Electrochemical
o f
M e t h o d
for
Urinary-tract
VIRGINIA A. LAMB, M.S.,
the
Early
D e t e c t i o n
Infections
HARRY P. DALTON, P H . D . , AND J U D D R. WILKINS, P H . D .
From the Virginia Commonwealth University, Medical College of Virginia, Richmond, Virginia
23298, and the National Aeronautics and Space Administration,
Langley Research Center, Hampton, Virginia 23665
ABSTRACT
Lamb, Virginia, A., Dalton, Harry P., and Wilkins, J u d d R.: Electrochemical
method for the early detection of urinary-tract infections. Am J Clin Pathol
66: 9 1 - 9 5 , 1976. A method for detecting the presence of bacteria in
urine based on measuring a change in potential between two electrodes was
tested in a clinical microbiology laboratory. Initial tests were conducted
with 13 bacteria commonly associated with urinary-tract infections; all of
the test organisms were detected within 2 - 9 hours. A linear relationship was
established between inoculum size and the time an increase in voltage
was observed on a strip-chart recorder. No response was seen with sterile
urine, but urine samples inoculated with Escherichia produced the expected
positive response. One hundred twenty-eight urine specimens from
hospitalized persons were simultaneously tested by the electrochemical
detection method (ECDM) and by conventional bacteriologic procedures.
Ninety-four per cent of 49 positive samples with counts of 105 organisms
per ml. or more were detected within 4 hours and 100% at 10 hours
with the ECDM. Twenty-nine specimens with counts less than 105 cells
per ml. were detected in 3.5 to 9 hours; two samples (8%) in this
group were positive within 4 hours. Fifty samples were negative for
bacterial growth, and no increase in voltage was found with these samples.
(Key words: Urine; Electrochemical method for detecting bacteria.)
bers of bacteria in the urine (> 100,000
colonies per ml.) remains one of the most
reliable indications of urinary-tract infection.1,2 While many methods for the
demonstration of bacteriuria have been
developed, there is still need for improvement in speed and ease of detection. In
this study an electrochemical microbial
detection method (ECDM) developed by
Received August 11, 1975; received revised the National Aeronautics and Space Admanuscript September 19, 1975; accepted for publica- ministration 6 ' 7 for the early detection of
tion September 19, 1975.
Address reprint requests to Dr. Dalton: Depart- coliforms in water was evaluated in a
ment of Clinical Pathology, Virginia Common- hospital laboratory for its ability to detect
wealth University, Medical College of Virginia,
significant bacteriuria.
MCV Station, Richmond, Virginia 23298.
URINARY-TRACT INFECTIONS represent one
of the most common human bacterial infections. It is estimated that about 4% of
women and 0.5% of men in the population
have bacteria in their urine. Unfortunately,
many of these persons have asymptomatic
infections, and the detection of large num-
91
LAMB, DALTON, AND WILKINS
92
Table 1. Average Detection Times Using
the Electrochemical Instrument
Average Time of
Positive Responset
Organism*
Escherichia coli
2 hr. 20 min
Proteus mirabilis
2hr. 51 min
Klebsiella pneumoniae
2 hr. 22 min
Enterobacter aerogenes
2 hr. 31 min
Citrobacter freundii
1 hr. 42 min
Pseudomonas aeruginosa
2hr. 30 min
Serratia marcescens
1 hr. 44 min
Acinetobacter calcoaceticus var.
7 hr. , 30 min
Iwoffi (Mima)
Acinetobacter calcoaceticus var.
anitratum (Herellea)
5 hr., 11 min
Alpha-streptococci (enteric)
4 hr. , 45 min
Alpha-streptococci
5 hr., 46 min
Staphylococcus aureus
9hr. , 2 min
Candida albicans
6hr. , 12 min
* Average is based on four runs per organism.
t 10" organisms per ml. of broth.
Materials and Methods
Cultures
Escherichia coli 12014, Enterobacter aerogenes 13882, Citrobacter freundii 8090,
and Proteus mirabilis 12453 were obtained
from the American Type Culture Collection (Rockville, Md.). Klebsiella pneumoniae,
Staphylococcus aureus, Acinetobacter calcoaceticus, varieties anitratum and Iwoffi,
Serratia marcescens, Pseudomonas aeruginosa,
alpha streptococci, and Candida albicans
were isolated from clinical specimens and
identified as described in the Manual of
Clinical Microbiology.3 Inocula for the
ECDM tests were prepared by making
tenfold dilutions of a 24-hour trypticase
soy broth culture of the above organisms
in sterile 0.05% peptone water. Three
milliliters of appropriate dilutions of the
organism were then added to 27 ml. of
phenol red broth (Difco) with 1% glucose;
all culture tubes were placed in a 35-C.
water bath. Viable counts were made by
spreading appropriate dilutions from a
tenfold series on trypticase soy agar (Difco)
and counting colonies after 24-hour incubation at 35 C.
