after 120 s
but still required wiping.
44
The reflectance meter displayed
close comparability with the hexokinase method and performed
accurately in the low (<4 mmol/L) and high (>10 mmol/L)
ranges. The Swedish company HemoCue announced the first
photometric HemoCue B system that used a capillary fill,
non-wipe glucometer.
45
It measured at two wavelengths of 660
and 840 nm, could use both battery and main electricity power
and had a innovative disposable microcuvette with dried
reagents (glucose dehydrogenase, coenzyme, diaphorase and a
tetrazolium salt) for blood collection and measuring. An
amount of 5 μL of capillary was needed and the coloured for-
mazan was formed within 20–240 s. This capillary fill technol-
ogy was simple to use but the downfall was that the reagent
microcuvettes had to be refrigerated and brought to room tem-
perature before use.
5
By 1992, Johnson & Johnson marketed the Lifescan OneTouch
IIa which was a reflectance meter that did not need timing from
application of sample to removal and it was simple, reliable, accur-
ate, precalibrated, produced results in 45 s and could store 250
results.
46 47
It was the most accurate meter at the time at low
blood glucose concentration (<4 mmol/L) but still could not fulfil
the ADA criteria. Boehringer Mannheim released the Accutrend,
which did not require wiping of sample and had a barcoded
reagent test that inhibited the use of strips from a different lot
code. It was already calibrated to give whole blood results, was
simple and needed a sample volume of 20 μL. The Accutrend
Mini was released in 1994 and the Accutrend Alpha in 1996.
39
The Glucometer Elite was sold by Bayer (Ames) in 1993 and
was the revised version of the Glucocard developed by ArkRay
in 1991. It used capillary fill technology and did not need to be
wiped, was small, compact and used electrode sensor technol-
ogy that required 5 μL sample volume.
39
In 1994, MediSense released the Companion II, which was
electrochemical. In 1996, the ADA lowered the acceptable vari-
ation from laboratory values from 15% to 5%, which proved to
be a challenge for manufacturers of bedside monitoring
devices.
48
Roche released its first biosensor blood glucose meter that
used the preferred glucose dehydrogenase and pyrroloquinoline-
quinone method and labelled it the ‘AccuChek Advantage’. Its
drawback was that it is interfered with by high concentrations of
maltose or galactose.
49
The remaining years of the 1990s saw the release of the
Glucometer Esprit biosensor with a 100 test memory that
downloaded information to the user’s computer with a Bayer
WinGlucofacts data management system in 1997. The electro-
chemical Medisense Precision QID was released in 1998. The
Glucowatch Biographe by Cygnus Inc was the first transdermal
glucose sensor approved by the US Food and Drug administra-
tion (FDA).
50
The Glucowatch Biographe used reverse ionto-
phoresis that measured the glucose in secreted subcutaneous
fluid with electrochemistry and had frequent monitoring with
alarms for low and high results. The Glucowatch failed,
however, since it took a long time to warm up, had a false
alarm, was inaccurate and caused increased sweating and skin
irritation. It was withdrawn from the market in 2008.
20
The ability to continuously monitor the blood glucose in dia-
betic patients throughout the day with minimal invasiveness and
interruption in daily living became a desire and then a possibility
with the introduction of continuous glucose monitoring systems
using ‘third generation technology’. These took readings every
10 min for up to 72 h with small third generation biosensors con-
taining catheters injected subcutaneously that measured the
glucose concentration in the subcutaneous fluid. This technology
was first described by Shichiri et al
50a
in 1982 and was marketed
by Minimed in 1999 but it did not show real-time data.
51
The FDA approved the Minimed Guardian REAL-Time
system by Medtronic, SEVEN by Dexcom and Freestyle
Navigator by Abbott, as these devices update glucose values
every minute or 5 min and the sensor can be used for up to 7
days.
41052
The GlucoDay Menarini and the SCGM Roche use
the microdialysis technique for continuous subcutaneous glucose
monitoring. Microdialysis prevents direct contact between the
interstitial fluid and transducer, improves precision and accuracy
with a lower signal drift but is still not as accurate as glucose
biosensors.
20
Continuous subcutaneous glucose monitoring has
not been formally assessed to be useful as yet but should
provide useful insight into glucose fluctuations throughout the
day so that treatment can be accurately tailored and glycaemic
control improved.
The 21st century did not disappoint with the introduction of
several devices. The OneTouch Ultra by Lifescan ( Johnson &
Johnson) in 2001, requiring only 1 mL of sample, used glucose
oxidase methodology and was one of the first glucometers to be
plasma calibrated.
