Urinary Tract

Ultrasonography

Ultrasound has found increased application to the lower genitourinary tract. It results in minimal discomfort; gives a 3-dimensional appreciation of the shape, size, and volume of organs and disease; and can provide direct intervention. Various transducers are available; high-megahertz transducers are required for superficial structures (for example, scrotal structures) to assess testicular disease (including tumors and torsion), while low-megahertz transducers are reserved for deep structures (for example, guiding percutaneous access for kidneys and bladders). Intervening tissue can significantly reduce image quality.

Transrectal ultrasound is valuable in evaluating the prostate to determine size and confirm digital information on the presence and stage of a suspected malignant tumor. Because of the low incidence of detecting malignancies (1.6–7%), mass screening programs are not cost-effective. Direct needle biopsies, with automatic biopsy mechanisms, are quick, well-tolerated, and result in reliable tissue cores and less pain than traditional needles (such as Tru-Cut) directed under digital palpation. Percutaneous drainage tubes, radioactive seed implants, and temperature coils used for cryosurgery of the prostate can be placed safely under transrectal ultrasonic guidance. Transrectal ultrasonography can yield unreliable images that often are misinterpreted by the novice. Pitfalls include faulty instrument settings, poor coupling caused by feces or gas, and unrecognized artifacts resulting from reverberation, deflection, shadowing, or enhancement.

Suprapubic ultrasonography can help to assess prostate anatomy, especially size and intravesical extension. It can help to evaluate the bladder for residual urine and for calculi that are questionable on plain abdominal radiographs. (Changing the patient's position can shift the position of a calculus.) Distal ureteral stones can be identified, especially when visualized through a full bladder used as an acoustic window. Double-J stents, incrustations, diverticula, and large malignant lesions can be identified. The procedure also can direct placement of suprapubic cystostomy drainage catheters.

Additional applications include endocavitary, color, Doppler, and dynamic ultrasonography. Endocavitary ultrasound, which includes transvaginal, transurethral (Figure 10–8), and transcystoscopic techniques, can delineate vaginal, urethral, and bladder disease. Endoureteral ultrasound can help in the identification of crossing vessels, preferably before an endopyelotomy. Color and Doppler ultrasound can assess blood flow as related to erectile dysfunction. Dynamic ultrasound can supplement urodynamic findings. Ultrasound applied to the lower genitourinary tract causes minimal discomfort and provides valuable information.

Figure 10–8. Multifocal bladder cancer. Left: Transurethral ultrasound. Right: Cystectomy specimen.

