Full text of DOI:10.1016/S0022-5347(05)64650-2

0022-5347/02/1682-0420/0
®
THE JOURNAL OF UROLOGY
Vol. 168, 420–424, August 2002
Copyright © 2002 by AMERICAN UROLOGICAL ASSOCIATION, INC.®
Printed in U.S.A.
Milestone in Urology
ASYMPTOMATIC INFECTIONS OF THE
URINARY TRACT*
By EDWARD H. KASS
(Introduced by)
MAXWELL FINLAND
BOSTON, MASS.
(From the Thorndike Memorial Laboratory and Second and Fourth
(Harvard) Medical Services, Boston City Hospital, and the Department of Medicine, Harvard Medical School)
(Reprinted with permission from Trans Am Physiol, 69: 56–64, 1956)
Infections of the urinary tract are second in frequency disappears with dilution. The significance of this indica-
only to infections of the respiratory tract. Yet, despite their tion of the presence of an inhibitor to bacterial multiplica-
relative frequency, and their apparent association, sup- tion in urine remains to be explored. In general, however,
ported by varying degrees of evidence, with such diverse
clinical states as hypertension, malignant nephrosclerosis,
toxemias of pregnancy, diabetes mellitus, our knowledge of
the incidence and pathogenesis of such infections is incom-
plete, and methods for their control seriously inade-
quate.1, 2
Several observers have found pyelonephritis in 12 to 20
per cent of autopsies; such pyelonephritis had been diag-
nosed during life in only about one-fifth of these cases.3, 4, 5
The discrepancy between the clinical and autopsy findings
seems too great to be accounted for on the basis of failure
to recognize the generally accepted clinical manifestations
of infection of the urinary tract. Alternatively, it may be
suggested that there is a high incidence of clinically atyp-
ical or inapparent infections of the urinary tract in the
population at large. The latter possibility has been inves-
tigated.
Initial observations were directed toward the problem of
determining by quantitative bacteriologic study whether it is
possible to distinguish contamination attendant upon the
collection procedure from actual multiplication of bacteria
within the urinary tract.
Several considerations suggest that contamination and
true bacilluria may be distinguished by counting the number
of bacteria in the urine.
When strains of the common pathogens of the urinary
tract are cultured in vitro in pooled specimens of sterile
urine, the bacteria generally grow readily to about 108
bacteria per ml. Even without added glucose, urine gener-
ally supports multiplication of the usual pathogens of the
urinary tract about as well as does nutrient broth, and
variations in pH and specific gravity within physiologic
ranges exert but slight effects on bacterial multiplication.
Occasional pools of urine, acidified to pH 5.0, are inhibi-
tory to bacterial multiplication but the inhibitory effect
bacteria growing in urine tend to multiply to maximal or
near maximal numbers after a sufficient time.
A similar phenomenon probably occurs in vivo. When the
urines of patients with symptoms generally recognized as
those of acute pyelonephritis, namely, chills, fever, flank pain
and dysuria, are examined by quantitative methods the num-
bers of bacteria in the urine are comparable to those found
after in vitro cultivation. Twenty-five consecutive patients
with shaking chills, fever, flank pain and dysuria were stud-
ied and all had more than 100,000 bacteria per ml. of their
urines; all but one had more than one million bacteria per ml.
Thus, the finding of large numbers of bacteria in the urine
suggests that bacteria had actually multiplied within the
urinary tract.
When the urines of female patients were collected by
catheterization, the numbers of bacteria found were dis-
tributed in parabolic fashion (Figure 1). Female medical
out patients were chosen at random, using no criteria for
selection other than that they were not regarded, by those
responsible for their care, as having active infections of the
urinary tract, and in this sense they, and other patients in
this study, are considered to be asymptomatic.
The data presented in Figure 1 indicate that there are
two separate but overlapping population groups. One
group, with bacterial counts between 0 and 10,000, pre-
sumably represents the range of contamination. The sec-
ond group, with more than 100,000 bacteria, would be
expected to encompass those in whom actual multiplica-
tion of bacteria had occurred. The two population groups
overlap at about 10,000 bacteria per ml. Thus, among 74
patients in whom the diagnosis of pyelonephritis was made
or suspected 95 per cent were found to have more than
100,000 bacteria per ml. of urine, and an additional 3 per
cent of this group had between 10,000 to 100,000 per ml.
Conversely, asymptomatic diabetic women and women
who were about to undergo surgery for repair of cystoceles
usually had less than 10,000 bacteria per ml. However, in
these groups, as might be expected, there were relatively
more patients whose urines contained more than 100,000
bacteria than were found in the unselected general group.