53
The OneTouch Ultra introduced the ability
to test in other sites of the body like the forearm or hand,
which reduced sampling pain and increased compliance. The
OneTouch Ultrasmart was released and had the ability to organ-
ise results in an electronic logbook. In 2002, the MediSense
SoftSense unit was launched in the UK.
54
This was fully auto-
mated and had the ability to collect the blood from multiple
sites with a built-in lancing device using vacuum technology.
The MediSense SoftSense had a backup port for the usual sam-
pling of glucose. The Ascensia Breeze (Bayer)
55
and the
Therasense by Freestyle
56
were launched in 2003, with the
former using autodiscs requiring 2–3μL of sample instead of
strips. The latter weighed only 38 g and used the more reliable
glucose dehydrogenase pyroquinolonequinone and coulometry
(the amount of electricity produced from a chemical reaction is
Moodley N, et al.J Clin Pathol 2015;0:1–7. doi:10.1136/jclinpath-2014-202672 5
Review
group.bmj.com on January 7, 2015 - Published by http://jcp.bmj.com/Downloaded from
measured). This required low sample volume, was more resistant
to changes in temperature and haematocrit variation and was
unaffected by high concentrations of paracetamol, uric acid and
vitamin C. The Therasense
56
required only 0.3 μL blood even
from alternative sites and results were precalibrated for results
equal to that of plasma. The Therasense could also be used with
the Precision Xceed and Optium Xceed test strips that used
glucose dehydrogenase-nicotinamide adenine dinucleotide to
improve specificity.
The Ascensia Contour
57
by Bayer was launched a year later
and used glucose dehydrogenase-flavin adenine dinucleotide and
capillary fill technology requiring a meagre 0.6 μL of blood.
The device produced plasma-calibrated results, with a wide ana-
lytical range of 0.6–33.3 mmol/L in just 15 s. The Ascensia
Breeze II
58
was released in 2007 and used 1 μL of blood but
had a more innovative data management system. The strip tech-
nology was also improved this year with Nova Biomedical’s
StatStrip
59
with multiwall electrode technology that reduced
interferences by haematocrit variations and non-glucose sub-
stances. Contour Didget improved the accessibility of results by
allowing the device to be connected to the Nintendo DS in
2009, improving compliance in children. The OneTouch
Ultravue
60
with a colour display and a reagent strip ejector was
also released in 2009 and had simple push buttons for diabetics
who were disabled. In 2006, a talking glucometer was also
released by BBI and was called the ‘SensoCard Plus’. This
device helped patients with diabetic retinopathy. Terumo also
released the Medisafe Fit with comparable technology.
61
And in
2012, the IT for glucose monitoring improved with the
Contour USB that could connect to Glucofacts Deluxe software
for superior display options. Error messages show up, as well as
the relation of results to meals and the high or low results. The
Trueone and Trueone Twist were the first disposable blood gluc-
ometers designed by Home Diagnostics.
5
Blood glucometers were not used for self-monitoring until
the 1980s, as they were not portable, required too many
operator-dependent steps, were not accurate and precise enough
and patients’ability to interpret the results and act on them was
questioned. Issues such as funding and who was responsible if
something went wrong had not been clarified. The first patient
to use a glucometer was an engineer named Richard
Bernstein.
62
He had been diagnosed with type 1 diabetes melli-
tus and was on insulin treatment, resulting in many hypogly-
caemic attacks. He was also starting to develop the
complications of diabetes. Since only doctors could access gluc-
ometers, he asked his wife who was a psychiatrist to get one. He
then used it to monitor his own blood sugars and adjust his
insulin accordingly. He found that this resolved the previously
frequent hypoglycaemic attacks and his health improved.
Bernstein then managed to convince Ames to try to market
their product to other patients but received no support from
journals, which refused to publish any of his studies. His frustra-
tion with the lack of support he received for his passion to instil
self-monitoring of blood glucose drove him to enter medical
school in his 40s and he later super-specialised in endocrinology
and had his own diabetic clinic. He introduced the trend of
allowing patients to monitor their own blood sugar levels.
62
Manufacturers then started aiming towards safe machines with
minimal errors that were easy to use with fewer operator-
dependent steps, portable, accommodated for disabilities in dia-
betics, stored results for clinicians to review and could perform
autocalibration and quality control. Data retrieval improved
with the ability to display it in graphical forms and averages and
it became downloadable to commonly used devices such as
smartphones. ISO 15197 was released in 2003, which stated the
requirements for acceptable glucometers to have 95% of results
within ±0.83 mmol/L of the reference glucose method for con-
centrations less than 4.2 mmol/L. Within 20% for higher con-
centrations was required which standardised the quality of
glucometers among manufacturers. Despite the wide use of self-
monitoring devices, there are still certain issues that are not
fully resolved such as patient adherence, financial burden and
there is still no convincing evidence that it is helpful in type 2
diabetes monitoring.
