Ask the Expert
Are urine cultures always necessary
before initiating therapy for urinary tract infections? What is appropriate initial therapy?
Stephen A. Lerner, MD
Wayne State University School of Medicine, Detroit, Michigan, USA
The etiology and treatment of urinary tract infections depends on
the sex and age of the patient, the clinical presentation of the infection, a history of prior urinary tract infections,
the patientís recent history of antibiotic therapy and sexual activity, and the presence or absence of complicating
conditions that are known to promote infection, account for persistence, or lead to recurrence. Asymptomatic bacteriuria
warrants antimicrobial treatment in children, pregnant women, and some immunosuppressed patients. However, most
patients without symptoms do not require antibiotic therapy. Therefore, I shall limit my comments to patients who
present with symptoms of urinary tract infection, such as dysuria, urgency, frequency, subrapubic pain, or hematuria.
The definition of urinary tract infection obviously includes bacteriuria.
However, the titer of bacteria in a urine specimen accepted as clinically significant has recently been revised.
Early classic studies indicated that a finding of more than 105 uropathogenic bacteria per ml of urine differentiated a true bladder infection
from contamination in women with pyelonephritis. More recent studies, however, have shown that cystitis, and even
pyelonephritis, can be caused in women by uropathogenic bacteria at titers in the range of 102-104/ml.1-3
In patients with upper urinary tract infection (i.e., pyelonephritis)
or complicating conditions (e.g., structural or functional urologic abnormalities, prior urinary catheterization
or urologic surgery, or in males), the spectrum of possible causative organisms is broad, and their antimicrobial
susceptibilities are relatively unpredictable. For such patients, therefore, a urine culture prior to therapy is
recommended to guide any necessary adjustment in therapy. On the other hand, in women with uncomplicated acute
cystitis, the predominant pathogens (Escherichia coli and Staphylococcus
saprophyticus) and their susceptibilities to the antimicrobial
agents most commonly used are sufficiently predictable so that therapy can be started without doing a urine culture.
Even some of the species of Enterobacteriaceae that are encountered less frequently in this setting are generally susceptible
to these agents.
Rapid non-culture diagnostic tests can provide additional support
of a presumptive diagnosis of urinary tract infection in any patient for whom treatment is considered. This approach
is especially important with patients who are presumed to have cystitis, because urine cultures may not be necessary
or cost effective. Pyuria as detected by the rigorous but rapid and inexpensive technique of quantitative microscopy
(greater than 10 white cells per l of uncentrifuged urine in a hemocytometer)5 is a sensitive indicator of urinary tract infection.4 When microscopy is not available, the leukocyte esterase test can be used
as a somewhat less sensitive alternative.4 Hematuria detected microscopically or bacteria visible in a gram-stained
smear of uncentrifuged urine (bacteriuria) is a less sensitive indictator than pyuria, but both are highly specific
for urinary tract infection.4
Thus, in a patient with urinary symptoms and no complicating factors, finding pyuria, hematuria, or bacteriuria
microscopically strongly indicates acute cystitis, and a culture need not be obtained.4
For such patients with uncomplicated cystitis, a 3-day course of
trimethoprim-sulfamethoxazole provides highly effective therapy.6 For patients with a history of allergy to sulfonamides or trimethoprim or
any other contraindication to either of these drugs, alternatives must be considered. The effectiveness of amoxicillin
(ampicillin) has been seriously compromised by the rise of resistance to these agents, even among E. coli isolates from
community-acquired urinary tract infections. Furthermore, short-course therapy with amoxicillin (and other F-lactams)
may be less effective than with agents such as trimethorim-sulfamethoxazole in eradicating E.
coli from the vaginal reservoir, so relapses may be more
Treatment of acute cystitis with a fluoroquinolone, such as ciprofloxacin, for 3 days is also highly effective.6 On the other hand,
the serious potential for selection of resistance to this class of drugs, which is critically important for the
treatment of more serious infections, provides strong incentive to limit their routine use for the treatment of
cystitis. Slightly less effective but certainly adequate alternatives include a 7-day course of nitrofurantoin
or a single dose of the recently approved fosfomycin tromethamine (MonurolTM, Forest Pharmaceuticals, Inc., St. Louis, MO).7 Single-dose antibiotic therapy has the advantages of greater compliance,
fewer adverse side effects, and less antibiotic exposure, which can select for resistance. Although resistance
to fosfomycin has been seen, there is as yet no evidence in clinical isolates of cross-resistance with other classes
of antibiotics. Because this class of drugs is used (in the U.S.) only for the indication of acute uncomplicated
cystitis and nitrofurantoin is used only for the treatment of lower urinary tract infection, emergence of resistance
to fosfomycin or nitrofurantoin is unlikely to compromise the effectiveness of antibiotic therapy for more serious
- Stamm WE. 1984. Eur J Clin Microbiol 3:279.
- Bollgren I, Engstrom CF, Hammarlind M et al. 1984.
Arch Dis Child 59:102.
- Johnson JR, Stamm WE. 1987. Infect Dis Clin No Am 1:773.
- Johnson JR, Stamm WE. 1989. Ann Intern Med 111:906.
- Rubin RH, Shapiro ED, Andriole VT et al. 1992. Clin
Infect Dis 15(suppl 1):216.
- Stamm WE. 1992. In Infectious Diseases, edited by SL
Gorbach, JG Bartlett, NR Blacklow. Philadelphia, PA: WB Saunders, 788.
- Reeves DS. 1995. Rev Contemp Pharmacother 6:71.