Community-acquired urinary tract infections (UTIs) account for approximately 8 million ambulatory visits and 1 million hospitalizations each year in the United States, making them one of the most common infections for which an antibiotic is prescribed in clinical practice. Another 1 million nosocomial UTIs are diagnosed annually, primarily indwelling urinary catheter–associated UTIs, accounting for an estimated 40% of all health care–associated infections (see Health Care–Associated Infections). Approximately half of all women experience a UTI by age 30 years; sexual activity is a major risk factor. Approximately 5% of otherwise healthy women who experience a UTI are at greater risk of developing future infections. Other UTI risk factors include structural and functional abnormalities, use of spermicidal agents and diaphragms, pregnancy, diabetes mellitus, obesity, urethral catheterization (or other urinary tract instrumentation), immunosuppression, and genetic factors.
UTIs are classified based on anatomic location as lower (cystitis), upper (pyelonephritis, perinephric abscess), or prostatitis. The term uncomplicated UTI refers to infections in nonpregnant women without structural or neurologic abnormalities or comorbidities. UTIs in men, pregnant women, and persons with foreign bodies (for example, indwelling catheters, calculi), kidney disease, immunocompromise, obstruction, urinary retention from neurologic disorders, health care–associated infections, or recent antibiotic use are considered to be complicated. Advanced age in the presence of other major comorbidities or with significant frailty may be considered a complicating factor in UTI, although age alone does not define a complicated versus uncomplicated infection. Designating an infection as complicated influences the choice and duration of antimicrobial therapy and extent of investigation. Nevertheless, the potential for uncomplicated UTIs to evolve into clinically severe disease should not be underestimated, nor should the urgency or seriousness of complicated UTIs be overstated.
Most infections occur by the ascending route. In 95% of these cases, UTIs are caused by a single bacterial species, mainly gram-negative aerobic bacilli originating from the bowel. Uropathogenic Escherichia coli accounts for 75% to 95% of UTIs in women. Less common urinary pathogens include other members of the Enterobacteriaceae family, streptococci (in particular Streptococcus agalactiae), enterococci, and staphylococci (most often Staphylococcus saprophyticus). UTIs occurring in hospitals and long-term care facilities frequently involve a more varied group of organisms (such as Enterobacter, Providencia, Morganella, Citrobacter, Serratia, and Pseudomonas). Isolation of Staphylococcus aureus from the urine may be related to instrumentation but should suggest the possibility of a hematogenous infection from a source outside the urinary tract.
In persons with symptoms of UTI, diagnosis in the outpatient setting is based on a combination of clinical features, determining if the presumed infectious process is in the lower or upper urinary tract, and the findings of significant pyuria (≥10 leukocytes/µL) and bacteriuria (bacteria in the urine). Pyuria can be detected by urine dipstick, which relies on the presence of leukocyte esterase. Although the sensitivity and specificity of dipstick testing are high (about 75% and 85%, respectively), pyuria may result from urinary tract disorders other than infection. The presence of leukocyte casts supports a diagnosis of pyelonephritis. Microscopic or gross hematuria may be present with a UTI but may also be encountered with nephrolithiasis and tumors. A positive nitrite test result signifies the presence of gram-negative bacteria capable of converting nitrates into nitrites but is negative in UTI caused by nonconverting organisms (Enterococcus, Staphylococcus, or Streptococcus species).
Quantitative cultures of a midstream, clean-void urine sample are the most accurate way to demonstrate bacteriuria in patients with suspected UTI. Because the microbiology is predictable and treatment courses are short, culture is not recommended in women with uncomplicated cystitis. Urine cultures are indicated in pyelonephritis, complicated cystitis, and recurrent UTIs; additionally, they are recommended in patients with histories of multiple antibiotic allergies and in those in whom the presence of a resistant organism is suspected (such as recent antibiotic treatment, health care–associated infection, previous multidrug-resistant UTI). The growth of 105 colony-forming units (CFU)/mL of urine is considered significant bacteriuria; however, lower CFU counts support a diagnosis in those with UTI symptoms.
In most adults, imaging studies are not required for diagnosis or treatment of UTIs. Imaging may be indicated when the diagnosis is unclear, when a structural abnormality or complication is suspected, or in patients with severe illness, immunocompromise, or lack of response to appropriate therapy. Ultrasonography can detect obstruction, whereas noncontrast helical CT is recommended for visualizing kidney stones. Although less sensitive than CT, kidney ultrasonography is less expensive, requires less radiation exposure, and can be used in pregnant women or if CT is unavailable. Contrast-enhanced CT (CT urography) is recommended when intrarenal or perinephric abscess is suspected.
Asymptomatic bacteriuria is defined as the presence of at least 105 CFU/mL of a uropathogen from two consecutive voided urine specimens in women or one specimen in men, or more than 102 CFU/mL of one bacterial species from a catheterized urine specimen in women or men, in all cases without local or systemic signs or symptoms of active infection. The prevalence of asymptomatic bacteriuria is as low as 1% to 5% in healthy premenopausal women (2%-10% in pregnant women) and nearly 100% in patients with long-term indwelling urinary catheters.