5
FUTURE OF POCT
POC testing (POCT) devices for glucose measurement have
evolved exponentially and have become widely available. The
main focus in future developments is to incorporate data man-
agement and connectivity via information technology. It is
worth noting that European regulations require manufacturers
of in vitro diagnostic devices to submit validation data to a noti-
fiable body, but this is not required for laboratory glucose
methods, underlining the uncertainty with which POCT devices
are often viewed. Methods with fewer interferences and more
specificity in POCT devices should be used and reliable non-
invasive methods for blood glucose monitoring would improve
patient comfort and compliance. The future is likely to produce
the fourth generation of glucose measuring devices with
advances in highly specific non-invasive continuous glucose
monitoring based on transdermal or optical methods.
Take home message
The evolution of glucose measurement over the past 100 years
reflects paradigms in the development of methods from static
bedside biochemical to dynamic real-time biosensor
measurement.
Handling editor CS Lee
Contributors All the authors contributed substantially to the final manuscript. The
idea and outline was conceived by TSP and the first draft produced by NM, UN and
MJT and edited extensively by TSP. All authors approved the final draft.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
1 Das AK, Shah S. History of diabetes: from ants to analogs. J Assoc Physicians India
2011;59(Suppl):6–7.
2 [No authors listed]. Matthew Dobson (1735?-1784) clinical investigator of diabetes
mellitus. JAMA 1968;205:698.
3 Zajac J, Shrestha A, Patel P, et al. The main events in the history of diabetes
mellitus. In: Poretsky L, ed. Principles of Diabetes Mellitus. 2nd edition edn.
New York: Springer, 2010:3–16.
4 Gale EAM. History to 1900. Diapedia 1104085132 rev. no. 28. ed, 2013.
5 Clarke SF, Foster JR. A history of blood glucose meters and their role in
self-monitoring of diabetes mellitus. Br J Biomed Sci 2012;69:83–93.
6 Hajdu SI. A note from history: discovery of the cerebrospinal fluid. Ann Clin Lab Sci
2003;33:334–6.
7 Venkatesh B, Scott P, Ziegenfuss M. Cerebrospinal fluid in critical illness. Crit Care
Resusc 2000;2:42–54.
8 Trommer CA. Unterscheidung von Gummin, Dextrin, Traubensucker und Rohrzucker.
Annalen der Chemie und Pharmacie 1841;39:360–2.
9 Fehling H. Die quantitative Bestimmung von Zucker und Stärkmehl mittelst
Kupfervitriol. Annalen der Chemie und Pharmacie 1848;72:106–13.
10 Maumené EJ. Sur un nouveau réactif pour distinguer la présence du sucre dans
certains liquides. J Pharm 1850;17:368–70.
6 Moodley N, et al.J Clin Pathol 2015;0:1–7. doi:10.1136/jclinpath-2014-202672
Review
group.bmj.com on January 7, 2015 - Published by http://jcp.bmj.com/Downloaded from
11 Benedict SR. A reagent for the detection of reducing sugars. J Biol Chem
1909;5:485–7.
12 Folin O, Wu H. A System of Blood Analysis. J Biol Chem 1919;38:81–110.
13 Nelson N. A photometric adaptation of the Somogyi method for the determination
of glucose. J Biol Chem 1944;153:375–80.
14 Rotblatt MD, Koda-Kimble MA. Review of drug interference with urine glucose
tests. Diabetes Care 1987;10:103–10.
15 Ceriotti G, De Nadai Frank A. An improved proceure for blood glucose
determination with o-toluidine. Clin Chim Acta 1969;24:311–13.
16 Neese JWD, Duncan P, Bayse D. Development and evaluation of a hexokinase/
glucose-6-phosphate dehydrogenase procedure for use as a national glucose
reference method. HEW Publication No. (CDC) 77-8330. Atlanta: Centers for
Disease Control, 1976.
17 Sacks DB. Carbohydrates. In: Burtis CA, Ashwood ER, Bruns D, eds. Tietz
fundamentals of clinical chemistry. 6th edn. St Louis, Missouri: Saunders Elsevier,
2008:389–93.
18 Updike SJ, Hicks GP. Reagentless substrate analysis with immobilized enzymes.
Science 1967;158:270–2.
