Urinary tract infection (UTI) may be the most common bacterial infection of humans. The syndrome is seen far more frequently in women than in men, affecting one in two women at least once in her lifetime. UTIs account for more than seven million office visits and one million visits to the emergency department in the United States each year; they are responsible for or complicate more than one million hospital admissions, and their diagnosis and treatment costs $1.6 billion a year. This chapter will focus on the ambulatory management of urinary tract infections.

The term UTI is a general one. For our purposes we will use it to indicate cystitis, an infection of the urinary bladder with or without involvement of immediately adjacent structures. The most common presenting symptom of a UTI is dysuria; it is important to remember that there are both infectious and noninfectious causes of dysuria. Noninfectious dysuria can be caused by chemical or physical trauma, as well as by atrophic vaginitis in post-menopausal women. Infectious causes include cystitis, urethritis and vaginitis. Symptoms of acute cystitis include acute dysuria, increased urinary frequency and urgency. Low back pain and suprapubic tenderness on exam may also be present. Patients with urethritis tend to present with a more gradual onset of mild dysuria and may have a vaginal discharge or bleeding (with concomitant cervicitis). They may also report a new sexual partner or have vulvovaginal lesions or cervicitis on exam. Patients with vaginitis often present with a vaginal discharge and complain of pruritis or odor; increased frequency and urgency are usually absent. These clinical syndromes are sufficiently different to permit the clinician to make a preliminary diagnosis based on history and physical exam. In a recent review, the likelihood ratio of UTI in a woman presenting with dysuria and frequency without vaginal discharge or irritation was 24.6.

The gold standard for the diagnosis of UTI is a positive urine culture, defined as the presence of at least 10⁵ colony-forming units (CFU)/ml of a pathogenic bacterium isolated from urine culture, although some studies suggest that 10³ CFU/ml is clinically significant in symptomatic men. Pyuria is present in almost all UTIs, and its absence should strongly suggest another diagnosis. In female patients with typical symptoms of acute cystitis, the presence of pyuria by urine dipstick testing is all that is required for a presumptive diagnosis and a urine culture is usually unnecessary. In men or women with complicated or recurrent UTIs (see below), a pretreatment culture should be obtained.

While the most accurate method of measuring pyuria is the leukocyte excretion rate, this test is impractical and is seldom performed. The presence of at least 10 WBC/mm³ of unspun urine (measured by hemocytometer) correlates well with significant leukocyte excretion rates and is considered the standard for the determination of significant pyuria.

Urine dipstick testing for the presence of leukocyte esterase is a fast and relatively reliable way to identify pyuria. Leukocyte esterase is present in primary neutrophil granules and acts as an indirect measure of the presence of activated WBCs. Dipstick sensitivity ranges from 75 to 96 percent and specificity ranges from 94 to 98 percent, giving the test a negative predictive value of 92 percent in some studies.

The direct microscopic examination of urine is inconvenient to perform and difficult to reproduce. However, the presence of at least one organism/oil immersion field in a Gram stained, unspun urine sample correlates well with a positive urine culture. Dipstick testing for the presence of nitrite is also available; false negative tests are common. As discussed above, the gold standard test is a positive urine culture. If urine is to be tested for bacteriuria, a clean-catch (or catheterization) specimen should be obtained. Women should wash the external genitalia three times with a cleansing agent and should collect the urine into a sterile container while in midstream. Washing the urethral meatus in men is recommended, but is of unproven necessity. Urine specimens should be refrigerated if not plated within two hours.

Escherichia coli is responsible for 80 to 85 percent of community-acquired UTIs and Staphylococcus saphrophyticus accounts for 10 to 15 percent. Other less frequent pathogenic organisms include Enterococcus, Klebsiella and Proteus species. Inpatient UTIs are caused by a much broader spectrum of organisms and will not be discussed here.

The initial evaluation of a patient with a urinary tract infection should focus on dividing complicated from uncomplicated infection. The term "complicated UTI" indicates that a functionally, metabolically, or anatomically abnormal urinary tract is present, or that a UTI is caused by an organism known to be resistant. Common complicating factors include urinary tract instrumentation or catheterization, diabetes mellitus, pregnancy, immunosuppression, uretero-vesical reflux or other urologic abnormalities, obstructive uropathy and azotemia.