It is apparent that the number of bacteria found at either
end of the distribution pattern is a function of the population
groups under study. Patients with manifest or suspected
* Supported by a grant from the National Institutes of Health.
The studies reported here were carried out with the assistance of
Miss Joanne Colburn and Miss Virginia Agrella. Various aspects
of the reported observations were studied in collaboration with
Drs. Thomas W. Mou, Egon Riss, Howard Levitin and Robert
Dandrow. The cooperation of Dr. Benjamin Tenney, Director of
Gynecology and Obstetrics, and Dr. David Hurwitz, Chief of the
Diabetes Clinic of the Boston City Hospital is gratefully acknowl-
edged.
420

ASYMPTOMATIC INFECTIONS OF THE URINARY TRACT
421
infection add to the numbers with high bacterial counts, and creases in bacterial numbers. Therefore, urines should be
those in whom infection is not considered to be present swell studied within the hour after they are obtained, unless
the ranks of those with relatively few bacteria in the speci- they are refrigerated immediately.
mens to be counted.
Using 100,000 bacteria per ml. as the dividing line be-
The reproducibility of these observations in the same tween contamination and infection, it has been found (Ta-
patients is shown in Figure 2. Sixty-seven of the asymp- ble 1) that the incidence of asymptomatic bacilluria in the
tomatic women were catheterized a second time 1 to 12 medical outpatient department of the Boston City Hospital
months after the first observation had been made, and the is about 6 per cent, the incidence in diabetic women is 18
two sets of counts were compared. Those patients with per cent, in women with cystocele 23 per cent, in pregnant
high numbers of bacteria in their urines in the first count women at term 11 per cent. Asymptomatic bacilluria oc-
had high numbers the second time with but one, as yet curs in 4 per cent of males in the medical outpatient
unexplained, exception. Those with low numbers of bacte- department and in 5 per cent of males in the diabetic clinic.
ria in the first count had second counts that fell in random Ninety-five per cent of patients with inlying catheters will
fashion over the entire low-count range. Two of the pa- have developed bacilluria within 96 hours, most of them
tients with 105 or more bacteria developed symptoms be- within the first 24 hours. The prophylactic use of antibi-
tween the two examinations and are not included in this otics has not prevented such infections.
chart. Only two of 14 specimens from patients whose uri-
Although pyuria is often relied upon as an index of the
nary counts were initially between 100 and 100,000 had presence of infection of the urinary tract, it is difficult to
more than 105 bacteria in their urines on second examina- get any group of clinicians to agree on the number of white
tion.6 It is noteworthy that frank dysuria and true bacil- cells that shall be called pyuria. Pyuria has been arbi-
luria occurred after a single catheterization in about 2 per trarily defined as the presence of five or more white blood
cent of the female patients studied.
cells per high-power field, using a centrifuged specimen of
For survey purposes, a count of 105 bacteria or more per urine. Pyuria, defined in this way, occurs in one-third to
ml. of urine has been designated arbitrarily as the dividing one-half of patients with true bacilluria, depending upon
line between true bacilluria and contamination. Obviously, the group, but occurs in only 2 per cent of those with less
for individual clinical purposes a reinvestigation of counts than 100,000 bacteria per ml. of urine (Table 2). Even if the
higher than 100 may at times be useful, but only occasion- criterion of pyuria is lowered to three white cells per high
ally, as has been shown, will patients whose urines are power field, the incidence in those with true bacilluria
found to contain less than 100,000 bacteria give higher rises by but 10 per cent. It is evident that pyuria is of value
counts on repeated study, and it is not yet clear that the diagnostically only when it is clearly present. Its absence
higher counts are unrelated to the catheterization neces- from any single specimen cannot be taken as evidence of
sary to collect the specimens. High counts show little ten- the absence of bacilluria.
dency to become low counts without treatment.
The Gram stain has long been a useful clinical guide to
The identification of the bacteria found in urines in the presence of infection and was subjected to quantitative
relation to bacterial numbers is of interest. In low count evaluation. The Gram-stained smears of uncentrifugated
specimens, coliform organisms are uncommon, and rarely urines were positive in 80 per cent of counts of 100,000 and
occur in predominance or as single species in pure culture. in 20 per cent of counts between 100 and 100,000 (Table 3).