The presence of pyuria accompanying asymptomatic bacteriuria is not an indication for antimicrobial treatment. Although bacteriuria increases the risk of symptomatic UTI, treatment of asymptomatic bacteriuria neither decreases the frequency of symptomatic infections nor improves other outcomes. Inappropriate treatment of asymptomatic bacteriuria is a major driver of antimicrobial resistance, particularly in health care facilities. Treatment of asymptomatic bacteriuria is, however, indicated in pregnant women and in patients scheduled to undergo an invasive procedure involving the urinary tract (Table 70).
Recommended first-line antibiotic regimens for uncomplicated cystitis (urinary frequency and urgency, dysuria, and suprapubic discomfort) should consider the increased rate of antimicrobial resistance of E. coli, the efficacy and advantages of short-course therapies, and the potential adverse effects (of the ecology and on patients). Preferred agents include nitrofurantoin (5 days), trimethoprim-sulfamethoxazole (3 days), and fosfomycin (1 dose, but expensive and less efficacious).
In geographic areas where trimethoprim-sulfamethoxazole resistance exceeds 20%, an alternative agent should be selected. The FDA recently indicated that fluoroquinolones should be reserved for other serious bacterial infections; however, fluoroquinolones (3 days) and β-lactam agents (including amoxicillin-clavulanate, cefdinir, cefaclor, and cefpodoxime-proxetil, each 3-7 days) are considered acceptable alternative second-line therapies. β-Lactams are not preferred if other recommended agents are available because they are less effective in eradicating infection. During pregnancy, the safest antibiotics are amoxicillin-clavulanate, cephalosporins, and nitrofurantoin. Tetracyclines and fluoroquinolones are contraindicated, and trimethoprim-sulfamethoxazole can only be used safely during the second trimester. Extended-spectrum β-lactamase–producing strains of Enterobacteriaceae causing cystitis have increased in frequency, especially in patients with recent antimicrobial or health care–facility exposure. Because of the greater risk of resistant and polymicrobial infections, urine culture and susceptibility testing are indicated in all patients with complicated cystitis. Fluoroquinolones are the preferred choice pending results, although fosfomycin and nitrofurantoin are reasonable options. The recommended treatment duration is 7 to 10 days rather than a 3-day, short-course regimen but is much less well defined. Other than in pregnant women, test of cure is not indicated in those reporting resolution of symptoms.
Lower urinary tract symptoms (frequency, urgency, and dysuria) often precede the onset of fever, chills, flank pain, and at times nausea and vomiting, which characterize acute pyelonephritis. Infection can usually be managed in the outpatient setting with oral antibiotics. Hospitalization is advised for patients with hemodynamic instability, obstructive disease, pregnancy, complicating comorbidities, known pathogen resistance requiring parenteral antibiotic therapy, inability to tolerate oral medications, or lack of reliable home supervision and clinical follow-up.
Every patient requires a urine culture with susceptibility testing obtained before initiation of empiric therapy. Fluoroquinolones (ciprofloxacin for 7 days or levofloxacin for 5 days for uncomplicated infections, 10-14 days in complicated infections) are the only oral agents recommended for empiric outpatient treatment, but an initial dose of a long-acting parenteral antibiotic (such as ceftriaxone, 1 g, or a once-daily aminoglycoside) should replace fluoroquinolones when local resistance rates exceed 10%. When a fluoroquinolone is contraindicated, trimethoprim-sulfamethoxazole twice daily for 14 days may be used after the pathogen is proven to be susceptible; trimethoprim-sulfamethoxazole should be avoided as initial empiric therapy because of the high level of E. coli resistance to this antibiotic in the community.
Depending on the risk of antimicrobial resistance and on recent antibiotic use, inpatient parenteral antimicrobial options include a fluoroquinolone, extended-spectrum cephalosporins (ceftriaxone or cefepime) or penicillins (piperacillin-tazobactam), or a carbapenem (meropenem, imipenem, or ertapenem). Again, fluoroquinolones are avoided for empiric therapy in severely ill patients with complicated pyelonephritis because of the increasing potential for resistance in E. coli and other aerobic gram-negative bacilli.
Therapy can be completed with active oral agents when an adequate clinical response has been observed. Patients with bacteremia do not require longer courses of treatment and may be converted to appropriate oral therapy when clinically stable.
Imaging studies are only necessary in patients with prolonged fever (>72 hours) or persistent bacteremia, in which complications such as obstruction or perinephric and intrarenal abscesses must be excluded. Routine follow-up urine cultures are only indicated in pregnant women.