19 Updike SJ, Hicks GP. The enzyme electrode. Nature 1967;214:986–8.
20 Yoo EH, Lee SY. Glucose biosensors: an overview of use in clinical practice. Sensors
(Basel) 2010;10:4558–76.
21 Oliver G. On bedside urinary tests: detection of sugar in the urine by means of test
papers. Lancet 1883;121:858–60.
22 Muller D. Studien iiber ein neues Enzym Glykoseoxydase. Biochem Z
1928;199:136–70.
23 Free AH, Free HM. Self testing, an emerging component of clinical chemistry. Clin
Chem 1984;30:829–38.
24 Comer. Semiquantitative specific test paper for glucose in urine. Anal Chem
1956;28:1748–50.
25 Free AH, Adams EC, Kercher ML, et al. Simple specific test for urine glucose. Clin
Chem 1957;3:163–8.
26 Hohenberger E, Kimling H. Compendium-urinalysis with test strips. Mannheim,
Germany: Roche Diagnostics Company Brochure, 2004.
27 Frost J. Innovations in participatory medicine: the advent of do-it-yourself blood
glucose monitoring. J Participat Med 2010;2:e7.
28 Kristensen GB, Sandberg S. Self-monitoring of blood glucose with a focus on
analytical quality: an overview. Clin Chem Lab Med 2010;48:963–72.
29 Kohn J. A rapid method of estimating blood-glucose ranges. Lancet
1957;273:119–21.
30 Middleton JE, Griffiths WJ. Rapid colorimetric micro-method for estimating glucose
in blood and C. S. F. using glucose oxidase. BMJ 1957;2:1525–7.
31 Moosa AA, Quortum HA, Ibrahim MD. Rapid diagnosis of bacterial meningitis with
reagent strips. Lancet 1995;345:1290–1.
32 Cheah JS, Wong AF. A rapid and simple blood sugar determination using the Ames
reflectance meter and Dextrostix system: a preliminary report. Singapore Med J
1974;15:51–2.
33 Yamada S. Historical achievements of self-monitoring of blood glucose technology
development in Japan. J Diabetes Sci Technol 2011;5:1300–6.
34 Stewart TC. Calibration of the Dextrometer/Dextrostix system for the assay of whole
blood glucose. Clin Chem 1980;26:1622–3.
35 Stewart TC, Kleyle RM. Statistical comparison of blood glucose as determined by
several test-strip procedures and by a hexokinase procedure. Clin Chem
1983;29:132–5.
36 Brooks KE, Rawal N, Henderson AR. Laboratory assessment of three new monitors
of blood glucose: Accu-Chek II, Glucometer II, and Glucoscan 2000. Clin Chem
1986;32:2195–200.
37 Walford S, Home PD, Alberti KG. A laboratory trial of two new blood-glucose
reflectance meters featuring automatic external calibration. Ann Clin Biochem
1984;21(Pt 2):116–19.
38 Forrest RD, Jackson CA, Yudkin JS. Screening for diabetes mellitus in general
practice using a reflectance meter system. The Islington Diabetes Survey. Diabetes
Res 1987;6:119–22.
39 Devreese K, Leroux-Roels G. Laboratory assessment of five glucose meters designed
for self-monitoring of blood glucose concentration. Eur J Clin Chem Clin Biochem
1993;31:829–37.
40 ADA. Consensus statement on self-monitoring of blood glucose. Diabetes Care
1987;10:95–9.
41 Clarke WL, Cox D, Gonder-Frederick LA, et al. Evaluating clinical accuracy of
systems for self-monitoring of blood glucose. Diabetes Care 1987;10:622–8.
42 Ross D, Heinemann L, Chantelau EA. Short-term evaluation of an electro-chemical
system (ExacTech) for blood glucose monitoring. Diabetes Res Clin Pract
1990;10:281–5.
43 Koschinsky T, Dannehl K, Gries FA. New approach to technical and clinical
evaluation of devices for self-monitoring of blood glucose. Diabetes Care
1988;11:619–29.
44 Schier GM, Moses RG, Gan IE, et al. An evaluation and comparison of Reflolux II
and Glucometer II, two new portable reflectance meters for capillary blood glucose
determination. Diabetes Res Clin Pract 1988;4:177–81.
45 Ashworth L, Gibb I, Alberti KG. HemoCue: evaluation of a portable photometric system
for determining glucose in whole blood. Clin Chem 1992;38(8 Pt 1):1479–82.
46 Trajanoski Z, Brunner GA, Gfrerer RJ, et al. Accuracy of home blood glucose meters
during hypoglycemia. Diabetes Care 1996;19:1412–15.