A.J.C.P. —Vol. 66
Electrochemical Detection Method
T h e equipment consisted of a test tube
(25 x 90 mm.) containing a reference
calomel electrode (Beckman Instruments,
Inc., Fullerton, Calif.) and a platinum
electrode connected to a strip-chart recorder. T h e platinum electrode was
formed by shaping a strip of platinum to
fit the circumference of the test tube. A
section of the platinum was positioned outside of the test tube for attachment to
the leads of the strip-chart recorder (model
19, Honeywell Industries Div., Fort Washington, Pa.). The reference electrode was
cemented to a diaphragmed plastic test
tube top and sterilized by exposure to
ultraviolet light for 45 minutes; the
platinum electrode and test tube were
steam-sterilized in the autoclave at 121 C.
for 15 minutes. The platinum lead was
connected to the negative terminal and the
reference lead to the positive terminal of
the strip-chart recorder. T h e strip-chart
recorder was operated at 0.5 volts fullscale with a chart speed of 10 minutes per
inch. The endpoint, or detection time,
was recorded as the time elapsed between
inoculation and the initial increase in
voltage and was read directly from the
strip-chart recording. Mean cell concentrations at the time of an endpoint were
105 to 106 cells per ml.
Urine Specimens
One hundred twenty-eight urine specimens were obtained from the bacteriology
laboratory. The charts of the patients were
examined to eliminate any specimens from
patients who at that time were receiving
antimicrobial therapy. These urines were
tested by both the electrochemical instrument and the standard bacteriologic procedure. Quantitative cultures of each
specimen were done by serially diluting the
urine and planting on blood agar plates
(Difco). The urine for the electrochemical
detection test was treated as follows: 12-
July 1976
93
DETECTION OF BACTERIA IN URINE
ml. urine samples were mixed with 15 ml.
double-strength phenol red broth and 2.7
ml. 10% glucose. The contents were mixed
well in the tube and placed in a 35-C.
water bath. As with the initial testing, the
leads from the strip-chart recorder were
connected to the proper electrodes.
Table 3. Detection Times of Various
Concentrations of Escherichia
from Urine Specimens
Strain No.
Results
Initial tests with the electrochemical
instrument were conducted with known
concentrations of the organisms that
commonly cause urinary tract infections
(Table 1). Detection times ranged from 2
to 9 hours with all of the test organisms.
A linear relationship between the bacterial
concentration of the inoculum and the time
an increase in voltage was detected on the
strip-chart recorder was demonstrated.
This graph consisted of a baseline (lag
phase), a slope (a build-up in potential),
and a leveling-off phase. Each increase in
inoculum size by one log reduced the lag
time by 6 0 - 8 0 minutes. When sterile urines
were tested the response obtained on the
strip-chart recorder was a continuous flat
baseline, indicating no increase in voltage
due to background substances. However,
sterile urines seeded with E. coli produced
a response with a slope on the stripchart recorder indicating an increase in
voltage.
20
52
7
127
128
14
103
116
119
38
70
15
106
109
31
46
No. of Organisms
per ml. Urine
(Plate Count)
3.0 x
2.8 x
2.4 x
1.0 x
5.1 x
1.3 x
1.4 x
1.2 x
4.0 x
1.7 x
1.3 x
1.0 x
4.0 x
1.1 x
1.3 x
5.0 x
109
109s
10
1088
108
10s
108
107
107
107
106
105
105
105
104
10
Time of
Detection by
Electrochemical
Method
22 min.
56 min.
15 min.
33 min.
35 min.
1 hr., 25 min.
1 hr., 30 min.
2 hr., 26 min.
59 min.
22 min.
1 hr., 14 min.
1 hr., 10 min.
1 hr., 50 min.
2 hr., 39 min.
3 hr., 20 min.
6 hr., 14 min.
One hundred twenty-eight urine specimens from hospitalized persons were
screened for the presence of bacteria with
this instrument. The samples were obtained from the hospital laboratory and
tested by the electrochemical instrument as
well as by conventional bacteriologic procedures. Forty-nine of these samples were
positive for large numbers of bacteria (105
organisms per ml. or more) (Table 2).
The detection times for these organisms
Table 2. Microorganisms in 49 Clinical ranged from 9 minutes to 10 hours, with
Urine Specimens with 100,000 or More
94% of the positive cultures being deOrganisms per Milliliter
tected within the first 4 hours. Twentynine specimens had counts of less than 105
No. Specimens
colonies per ml., and the detection times
for this group ranged from 3.5 to 9 hours.