Conversely, high count specimens seldom contain diphthe- Clinicians have long depended upon the Gram stain as a
roids, saprophytic Neisseriæ, Sarcinæ and the other com- screening device to distinguish infection from contamina-
mon contaminants found in the urethra. Instead, high tion, and its use is amply justified in these observations.
counts are usually associated with the presence of the The intrinsic error in the procedure is about 20 per cent.
common pathogens of the urinary tract which are present
Details of the clinical analysis of the patients with and
in large numbers and are most often found in pure culture. without asymptomatic bacilluria can be summarized
Urines may be stored at room temperature for one hour, briefly. The incidence of hypertension is but slightly higher
or at refrigerator temperature for 48 hours without signif- in those with bacilluria than in those without. Parity and
icant changes in the bacterial population. Storage at room age are not significantly different in the two groups. A
temperature for two hours or longer leads to distinct in- history of inlying catheterization, instrumentation, previ-
FIG. 1

422
ASYMPTOMATIC INFECTIONS OF THE URINARY TRACT
ous infection, etc., is found in about one-half of those with acute pyelonephritis, of drugs such as mandelic acid, and
asymptomatic bacilluria and in less than 10 per cent of nitrofurantoin, which have negligible activity at tissue
those without true bacilluria.
levels, but are strongly bactericidal in urine under proper
Direct evidence to indicate that asymptomatic bacilluria conditions.
is associated with pyelonephritis is as yet not available.
In certain instances, the bacterial count of the urine may
Several considerations suggest that such an association is fall below the range that is characteristic of infections, and
not unlikely. First, it is common clinical observation that these instances may be cited briefly:
inlying catheterization and instrumentation may be fol-
lowed by frank pyelonephritis, indicating that bacteria
(a) If a bacteriostatic agent is present in the urine;
(b) If the rate of urine flow is rapid, the numbers of
can, at times, progress from the bladder to the kidney and bacteria discharged from the kidney small, and pooling of
cause symptoms, regardless of the specific route by which urine in the bladder for a time sufficient to permit multi-
they do so. Secondly, we have observed patients in whom plication of bacteria to maximal numbers has not occurred;
asymptomatic bacilluria was demonstrated in this study,
(c) If the micro-organism involved is fastidious in its
who suddenly, and without obvious predisposing incident, growth requirements and grows poorly in urine (this has
weeks or months later developed fever, dysuria and flank been a rare occurrence, and thus far has been limited to
pain.
It may be suggested, then, that bacteria may also reach
organisms not commonly found in the urinary tract);
(d) If there is obstruction of the ureter, thus interfering
the kidney in the course of asymptomatic bacilluria with- with the discharge of bacteria into the bladder;
out causing marked symptoms. The infected kidney would
(e) If the infection is limited to areas of the kidney not
be expected to discharge bacteria into the urine, and a directly accessible to renal tubules.
cycle would be established which could persist for long
In summary, contamination may be distinguished from
periods of time. Such an hypothesis may explain some true bacilluria by quantitation of the numbers of bacteria
puzzling aspects of the problem of pyelonephritis. It could in the urine. Asymptomatic infections of the urinary tract
account, for example, for the pathogenesis of chronic pye- occur frequently in the outpatient department of a munic-
lonephritis, and could account for the efficacy in cases of ipal general hospital. Instrumentation and catheterization
FIG. 2

ASYMPTOMATIC INFECTIONS OF THE URINARY TRACT
423
TABLE 1.—OCCURRENCE OF URINARY COUNTS Ͼ 105 IN “ASYMPTOMATIC”
OUTPATIENT DEPARTMENT POPULATION
1. 6% of 337 females in Medical Outpatient Department
2. 18% of 54 diabetic females
3. 23% of 52 females with cystocele
4. 11% of 152 pregnant females at term
5. 4% of 102 males in Medical Outpatient Department
6. 5% of 37 diabetic males
7. 95% of 22 patients with inlying catheters for 96 hours
TABLE 2.—INCIDENCE OF PYURIA (Ͼ5 WBC/HPF) IN ASYMPTOMATIC FEMALE PATIENTS
Number
Number of
Bacteria
Number of
Patients
Per Cent
“Pyuria”
with Ͼ5
wbc/hpf
0
120
60
38
14
7
4
0
3
0
10Ϫ1
101-2
102-1
103-1
104-5
105-6
Ͼ106
0
0
1*
0
7
0
5
0
0
5
12
2
40
33
4
Total . . . . . . . . . . . . . . . .
261
11
* On subsequent count this patient had Ͼ 106 bacteria.
