The Etiology of Urinary Tract Infection:
Traditional and Emerging Pathogens
Allan Ronald, MD
The microbial etiology of urinary infections has been
regarded as well established and reasonably consis-
tent. Escherichia coli remains the predominant uro-
pathogen (80%) isolated in acute community-ac-
quired uncomplicated infections, followed by Staph-
ylococcus saprophyticus (10% to 15%). Klebsiella,
Enterobacter, and Proteus species, and enterococci
infrequently cause uncomplicated cystitis and pyelo-
nephritis.
The pathogens traditionally associated with UTI
are changing many of their features, particularly be-
cause of antimicrobial resistance. The etiology of UTI
is also affected by underlying host factors that com-
plicate UTI, such as age, diabetes, spinal cord injury,
or catheterization. Consequently, complicated UTI
has a more diverse etiology than uncomplicated UTI,
and organisms that rarely cause disease in healthy
patients can cause significant disease in hosts with
anatomic, metabolic, or immunologic underlying dis-
ease. The majority of community-acquired symptom-
atic UTIs in elderly women are caused by E coli.
However, gram-positive organisms are common,
and polymicrobial infections account for up to 1 in
3 infections in the elderly. In comparison, the most
common organisms isolated in children with un-
complicated UTI are Enterobacteriaceae. Etiologic
pathogens associated with UTI among patients with
diabetes include Klebsiella spp., Group B strepto-
cocci, and Enterococcus spp., as well as E coli. Pa-
tients with spinal cord injuries commonly have E coli
infections. Other common uropathogens include
Pseudomonas and Proteus mirabilis.
Recent advances in molecular biology may facili-
tate the identification of new etiologic agents for UTI.
The need for accurate and updated population sur-
veillance data is apparent, particularly in light of con-
cerns regarding antimicrobial resistance. This infor-
mation will directly affect selection of empiric ther-
apy for UTI. Am J Med. 2002;113(1A):14S19S. 2002
by Excerpta Medica, Inc.
From the University of Manitoba Faculty of Medicine, Winnipeg,
Manitoba, Canada.
Requests for reprints should be addressed to Allan Ronald, MD, Uni-
versity of Manitoba Faculty of Medicine, Winnipeg, Manitoba, Canada.
14S 2002 by Excerpta Medica, Inc.
All rights reserved.
T
he microbial etiology of urinary infections has for
several decades been regarded as well established,
reasonably consistent, and of limited interest.
Escherichia coli remains the predominant uropathogen
isolated. Although new pathogens have been appearing
with remarkable frequency in other illnesses, research to
identify new agents underlying unexplained urinary tract
clinical syndromes has been limited.1 Stemming from re-
cent advances in molecular biology, clues to a variety of
potential new pathogens are being identified, and expec-
tations for identifying new and important etiologic
agents for urinary syndromes will be substantial over the
next several years. In addition, the pathogens tradition-
ally associated with urinary tract infection (UTI) are
changing many of their features, particularly because of
antimicrobial resistance. As a result, empiric treatment
will undergo change during the next several years in an
attempt to limit the occurrence of resistance and prevent
its spread.
The etiology of UTI is also affected by underlying host
factors that complicate UTI, such as age, diabetes, spinal
cord injury, or catheterization. Consequently, compli-
cated UTI has a more diverse etiology than uncompli-
cated UTI, and organisms that rarely cause disease in
healthy patients can cause significant disease in hosts with
anatomic, metabolic, or immunologic underlying disease
(Table 1). This article reviews the traditional etiology of
UTI and provides an overview of the pathogenesis of E
coli and nonE coli UTIs. New diagnostic strategies and
interventions will almost certainly arise from our ex-
panding knowledge.
UNCOMPLICATED URINARY TRACT
INFECTIONS
The etiology of uncomplicated UTIs has generally re-
mained constant over the past 2 decades, albeit with a
noticeable increase in antimicrobial resistance in both
community-acquired and nosocomial UTIs. The etiology
of complicated UTI is more diverse and directly affected
by underlying host characteristics than that of uncompli-
cated UTI.