The clinical presentation of cystitis in women usually consists of the classic triad of dysuria, urgency and frequency, and the organisms causing acute uncomplicated cystitis in women are highly predictable. This permits presumptive diagnosis and empiric treatment. In women with typical symptoms and pyuria by dipstick testing, no urine culture is necessary. A short course of antibiotics should be prescribed, and no follow-up appointment is needed unless symptoms persist or recur.

Uncomplicated cystitis may be effectively treated with a wide variety of antibiotics. Most guidelines recommend the use of TMP/SMX, based on urinary drug excretion rates, safety, cost and effects on vaginal flora. Recent surveys, however, indicate increasing rates of resistance among urinary pathogens. A large study of uncomplicated UTI among women at an HMO showed that resistance to TMP/SMX among E. coli isolates rose from 9 percent in 1992 to 18 percent in 1996. Rates of resistance to ampicillin and cephalothin were also rising. This study, of course, could not include women in whom no urine cultures were sent. Another study of antibiotic resistance among patients in the emergency department of a tertiary care center demonstrated TMP/SMX resistance in 15 percent of all coliform isolates; diabetes, recent hospitalization and use of antibiotics were independent risk factors for antimicrobial resistance. A retrospective cohort study of women with cystitis seen in ambulatory care settings in Michigan noted that TMP/SMX resistance rates in E. coli isolates rose from 8.1 percent in 1992 to 15.8 percent in 1999. The risk of resistance was 16 times higher in women who had recently taken TMP/SMX.

Based on these and similar data, it is reasonable to reconsider the management of acute uncomplicated UTI in women, balancing concerns about the creation of fluoroquinolone resistance and efficacy of TMP/SMX. In the past, fluoroquinolones were reserved for patients with recurrent infections, treatment failures, known resistant organisms or allergies to alternate antibiotics. Guidelines from the Infectious Diseases Society of America, published in 1999, expanded these indications to include women living in communities where the rate of resistance to TMP/SMX is greater than 20 percent. We recommend modifying this approach; fluoroquinolones should also be used as first-line agents in women with a history of recent antibiotic use.

There have been numerous studies comparing seven-day, three-day and single- dose regimens. For most antibiotics, three-day regimens are optimal, with similar cure rates, lower cost and fewer side effects when compared to seven-day regimens. Single dose therapy is effective with some agents, but cure rates are lower and recurrence more frequent. When using beta-lactams, seven-day regimens are required to produce optimal cure rates.

In summary, most women presenting with the typical symptoms of cystitis, pyuria by dipstick, no complicating features and no sulfa allergies should be treated empirically with three days of TMP/SMX (Bactrim DS bid x three days). Trimethoprim alone can be used in sulfa-allergic patients. Non-pregnant women with recent antibiotic use or history of resistant organisms should be treated with ciprofloxacin (250 mg bid x 3 days). Patients with unresolved symptoms after appropriate treatment should return to clinic for urine cultures. Those who were initially treated with TMP/SMX may be switched to a fluoroquinolone while culture results are pending.

Cystitis is far more common in women than in men, and most of the research on UTIs has focused on women. In the past, UTIs in men were considered "complicated" by definition, as they were felt to occur only in the presence of an anatomic abnormality. However, several studies have shown that anal intercourse, lack of circumcision and intercourse with a female partner whose vaginal flora is colonized with uropathogens can predispose to cystitis in young men. Symptoms of cystitis in men can be irritative (frequency, urgency, dysuria) or obstructive (hesitancy, dribbling, poor stream). The presence of urethral discharge points to urethritis rather than cystitis.

In a man with suspected cystitis, pretreatment urine cultures should be obtained. In symptomatic men, 10³ CFU/ml is considered significant bacteriuria. A seven day course of antibiotics is recommended - it is reasonable to start with empiric treatment such as TMP/SMX (Bactrim DS bid x seven days) and to adjust based on culture results if necessary. Urologic evaluation can be reserved for men with recurrent infection.

Recurrences are defined as two or more infections in a six-month period, or at least three infections in a 12-month period. It is estimated that 20 percent of young women with an initial bout of cystitis have recurrences, and that 90 percent of these are due to repeat exogenous infections. These patients rarely have anatomic or functional abnormalities of the urinary tract, making imaging studies and cystoscopy of little help. Colonization of the vaginal introitus and peri-urethral areas with uropathogenic strains of bacteria has been implicated in recurrent infection in women. The responsible organisms can be treated with the same regimens used for uncomplicated cystitis.