Recurrent UTI is defined as three episodes of UTI in the preceding 12 months or two episodes in the preceding 6 months. Recurrent UTI is common in women. A recurrent UTI may be a relapse or reinfection. Relapse is defined as an infection caused by the same strain (by repeat culture) as the initial UTI and occurs within 2 weeks of completing initial therapy. Relapse suggests infection with a resistant strain of bacteria, incomplete treatment of an infection of the upper urinary tract, or a structural abnormality, including renal calculi.
Reinfection is diagnosed if the UTI is caused by a different strain than that causing the initial infection or if a sterile urine culture was documented between episodes. Most recurrences are reinfections.
Reinfection, the most common type of recurrent UTI, is generally caused by a bacterial strain separate from the original infection and presents more than 2 weeks after cessation of treatment for the previous infection. Symptomatic relapsed infection requires a urine culture. Assuming the organism is sensitive, patients are treated for infection of the upper urinary tract for 7 to 10 days with the same antibiotic as prescribed for the previous infection or, if bacterial resistance is discovered, an alternative agent. Likewise, the same first-line antimicrobial agent can be given for reinfections, although an alternative antibiotic should be used if the recurrence occurs within 6 months, particularly if the original agent was trimethoprim-sulfamethoxazole, because of the increased chance of resistance.
Although fluoroquinolone antibiotics are no longer recommended as first-line agents for the treatment of cystitis because of increasing concerns for potential adverse effects and uropathogen antimicrobial resistance development, ciprofloxacin and levofloxacin are the preferred antimicrobial agents when trimethoprim-sulfamethoxazole local resistance rates are high (>20%) or the patient has been treated with an antibiotic for a UTI within the previous 3 months. Having recently received antibiotics defines this patient's UTI as complicated, warranting 7 to 10 days of treatment with a fluoroquinolone antibiotic
Nitrofurantoin, trimethoprim-sulfamethoxazole, fosfomycin, and oral β-lactams are not recommended as first-line empiric oral therapy in complicated cystitis because of concerns regarding resistance to these agents. In the case of culture-proven sensitivity, these agents can be used in the treatment of complicated UTI.
Strategies to prevent infection recurrence include avoidance of spermicide contraceptives, urination soon after intercourse, topical vaginal estrogens, ascorbic acid (vitamin C), and methenamine salts. Cranberry products have not been proven effective in controlled trials.
Prophylactic daily antimicrobial agents have been found to reduce the risk of recurrences by nearly 95%; they are an option in women who have had three or more UTIs in the previous 12 months, or two or more in the previous 6 months, and have received no benefit from other prevention efforts. Prophylactic therapy should be considered in pregnant patients who have required treatment for cystitis or asymptomatic bacteriuria to prevent recurrence during pregnancy. Antibiotic complications and potential emergence of resistance must be considered. Approximately 50% of patients revert to previous recurrence patterns within 6 months of prophylaxis discontinuation. Recommended prophylactic antibiotics include nitrofurantoin, trimethoprim-sulfamethoxazole, trimethoprim, cephalexin, or fosfomycin. Fluoroquinolones are very effective but not recommended. Other options include postcoital antimicrobial prophylaxis and self-diagnosis with self-treatment.
Benign prostatic hyperplasia resulting in urinary obstruction and altered urine flow is the most common reason for the increased incidence of UTIs in men older than 60 years. Other risk factors include unprotected sexual intercourse, chronic indwelling urinary catheters, and transrectal prostate biopsy. Approximately 5% of men develop chronic prostatitis after acute infection.
Presenting symptoms include sudden onset of fever, pelvic or perineal pain, urinary frequency and dysuria, and increasing obstructive symptoms. Acute bacterial prostatitis frequently presents as a severe systemic infection and is the most common cause of bacteremia in older men. Cautious digital rectal examination of the prostate reveals a boggy and tender gland. Urinalysis and culture are required to confirm the diagnosis. Although pyuria may be present for reasons other than infection, its absence strongly indicates no infection. Prostate-specific antigen test results may be elevated because of inflammation of the gland and should be avoided in the setting of presumed infection.
Hospitalized patients and those with severe infection require blood cultures. Gram-negative uropathogens account for about 80% of infections, two thirds of which are E. coli; Proteus, Enterobacter, Serratia, Klebsiella, and sometimes Pseudomonas and enterococcal species compose most of the other pathogens. In men 35 years or younger, sexually transmitted infections, including Neisseria gonorrhoeae and Chlamydia trachomatis, must be considered.
Fluoroquinolone antibiotics (ciprofloxacin, levofloxacin) may be the preferred oral agents for treating acute bacterial prostatitis but should not be used if recent genitourinary instrumentation was performed, especially transrectal prostate biopsies, because most E. coli strains are now resistant to fluoroquinolones. Trimethoprim-sulfamethoxazole also has good tissue penetration and is a viable treatment option. Treatment duration is typically 4 to 6 weeks. Hospitalized patients should initially receive a broad-spectrum parenteral antibiotic, such as an extended-spectrum penicillin or cephalosporin, with the possible addition of an aminoglycoside. Imaging studies are not recommended unless a prostatic abscess is suspected clinically.