The majority of acute community-acquired, uncom-
plicated infections in the United States and abroad are
caused by E coli (80%) or Staphylococcus saprophyticus
(10% to 15%).2 Klebsiella, Enterobacter, and Proteus spe-
cies and enterococci infrequently cause uncomplicated
cystitis and pyelonephritis.3 Fungal pathogens, and par-
0002-9343/02/$20.00
PII S0002-9343(02)01055-0
 A Symposium: Traditional and Emerging Pathogens in UTI/Ronald
Table 1. Etiology of Uncomplicated Versus Complicated Uri-
nary Tract Infection (UTI)
Uncomplicated UTI
Pathogens Complicated UTI Pathogens
Escherichia coli Escherichia coli
Staphylococcus saprophyticus Klebsiella spp.
Klebsiella spp. Enterobacter cloacae
Enterococcus faecalis Serratia marcescens
Proteus mirabilis
Pseudomonas aeruginosa
Enterococcus faecalis
Group B streptococci
ticularly Candida albicans or other Candida species, ac-
count for up to 10% of positive urine cultures in tertiary
care facilities in the United States,4 but most of these are
in complicated UTIs.
Susceptibility patterns for the bacteria causing acute
uncomplicated UTIs in women have generally been pre-
dictable until recently, with most susceptible to tri-
methoprim-sulfamethoxazole (TMP-SMX). Conse-
quently, the traditional approach to therapy has been em-
piric short-course regimens with TMP-SMX.2,5
However, increasing antimicrobial resistance among
uropathogens causing acute uncomplicated cystitis has
had important ramifications for traditional empiric ap-
proaches. A recent 5-year study (1992 to 1996) investi-
gated antimicrobial susceptibility patterns among
4,000 individual UTI isolates from women with acute
uncomplicated cystitis.6 Despite no differences in the dis-
tribution of uropathogens over the 5 years, there was a
significant increase in the prevalence of resistance of E coli
to TMP-SMX (from 9% to 18%), cephalothin (from 20%
to 28%), and ampicillin (from 26% to 34%). Resistance
to nitrofurantoin and ciprofloxacin among E coli re-
mained at 1% throughout the 5 years. A 1999 nation-
wide analysis of laboratory outpatient UTI isolates7 iden-
tified ampicillin resistance among E coli nearing 40%. Re-
sistance among the fluoroquinolones remained low, at
approximately 3%.
TMP-SMX susceptibility rates to E coli range from
78% in the western United States to 88% to 89% in the
northeastern United States.7 In addition to geographic
variation, specific risk factors associated with increased
TMP-SMX resistance to E coli include current or recent
use of TMP-SMX (RR 5.1), diabetes (RR 3.1), recent hos-
pitalization (RR 2.5), and current use of any antibiotic
(RR 4.5).8 According to research from Finland, the type 1
dihydrofolate (DHFR) gene has been identified as the
most common TMP resistance gene among E coli
strains.9 Similar resistance and geographic patterns have
emerged with acute community-acquired pyelonephritis,
in that susceptibility to TMP-SMX decreased by 18% in a
10-year period, from 100% during 1985 to 19875 to 82%
July 8, 2002
during the period 1994 to 1997.10 It should be noted that
fluoroquinolone susceptibility remains high (99%).5
The increasing resistance trends are likely to have im-
portant clinical implications to the empiric use of TMP-
SMX for uncomplicated UTI and pyelonephritis. In fact,
recent research has already demonstrated significantly
greater bacteriologic and clinical cure rates for a 7-day
course of the fluoroquinolone ciprofloxacin than a 14-
day regimen of TMP-SMX for the treatment of acute un-
complicated pyelonephritis in premenopausal women.10
E coli resistance to TMP-SMX was significantly greater
(18%) than to ciprofloxacin (0%; P 0.001), and there
were significantly greater rates of clinical failures among
patients receiving TMP-SMX who had TMP-SMX resis-
tant E coli infections (65%) as compared with susceptible
E coli infections (8%). Lower clinical cure rates and di-
minished bacterial eradication have also been demon-
strated in TMP-SMXtreated women with uncompli-
cated UTI.10 Clinical failure rates of TMP-SMX for cysti-
tis increased from 3% to 13% among susceptible E coli to
27% to 40% among resistant strains.11 As a result, in geo-
graphic locations with known TMP-SMX resistance
prevalence in the range of 10% to 20%, alternative anti-
microbials, including a fluoroquinolone (3 days), nitro-
furantoin (7 days), or fosfomycin (single-dose treat-
ment), are preferred for cystitis.12
Although susceptibility is of critical importance in se-
lecting an appropriate antimicrobial agent, clinicians
must also first consider the established range of coverage
for each agent. For example, despite established suscepti-
bility, nitrofurantoin has not been shown to be effective
in patients with acute pyelonephritis. In addition, nitro-
furantoin is rarely the agent of choice in patients diag-
nosed with urinary infections other than acute cystitis. As
a result, selection of an antibacterial must consider both
susceptibility and known efficacy for the clinical syn-
drome.