TABLE 3.—GRAM STAIN AND COUNTS IN CONTROL URINES
Positive Gram Stain
Count
Total No.
No.
Per Cent
0
200
83
1
4
0.5
5
20
80
106-2
102-4
105-Ͼ6
25
5
32
25
play an important predisposing role in these infections.
There are as yet no distinguishing features that permit
clinical recognition of the presence of asymptomatic bacil-
luria; only bacteriologic study has proved adequate for this
purpose.
3. MANSFIELD, J. S., MALLORY, G. K., and ELLIS, L. B.: The differ-
ential diagnosis of Bright’s disease: clinicopathological corre-
lation, New England J. Med., 229, 387, 1943.
4. RHOADS, P. S., BILLINGS, C. E., and O’CONOR, V. J.: Antibacterial man-
agement of urinary tract infections. J.A.M.A., 148, 165,
1952.
5. JACKSON, G. G., DALLENBACH, F. D., and KIPNIS, G. P.: Pyelone-
phritis. Correlation of clinical and pathologic observations in
the antibiotic era. M. Clin. North America, 39, 297, 1955.
6. SANFORD, J. P., FAVOUR, C. B., MAO, F. H., and HARRISON, J. H.:
Evaluation of “positive” urine culture. An approach to the
differentiation of significant bacteria from contaminants,
Am. J. Med., 20, 88, 1956.
REFERENCES
1. BEESON, P. B.: Factors in the pathogenesis of pyelonephritis.
Yale J. Biol. & Med., 28, 81, 1955.
2. KASS, E. H.: Chemotherapeutic and antibiotic drugs in the manage-
ment of infections of the urinary tract, Am. J. Med., 18, 764, 1955.
DISCUSSION
DR. LOWELL A. RANTZ: I would like to compliment Dr. Kass for reundertaking study of a field of the greatest
clinical importance.
We too have been interested in this subject for many years and can amply confirm the observations that
asymptomatic infections of the urinary tract, particularly in women, are exceedingly common. The use of
quantitative methods for the study of organisms in the urine is extremely important. What Dr. Kass has said
about multiplication of organisms in the urine is quite correct. If you simply compare the number of organisms
per ml. in urine obtained by ureteral catheterization with that obtained simultaneously from the bladder, it will
be apparent that there is a many-fold increase during the period of retention in the bladder.
A study we did some years ago casts some light on the significance of the bacteriuria described here. If the discovery
of large numbers of bacteria in the urine in the bladder of asymptomatic females is followed by subsequent ureteral
catheterization, between one and two per cent will have many microorganisms and pus coming from one or both
ureters. This indicates that there is a fairly high incidence at all times of asymptomatic pyelonephritis among women.
Even if it is as low as one per cent, it is apparent that there are a truly enormous number of cases of this disease and
the whole matter deserves a very great deal of consideration and investigation.

424
ASYMPTOMATIC INFECTIONS OF THE URINARY TRACT
DR. EDWARD H. KASS: I just want to thank Dr. Rantz for his comments. He has promised to send me a reprint
of this fine study.
We have made a few similar observations which completely confirm the findings of Dr. Rantz. Incidentally, as
an interesting clinical sidelight, when the material obtained from the ureteral catheter has very high counts of
bacteria in it, the large numbers are usually due to the presence of hydronephrosis. One can make a diagnosis
of hydronephrosis long before the results of the radiologic study are available.
DR. WILLIAM S. TILLETT: I am reminded of an observation that the late Dr. Longcope made about infection of the
urinary tract, which I think he never published, but it is an interesting one and we, on our service, have
confirmed it. That is that he could clear up infections of the urinary tract by the administration of non-absorbable
sulfonamide drugs given by mouth.
As I say, we have confirmed this many times and I think it throws some light on the pathogenesis of some
urinary infections. I have always thought that it invited further study in that it suggests that infections in the
pelvis of the kidney might be primarily derived from the intestinal tract. Since Dr. Kass has manifested interest
in the subject I mention it because he might be interested in pursuing this lead which might add information to
the mechanism of urinary infection.
DR. J. P. SANFORD: I would like to congratulate Dr. Kass on his excellent study. As some of you may recall we
have carried out similar studies and have come to much the same conclusions.
However, our quantitative figures differ slightly from his in that we have felt that contamination occurred at
lower numbers. In other words, we felt that numbers of bacteria somewhere around 10,000 were significant for
infection.
Actually, I think, on further study, his figures come closer. The number is somewhere between 10,000 and
100,000 rather than between 1,000 and 10,000, as we have felt earlier.