47 Strowig SM, Raskin P. Improved glycemic control in intensively treated type 1
diabetic patients using blood glucose meters with storage capability and
computer-assisted analyses. Diabetes Care 1998;21:1694–8.
48 Tonyushkina K, Nichols JH. Glucose meters: a review of technical challenges to
obtaining accurate results. J Diabetes Sci Technol 2009;3:971–80.
49 Thomas LE, Kane MP, Bakst G, et al. A glucose meter accuracy and precision
comparison: the FreeStyle Flash versus the Accu-Chek Advantage, Accu-Chek
Compact Plus, Ascensia Contour, and the BD Logic. Diabetes Technol Ther
2008;10:102–10.
50 Garg SK, Potts RO, Ackerman NR, et al. Correlation of fingerstick blood glucose
measurements with GlucoWatch biographer glucose results in young subjects with
type 1 diabetes. Diabetes Care 1999;22:1708–14.
50a Shichiri M, Kawamori R, Yamasaki Y, et al. Wearable artificial endocrine pancrease
with needle-type glucose sensor. Lancet 1982;2:1129–31.
51 Rebrin K, Steil GM. Can interstitial glucose assessment replace blood glucose
measurements? Diabetes Technol Ther 2000;2:461–72.
52 Tavris DR, Shoaibi A. The public health impact of the MiniMed Continuous Glucose
Monitoring System (CGMS)-an assessment of the literature. Diabetes Technol Ther
2004;6:518–22.
53 Ito T, Kamoi K, Minagawa S, et al. Patient perceptions of different lancing sites for
self-monitoring of blood glucose: a comparison of fingertip site with palm site using
the OneTouch Ultra Blood Glucose Monitoring System. J Diabetes Sci Technol
2010;4:906–10.
54 Clarke P, Coleman MA, Holt RI. Alternative site self blood glucose testing is preferred
by women with gestational diabetes. Diabetes Technol Ther 2005;7:604–8.
55 Lekarcyk J, Ghiloni S. Clinical performance and ease of use of the Ascensia Breeze.
Diabetes Technol Ther 2004;6:503–4.
56 Feldman B, Brazg R, Schwartz S, et al. A continuous glucose sensor based on wired
enzyme technology—results from a 3-day trial in patients with type 1 diabetes.
Diabetes Technol Ther 2003;5:769–79.
57 Kilo C, Pinson M, Joynes JO, et al. Evaluation of a new blood glucose monitoring
system with auto-calibration. Diabetes Technol Ther 2005;7:283–94.
58 Freckmann G, Schmid C, Ruhland K, et al. Integrated self-monitoring of blood
glucose system: handling step analysis. J Diabetes Sci Technol 2012;6:938–46.
59 Tendl KA, Christoph J, Bohn A, et al. Two site evaluation of the performance of a
new generation point-of-care glucose meter for use in a neonatal intensive care
unit. Clin Chem Lab Med 2013;51:1747–54.
60 Chang A, Orth A, Le B, et al. Performance analysis of the OneTouch UltraVue blood
glucose monitoring system. J Diabetes Sci Technol 2009;3:1158–65.
61 Kuo CY, Hsu CT, Ho CS, et al. Accuracy and precision evaluation of seven
self-monitoring blood glucose systems. Diabetes Technol Ther 2011;13:596–600.
62 Bernstein RK. Blood glucose control. Arch Intern Med 1981;141:267–8.
Moodley N, et al.J Clin Pathol 2015;0:1–7. doi:10.1136/jclinpath-2014-202672 7
Review
group.bmj.com on January 7, 2015 - Published by http://jcp.bmj.com/Downloaded from
a history of glucose measurement
Historical perspectives in clinical pathology:
Tahir S Pillay
Nareshni Moodley, Unathi Ngxamngxa, Magdalena J Turzyniecka and
published online January 7, 2015J Clin Pathol
http://jcp.bmj.com/content/early/2015/01/07/jclinpath-2014-202672
Updated information and services can be found at:
These include:
References
#BIBL
http://jcp.bmj.com/content/early/2015/01/07/jclinpath-2014-202672
This article cites 56 articles, 25 of which you can access for free at:
service
Email alerting box at the top right corner of the online article.
Receive free email alerts when new articles cite this article. Sign up in the
Collections
Topic Articles on similar topics can be found in the following collections
(126)JCP Education
Notes
http://group.bmj.com/group/rights-licensing/permissions
To request permissions go to:
http://journals.bmj.com/cgi/reprintform
To order reprints go to:
http://group.bmj.com/subscribe/
To subscribe to BMJ go to:
group.bmj.com on January 7, 2015 - Published by http://jcp.bmj.com/Downloaded from