Finally, approximately 25% to 35% of initial UTI epi-
sodes are followed by recurrent episodes within 3 to 6
months.13 In up to 60% of cases of recurrent UTI (RUTI),
the second infection is caused by a strain identical to that
which caused the prior infection. With E coli, the
6-month risk of a second UTI is 23.7% versus 7.7% for
nonE coli infections (P 0.02).14
In conclusion, although the etiology of UTI remains
constant, knowledge of local resistance trends is now an
integral component in the successful empiric treatment
of uncomplicated UTI. Selection of a therapeutic agent
must now take into account geographic location and in
vitro susceptibility, in addition to potential adverse ef-
fects and cost-effectiveness. Substituting a fluoroquino-
lone or nitrofurantoin for TMP-SMX in uncomplicated
UTI may be necessary.
THE AMERICAN JOURNAL OF MEDICINE Volume 113 (1A) 15S
 A Symposium: Traditional and Emerging Pathogens in UTI/Ronald
COMPLICATED UTI
Overview
Underlying host factors, including age, diabetes, spinal
cord injury, or catheterization, have a significant impact
on the etiology of UTIs. Less virulent organisms that
rarely cause disease in an anatomically or metabolically
normal urinary tract can cause significant illness and in-
vasive disease in an abnormal urinary tract. In addition,
research in Spain has identified an increase in group B
streptococcal bacteremia in nonpregnant adults during
the 19851994 time period, particularly among elderly
patients and patients with diabetes and other underlying
comorbidities.15 International studies also suggest an in-
crease in Candida spp. and Enterococcus spp. as uropatho-
gens.16
Pediatric Populations
The most common organisms isolated in children with
uncomplicated UTI are Enterobacteriaceae.17 In con-
trast, children with comorbidities are more likely to have
complicated infections, frequently caused by organisms
not otherwise associated with UTI among healthy chil-
dren. Staphylococcus aureus is more commonly seen
among children with indwelling catheters or other
sources of infection than healthy children. Coagulase-
negative staphylococci and Candida spp. are commonly
associated with infections after instrumentation of the
urinary tract.17
Nosocomial UTI (NUTI) has been associated with ure-
thral instrumentation among pediatric patients. How-
ever, a recent prospective study in Canada determined
that only 50% of pediatric NUTI had prior instrumenta-
tion.18 The most common uropathogens included E coli
(28%), Candida spp. (18%), Enterococcus (13%), gram-
negative nonfermenters (13%), Enterobacter (approxi-
mately 10%), and Pseudomonas (9.7%). It is important to
note that fungi have become the second most common
pathogen isolated in pediatric NUTI.
More than 90% of acute nonobstructive E coliin-
duced pyelonephritis in children is caused by E coli pos-
sessing P fimbriae.19 P fimbriae allow the bacteria to ad-
here to the uroepithelial cell lining; children with acute
nonobstructive pyelonephritis caused by P-fimbriated
E coli have heavy colonization of P-fimbriated E coli in
both periurethral areas and stool. In fact, this study dem-
onstrated that poor sanitary habits from hospital workers
in a neonatal ward facilitated epidemic outbreaks of acute
pyelonephritis.