Women with recurrent cystitis should have urine cultures performed, and attention should be paid to the frequency and setting of cystitis. If cystitis is temporally linked to sexual intercourse and is very frequent, post-coital prophylaxis with single strength TMP/SMX or nitrofurantoin is recommended. Post-coital urinary voiding may also be helpful. Patients with frequent recurrent infection not temporally associated with intercourse may require continuous prophylaxis. Regimens include single strength TMP/SMX, nitrofurantoin 50 mg or cephalexin 250 mg given daily or three times a week for six to 12 months.

Alternatives to prophylaxis include intermittent patient-initiated therapy, where medication is kept at home and a single dose or three-day regimen is started when the patient recognizes typical symptoms of cystitis. In post-menopausal women, atrophic vaginitis and the alteration of vaginal flora may contribute to recurrent infection; topical vaginal steroid or estrogen creams may be of benefit.

A complicated UTI indicates a functionally, anatomically or metabolically abnormal urinary tract, or the presence of a suspected resistant pathogen. Urine cultures are essential in this setting because the list of potential pathogens is long, and antibiotic resistance is common. Patients require longer courses of therapy, usually lasting at least 10 to 14 days. In the absence of systemic illness, these infections can be treated in the ambulatory setting, although careful follow-up is essential. While the urine culture is pending, treatment should be initiated with a broad-spectrum antibiotic, such as a fluoroquinolone. Once susceptibilities are known, antimicrobials can be tailored to a narrower spectrum.

Patients with acute uncomplicated pyelonephritis typically present with symptoms of cystitis, fever, chills and flank pain. Symptoms of nausea, vomiting and malaise are also common. Acute pyelonephritis can be life-threatening, and sequelae of untreated infection include renal scarring, impaired renal function, renal abscess formation and sepsis. In reliable patients with mild symptoms, without nausea and vomiting, signs of significant dehydration or bacteremia, outpatient therapy is an option. Severe symptoms, inability to tolerate oral medications and/or orthostasis are indications for admission and intravenous antibiotics.

Uropathogenic strains of E. coli are the most common cause of this syndrome, but urine cultures should be performed in all patients suspected of having pyelonephritis. Determination of pyuria, bacteriuria and hematuria can also assist in the in-office diagnosis. Blood cultures are not indicated in the outpatient setting, but should be performed in all patients sick enough to require admission.

Initial outpatient therapy should be a 10 to 14 day course of TMP/SMX (or a fluoroquinolone in the case of sulfa allergy, diabetes, or recent antibiotic use). Close follow-up is required. Symptoms should resolve within 72 hours - if they do not, a complicated infection may be present. Repeat cultures and radiologic imaging with CT or ultrasound should be performed to rule out abscess formation, obstruction or the presence of calculi. Follow-up culture two weeks after the completion of therapy is generally recommended.

Diabetic patients comprise a large proportion of our outpatient population and deserve special attention. Diabetics are more prone to UTIs and to upper urinary tract infections. The reason for this predisposition is not completely understood, but the most important factor is likely to be bladder dysfunction caused by diabetic neuropathy. Studies are limited, but many experts believe that asymptomatic bacteriuria in diabetics should be treated because of the frequency and severity of upper urinary tract infections in these patients.

A diabetic patient with symptoms of cystitis should be managed differently than a non-diabetic. Urine cultures should be routinely obtained prior to treatment, and a two-week antibiotic course is recommended. As noted, increasing TMP/SMX-resistance has changed our recommendations in favor of a fluoroquinolone such as ciprofloxacin. Follow-up cultures two weeks after the completion of therapy are also indicated.

Urinary tract infections in the elderly are common, and the prevalence of asymptomatic bacteriuria rises with age in both men and women. Acute uncomplicated cystitis in an elderly patient can be handled exactly as recommended in younger patients. The significance of asymptomatic bacteriuria in the elderly has been the subject of great debate. Initial series demonstrated a link between asymptomatic bacteriuria and mortality in nursing home patients, but subsequent studies suggest that underlying disease processes may have confounded this observation. Current thinking is that therapy for asymptomatic bacteriuria should be reserved for immunosuppressed patients and those undergoing genitourinary instrumentation.