Escherichia
15
In this group two samples (8%) were posiProteus
8
Klebsiella
3
tive within 4 hours (Fig. 1). There was no
Enterobacter
3
bacterial growth in 50 samples. In five of
Staphylococcia
3
Serratia
2
the 50 specimens, there was a slope
Pseudomonas
1
registered on the strip-chart recorder with
Streptococcus
1
no bacteria growing initially from the urine
Candida
1
specimen. When these samples were reMixed infections
12
run, the positive responses were found to
TOTAL
49
be due to contaminated media.
94
LAMB, DALTON, AND WILKINS
A.J.C.P.—Vol.
66
10"
10°
^ 107
e
10
10'
I 10'
10'
+
+
102
10'
10"
4
5
Lag time, hours
FIG. 1. Comparison of bacterial concentrations in positive urine specimens and length of the lag
phase. + = Gram-positive, — = Gram-negative, • = mixed bacteria.
Unlike the initial test using pure cultures, less of a linear relationship was
established between the concentration of
bacteria in the urine and the length of
detection time, as shown in Table 3. For
example, 109 and 107 concentrations of
Escherichia produced similar responses of
22 minutes; however, a 4-hour cutoff point
allowed for the detection of 94% of the
positive urines (10 5 colonies per ml. or
more). This time limit included only 8% of
the doubtful (less than 105 colonies per ml.)
and eliminated 100% of the negatives (no
growth) (Fig. 1).
Discussion
The electrochemical detection method
(ECDM) utilized the placing of a reference
and platinum electrode directly into broth
and detecting the presence of viable bacteria by measuring a voltage change. T h e
ECDM is not to be confused with the
impedance technic, as described by Ur and
Brown, which measures resistance to the
electrical flow of a current passed through
a solution. 5 The principle of the electrochemical method for detecting organisms
producing molecular hydrogen from the
metabolism of carbohydrates has been described by Wilkins and associates6. Characteristically, the voltage response for r e producing bacteria (Escherichia, Enterobacter, Proteus, etc.) was 0.4 to 0.5 volts
and was observed for a number of isolates
obtained from urine. Non-hydrogen-producing bacteria from urine (Staphylococcus, Pseudomonas, Streptococcus,
etc.) also produced a response with the
ECDM, but considerably less than the r e p r o d u c i n g organisms, viz.,
100-150
millivolts. The mechanism of this second
response is not known at this time.
Although positive urine specimens were
detected rapidly using the electrochemical
instrument, there was less of a correlation
between the bacterial concentration in
the urine and the length of the detection
time than there was with the initial pure
July 1976
DETECTION OF BACTERIA IN URINE
cultures. Individual urines may contain accelerating as well as inhibitory factors that
influence the electrochemical responses of
the test organisms. T h e factors could be
medications the individual is taking or
the retention of antibiotic levels from previous therapy, which may produce some
bacteriostatic action in the urine.
Natural inhibitory factors could also
contribute to variations in the responses.
The pH of the urine could be of a level
to suppress the usual growth rates of an
organism. 4 Also, the large number of
bacterial strains may account for many
variations in response times. Within a
single species of bacteria there are differences in metabolic rates. A similar inhibitory effect in which the correlation between fecal coliform counts from estuarine
water samples and detection times was
poor was observed by Wilkins; on the other
hand, good agreement was observed when
fresh water samples were tested. 7
Our results indicate that the electrochemical detection method was accurate
within the 90% range, and sensitivity of
the test was excellent, as all of the tested
urines containing any concentration of
bacteria were detected. T h e absence of
95
nonspecific reactions should also be
emphasized, as all positive responses were
caused only by the presence of bacteria
in the urine. Because of these results,
this procedure may be applicable for the
screening of urine specimens from persons
in outpatient clinics, schools, antenatal
clinics, epidemiology programs, and medical offices where sophisticated bacteriologic
facilities are not available.
References
1. Craig WA, Kunin CM, DeGroot J: Evaluation
of new urinary tract infection screening
devices. Appl Microbiol 26:196-201, 1973
2. Kass EH: Progress in Pyelonephritis. Philadelphia, F. A. Davis, 1965
3. Lennette EH, Spaulding EH, Traunt JP:
Manual of Clinical Microbiology. Second
edition. Washington, D.C., American Society
of Microbiology, 1974
4. Roberts AP, Robinson RE, Beard RW: Some
factors affecting bacterial colony counts in
urinary tract infection. Br Med J 1:400403, 1967
5. Ur A, Brown FJ: Impedance monitoring of
bacterial activity. J Med Microbiol 8:19-25,
1975
6. Wilkins JR, Stoner GE, Boykin EH: Microbial
detection method based on sensing molecular
hydrogen. Appl Microbiol 27:949-952, 1974
7. Wilkins JR, Boykin EH: Electrochemical method
for the early detection and monitoring of conforms in water. Am Water Works J (in press)