Elderly
The etiology of UTI among the elderly depends on their
overall health status and whether they reside indepen-
dently in the community or require long-term care ar-
rangements. The majority of community-acquired
symptomatic uncomplicated infections in elderly women
16S July 8, 2002 THE AMERICAN JOURNAL OF MEDICINE Volume 113 (1A)
are caused by E coli.20 S saprophyticus is a rare uropatho-
gen in this subpopulation. Gram-positive organisms are
common and account for 10% to 20% of UTIs in the
elderly. Polymicrobial infections affect up to 1 in 3 pa-
tients. In contrast, the high rate of urinary catheterization
among the elderly in long-term care facilities (7%),21 in
combination with the frequent use of systemic antibiot-
ics, increases the risk of UTI caused by resistant gram-
negative rods, that is, Pseudomonas aeruginosa, Proteus
spp., and nonfermenting gram-negative rods.
Patients with Diabetes
UTIs are considered to be 1 of the top 10 concurrent or
complicating illnesses with diabetes.22 Etiologic patho-
gens associated with UTI among patients with diabetes
include Klebsiella spp., group B streptococci, and Entero-
coccus, as well as E coli. It is believed that Klebsiella and
group B streptococcus are 2 to 3 times more common
among patients with diabetes than patients without dia-
betes.22 A recent prospective study found E coli was the
most frequent causative agent among patients both with
diabetes (56.1%) and without diabetes (56.8%).23 As yet,
the reasons underlying an increased propensity to cause
UTI among most uropathogens are unknown.
Women with diabetes who have asymptomatic bacte-
riuria are at increased risk of developing symptoms.22 It
has been suggested that organisms currently found in
asymptomatic women with diabetes have the potential to
invade the renal parenchyma and cause illness. Patients
with diabetes are more likely to have UTIs by non-albi-
cans Candida species than patients without diabetes. Fi-
nally, the high rate of bladder catheterization among pa-
tients with diabetes also impacts etiologic factors.
Patients with Spinal Cord Injuries or Catheters
Patients with spinal cord injuries commonly present with
E coli infections. Other common uropathogens include
Pseudomonas and Proteus mirabilis. Patients with in-
dwelling catheters or structural abnormalities of the uri-
nary tract are more vulnerable to infection by P mirabi-
lis.24,25 Unique virulence factors produced by this organ-
ism contribute to its pathogenicity. E coli remains the
predominant causative agent. Other common pathogens
include Candida species, enterococci, P aeruginosa, Kleb-
siella spp., Enterobacter spp., and S aureus.
A recent microbiologic survey of long-term care
facilities identified numerous gram-negative isolates
(P aeruginosa, P mirabilis, E coli, and Klebsiella pneu-
moniae), as well as gram-positive isolates (enterococci
and S aureus).21 Approximately 7% of residents had uri-
nary catheters; 38% had recently taken, or were currently
receiving, antibiotics, facilitating significant antibiotic
pressures within this population.
Results from a surveillance database in a Midwest uni-
versity hospital covering 1982 through 1991 demon-
strated a significant increase in NUTI, from 2.63 of 1,000
 A Symposium: Traditional and Emerging Pathogens in UTI/Ronald
Table 2. Inherited or Acquired Host Susceptibility Factors
Genetic Biologic
Blood group antigen Congenital abnormalities
Nonsecretor status Urinary obstruction
Increased density Calculi
adhesion receptors Diabetes mellitus
Anatomy
Residual urine
Atrophic vaginitis
Urinary incontinence
Prior history of UTI
Maternal history of UTI
Childhood history of UTI
Catheters/stents/foreign materials
Condom catheters
Immunologic abnormalities (HIV)
Renal transplant
HIV human immunodeficiency virus; UTI urinary tract infection.
patient-days to 4.35 of 1,000 patient-days (P 0.0023).26
An overwhelming majority (88%) of NUTIs were cathe-
ter related. There was a reduction in the prevalence of
UTI caused by E coli, Proteus spp., and Pseudomonas spp.,
and an increase in UTI caused by yeasts, K pneumoniae,
and group B streptococcus.
PATHOGENESIS OF UTI
Individual susceptibility to UTI is complex, depending
on genetic, biologic, and behavioral factors (Table 2).
The interaction between bacterial virulence and host de-
fense factors can ultimately result in UTI. Each bacterial
species has distinct pathogenic mechanisms that facilitate
UTI. Colonization is determined by the specific bacterial
adhesive characteristics, the receptor repertoire on the
epithelial surface, and the surrounding fluids. Vaginal
colonization of uropathogens frequently precedes UTI.
Genotypic traits for epithelial cell receptivity have been
identified as an important susceptibility factor in UTI.27
Pathogenesis of E coli Infections
A critical initial step in the pathogenesis of both acute and
recurrent UTI involves colonization of the urinary tract
or vaginal introitus with E coli. Uropathogenic E coli
strains generally possess filamentous surface adhesive or-
ganelles called fimbriae or pili. In the case of type 1 fim-
briae, the adhesin molecule FimH, which is located at the
tip of the type 1 pilus, directly interacts with host recep-
tors and facilitates bacterial attachment on the luminal
surface of the bladder epithelium. This may also mediate
the internalization of bacteria into epithelial cells, where
E coli can replicate and escape host defense mechanisms.
Research has shown that women with RUTI have 3-fold
more E coli adhering to vaginal, buccal, and voided uro-
epithelial cells than women without recurrent infection.
July 8, 2002
Behavioral Other
Sexual intercourse sMental status
Use of diaphragm
Use of spermicides
Antimicrobial use
Further, women with RUTI also have longer durations of
vaginal colonization with uropathogenic E coli,28 even
during asymptomatic periods.29,30
There is evidence for an intestinal habitat for uro-
pathogenic bacteria,31 as strains of uropathogenic E coli
have been found in the colonic microflora. In addition,
new molecular assays have identified 29 virulence factor
genes of E coli.32 Research comparing UTI among non-
immunocompromised, immunocompromised, and re-
nal disease patients found E coli was the most frequently
isolated pathogen among all groups.33 All virulence fac-
tors of E coli were more common among nonimmuno-
compromised versus immunocompromised patients; the
highest degree of discrimination between the groups in-
volved genes for bacterial adhesive proteins. It was sug-
gested that less virulent E coli strains can cause UTI more
frequently in immunocompromised patients or patients
with renal disease. Additional research also found low
virulence strains to be prevalent in patients with diabetes
and signs of renal complications.34
A limited number of P-fimbriated E coli strains have
been strongly associated with pediatric pyelonephritis.
Each E coli clone has distinct characteristics, including
O and K serotype, similar outer membrane protein pat-
terns, and hemolysins. Some O and K serotypes of E coli
are more common uropathogens than others. Neverthe-
less, P fimbriation appears to be a virulence factor. Fur-
ther, there appears to be a correlation between coloniza-
tion of the gastrointestinal tract with P-fimbriated E coli
and acute pyelonephritis.19
A recent surveillance study in Spain assessed the etiol-
ogy of community-acquired UTI for the E coli serotype
O15:K52:H1 over a 1-year period.35 This specific serotype
accounted for only 1.4% of community-acquired E coli
UTIs, but these infections were more likely to present as
THE AMERICAN JOURNAL OF MEDICINE Volume 113 (1A) 17S
 A Symposium: Traditional and Emerging Pathogens in UTI/Ronald
pyelonephritis. E coli O15:K52:H1 strains had a limited
repertoire of virulence factors, but extensive and diverse
antimicrobial resistance profiles.
There are, as yet, no prospective population-based
studies on the pathogenesis of UTI among patients with
diabetes. In some patients with diabetes, Tamm-Horsfall
protein is markedly reduced, allowing adherence and cell
entry of type 1 fimbriated E coli.22 Whereas P fimbriation
and hemolysin production have been identified as impor-
tant virulence factors among patients with diabetes,33
there is, as yet, no evidence identifying differences in bac-
terial virulence factors for E coli among patients with or
without diabetes.
Pathogenesis of NonE coli Infections
In contrast to the pathogenesis of E coli UTI, S saprophy-
ticus and enterobacterial species adhere to uroepithelial
cells through different adhesive mechanisms. After at-
tachment, Proteus spp., K pneumoniae, and S saprophyti-
cus each produce urease, which catalyzes the hydrolysis of
urea in urine and causes the release of ammonia and CO2.
As a result, the urinary pH is elevated, which can eventu-
ally lead to the formation of bladder or kidney stones.
CONCLUSIONS
The majority of UTIs are uncomplicated infections. UTIs
are considered complicated when patients have func-
tional, metabolic, or structural abnormalities. Through-
out the recent past, the most common causative patho-
gens associated with uncomplicated cystitis or acute pye-
lonephritis have been, and remain, E coli (80%) and S
saprophyticus (5% to 15%). In comparison, the etiology
of complicated UTI is more diverse and frequently
polymicrobial in nature. Specific host factors, such as the
extremes of age (pediatric or elderly), pregnancy, diabe-
tes, catheterization, or spinal cord injuries, can impact on
etiology. Patients with diabetes are more likely to have
infections involving Klebsiella spp., whereas patients with
urinary catheters are more frequently diagnosed with
Pseudomonas infection. Although there have been mini-
mal changes in the predominant uropathogens over the
past decades, there have been significant changes in the
resistance patterns to antimicrobials. The changes in re-
sistance patterns need to be considered when determin-
ing the most appropriate empiric therapy. This is partic-
ularly important given the increasing resistance to TMP-
SMX.
Recent advances in molecular biology may facilitate
the identification of new etiologic agents for UTI. A re-
newed interest in the etiology and management of UTI
has surfaced over the past few years. The need for accurate
and updated population surveillance data is apparent,
particularly in light of concerns regarding antimicrobial
resistance. This information will ultimately direct selec-
tion of empiric therapy for UTI.
18S July 8, 2002 THE AMERICAN JOURNAL OF MEDICINE Volume 113 (1A)
REFERENCES
1. Ronald AR, Nicolle LE, Stamm WE, et al. Urinary tract
infection in adults: research priorities and strategies. Int J
Antimicrob Agents. 2001;17:343348.
2. Stamm WE, Hooton TM. Management of urinary tract in-
fections in adults. N Engl J Med. 1993;329:1328 1334.
3. Hooton TM. Practice guidelines for urinary tract infections
in the era of managed care. Int J Antimicrob Agents. 1999;
11:241245.
4. Sobel JD, Kauffman CA, McKinsey D, et al. Candiduria: a
randomized, double-blind study of treatment with flucon-
azole and placebo. Clin Infect Dis. 2000;30:19 24.
5. Hooton TM, Stamm WE. Diagnosis and treatment of un-
complicated urinary tract infection. Infect Dis Clin North
Am. 1997;11:551581.
6. Gupta K, Scholes D, Stamm WE. Increasing prevalence of
antimicrobial resistance among uropathogens causing
acute uncomplicated cystitis in women. JAMA. 1999;281:
736 738.
7. TSN Database, MRL Pharmaceutical Services.
8. Wright SW, Wrenn KD, Haynes ML. Trimethroprim-sulfame-
thoxazole resistance among urinary coliform isolates. J Gen
Intern Med. 1999;14:606 609.
9. Heikkila E, Renkonen OV, Sunila R, et al. The emergence
and mechanisms of trimethroprim resistance in Escherichia
coli isolated from outpatients in Finland. J Antimicrob Che-
mother. 1990;25:275283.
10. Talan DA, Stamm WE, Hooton TM, et al. Comparison of
ciprofloxacin (7 days) and trimethoprim-sulfamethoxazole
(14 days) for acute uncomplicated pyelonephritis in women:
a randomized trial. JAMA. 2000;283:15831590.
11. McCarty JM, Richard G, Huck W, et al. A randomized trial
of short-course ciprofloxacin, ofloxacin, or trimethoprim/
sulfamethoxazole for the treatment of acute urinary tract
infection in women: Ciprofloxacin Urinary Tract Infection
Group. Am J Med. 1999;106:292299.
12. Warren JW, Abrutyn E, Hebel JR, et al. Guidelines for
antimicrobial treatment of uncomplicated urinary tract in-
fection. Infect Dis Clin North Am. 1997;11:551581.
13. Mabeck CE. Treatment of uncomplicated urinary tract in-
fection in nonpregnant women. Postgrad Med J. 1972;48:
69 75.
14. Foxman B, Gillespie B, Koopman J, et al. Risk factors for
second urinary tract infection among college women. Am J
Epidemiol. 2000;151:1194 1205.
15. Munoz P, Llancaqueo A, Rodriguez-Creixems M, et al.
Group B streptococcus bacteremia in nonpregnant adults.
Arch Intern Med. 1997;157:213216.
16. Khan SW, Ahmed A. Uropathogens and their susceptibility
pattern: a retrospective analysis. J Pak Med Assoc. 2001;
51:98 100.
17. Schlager TA. Urinary tract infections in children younger
than 5 years of age: epidemiology, diagnosis, treatment,
outcomes and prevention. Paediatr Drugs. 2001;3:219
227.
18. Langley JM, Hanakowski M, Leblanc JC. Unique epidemi-
ology of nosocomial urinary tract infection in children. Am J
Infect Control. 2001;29:94 98.
19. Tullus K, Horlin K, Svenso SB, Kallenius G. Epidemic out-
breaks of acute pyelonephritis caused by nosocomial
spread of P fimbriated Escherichia coli in children. J Infect
Dis. 1984;150:728 736.
20. Beyer I, Mergan A, Benoit F, Theunissen C, Pepersack T.
Management of urinary tract infections in the elderly. Z
Gerontol Geriatr. 2001;34:153157.
 A Symposium: Traditional and Emerging Pathogens in UTI/Ronald
21. Smith PW, Seip CW, Schaefer SC, Bell-Dixon C. Microbi-
ologic survey of long-term care facilities. Am J Infect Con-
trol. 2000;28:8 13.
22. Ronald A, Ludwig E. Urinary tract infections in adults with
diabetes. Int J Antimicrob Agents. 2001;17:287292.
23. Bonadio M, Meini M, Gigli C, et al. Urinary tract infection in
diabetic patients. Urol Int. 1999;63:215219.
24. Mobley HL, Island MD, Massad G. Virulence determinants
of uropathogenic Escherichia coli and Proteus mirabilis.
Kidney Int Suppl. 1994;47:S129 S136.
25. Coker C, Poore CA, Li X, Mobley HL. Pathogenesis of
Proteus mirabilis urinary tract infection. Microbes Infect.
2000;2:14971505.
26. Bronsema DA, Adams JR, Pallares R, Wenzel RP. Secular
trends in rates and etiology of nosocomial urinary tract infec-
tions at a university hospital. J Urol. 1993;150:414 416.
27. Schaeffer AJ, Rajan N, Cao Q, et al. Host pathogenesis in
urinary tract infections. Int J Antimicrob Agents. 2001;17:
245251.
28. Stapleton A. Prevention of recurrent urinary-tract infections
in women. Lancet. 1999;353:7 8.
29. Stamey TA, Sexton CC. The role of vaginal colonization
with Enterobacteriaceae in recurrent urinary tract infec-
tions. J Urol. 1975;113:214 217.
July 8, 2002
30. Stamey TA, Timothy MM. Studies of introital colonization in
women with recurrent urinary infections. 1. The role of
vaginal pH. J Urol. 1975;114:261263.
31. Goetz GS, Mahmood A, Hultgren SJ, et al. Binding of pili
from uropathogenic Escherichia coli to membranes se-
creted by human colonocytes and enterocytes. Infect Im-
mun. 1999;67:6161 6163.
32. Johnson JR, Stell AL. Extended virulence genotypes of
Escherichia coli strains from patients with urosepsis in re-
lation to phylogeny and host compromise. J Infect Dis.
2000;181:261272.
33. Karkkainen UM, Ikaheimo R, Katila ML, et al. Low virulence
of Escherichia coli strains causing urinary tract infection in
renal disease patients. Eur J Clin Microbiol Infect Dis. 2000;
19:254 259.
34. Brauner A, Katouli M, Ostenson CG. P-fimbriation and hae-
molysin production are the most important virulence fac-
tors in diabetic patients with Escherichia coli bacteraemia:
a multivariate statistical analysis of seven bacterial viru-
lence factors. J Infect. 1995;31:2731.
35. Prats G, Navarro F, Mirelis B, et al. Escherichia coli sero-
types O15:K52:H1 as a uropathogenic clone. J Clin Micro-
biol. 2000;38:201209.
THE AMERICAN JOURNAL OF MEDICINE Volume 113 (1A) 19S