Acute UTIs: Urethritis, Cystitis, and Pyelonephritis

Epidemiology Epidemiologically, UTIs are subdivided into catheter-associated (or nosocomial) infections and non-catheter-associated (or community-acquired) infections. Infections in either category may be symptomatic or asymptomatic. Acute community-acquired infections are very common and account for more than 7 million office visits annually in the United States. These infections occur in 1 to 3% of schoolgirls and then increase markedly in incidence with the onset of sexual activity in adolescence. The vast majority of acute symptomatic infections involve young women; a prospective study demonstrated an annual incidence of 0.5 to 0.7 infections per patient-year in this group. Acute symptomatic UTIs are unusual in men under the age of 50. The development of asymptomatic bacteriuria parallels that of symptomatic infection and is rare among men under 50 but common among women between 20 and 50. Asymptomatic bacteriuria is more common among elderly men and women, with rates as high as 40 to 50% in some studies. Etiology Many different microorganisms can infect the urinary tract, but by far the most common agents are the gram-negative bacilli. Escherichia coli causes 80% of acute infections in patients without catheters, urologic abnormalities, or calculi. Other gram-negative rods, especially Proteus and Klebsiella and occasionally Enterobacter, account for a smaller proportion of uncomplicated infections. These organisms, plus Serratia and Pseudomonas, assume increasing importance in recurrent infections and in infections associated with urologic manipulation, calculi, or obstruction. They play a major role in nosocomial, catheter-associated infections (see below). Proteus spp., by virtue of urease production, and Klebsiella spp., through the production of extracellular slime and polysaccharides, predispose to stone formation and are isolated more frequently from patients with calculi. Gram-positive cocci play a lesser role in UTIs. However, Staphylococcus saprophyticus—a novobiocin-resistant, coagulase-negative species—accounts for 10 to 15% of acute symptomatic UTIs in young females. Enterococci occasionally cause acute uncomplicated cystitis in women. More commonly, enterococci and Staphylococcus aureus cause infections in patients with renal stones or previous instrumentation or surgery. Isolation of S. aureus from the urine should arouse suspicion of bacteremic infection of the kidney. About one-third of women with dysuria and frequency have either an insignificant number of bacteria in midstream urine cultures or completely sterile cultures and have been previously defined as having the urethral syndrome. About three-quarters of these women have pyuria, while one-quarter have no pyuria and little objective evidence of infection. In the women with pyuria, two groups of pathogens account for most infections. Low counts (102 to 104/mL) of typical bacterial uropathogens such as E. coli, S. saprophyticus, Klebsiella, or Proteus are found in midstream urine specimens from most of these women. These bacteria are probably the causative agents in these infections because they can usually be isolated from a suprapubic aspirate, are associated with pyuria, and respond to appropriate antimicrobial therapy. In other women with acute urinary symptoms, pyuria, and urine that is sterile (even when obtained by suprapubic aspiration), sexually transmitted urethritis-producing agents such as Chlamydia trachomatis, Neisseria gonorrhoeae, and herpes simplex virus are etiologically important. These agents are found most frequently in young, sexually active women with new sexual partners. The causative role of several more unusual bacterial and nonbacterial pathogens in UTIs remains poorly defined. Ureaplasma urealyticum has frequently been isolated from the urethra and urine of patients with acute dysuria and frequency but is also found in specimens from many patients without urinary symptoms. Ureaplasmas probably account for some cases of urethritis and cystitis. U. urealyticum and Mycoplasma hominis have been isolated from prostatic and renal tissues of patients with acute prostatitis and pyelonephritis, respectively, and are probably responsible for some of these infections as well. Adenoviruses cause acute hemorrhagic cystitis in children and in some young adults, often in epidemics. Although other viruses can be isolated from urine (e.g., cytomegalovirus), they are thought not to cause acute UTI. Colonization of the urine of catheterized or diabetic patients by Candida and other fungal species is common and sometimes progresses to symptomatic invasive infection (Chap. 187). Mycobacterial infection of the genitourinary tract is discussed in Chap. 150. Pathogenesis and Sources of Infection The urinary tract should be viewed as a single anatomic unit that is united by a continuous column of urine extending from the urethra to the kidney. In the vast majority of UTIs, bacteria gain access to the bladder via the urethra. Ascent of bacteria from the bladder may follow and is probably the pathway for most renal parenchymal infections. The vaginal introitus and distal urethra are normally colonized by diphtheroids, streptococcal species, lactobacilli, and staphylococcal species but not by the enteric gram-negative bacilli that commonly cause UTIs. In females prone to the development of cystitis, however, enteric gram-negative organisms residing in the bowel colonize the introitus, the periurethral skin, and the distal urethra before and during episodes of bacteriuria. The factors that predispose to periurethral colonization with gram-negative bacilli remain poorly understood, but alteration of the normal vaginal flora by antibiotics, other genital infections, or contraceptives (especially spermicide) appears to play an important role. Loss of the normally dominant H2O2-producing lactobacilli in the vaginal flora appears to facilitate colonization by E. coli. Small numbers of periurethral bacteria probably gain entry to the bladder frequently, a process that is facilitated in some cases by urethral massage during intercourse. Whether bladder infection ensues depends on interacting effects of the pathogenicity of the strain, the inoculum size, and the local and systemic host defense mechanisms. Under normal circumstances, bacteria placed in the bladder are rapidly cleared, partly through the flushing and dilutional effects of voiding but also as a result of the antibacterial properties of urine and the bladder mucosa. Owing mostly to a high urea concentration and high osmolarity, the bladder urine of many normal persons inhibits or kills bacteria. Prostatic secretions possess antibacterial properties as well. Polymorphonuclear leukocytes enter the bladder epithelium and the urine soon after infection arises and play a role in clearing bacteriuria. The role of locally produced antibody remains unclear. Hematogenous pyelonephritis occurs most often in debilitated patients who are either chronically ill or receiving immunosuppressive therapy. Metastatic staphylococcal or candidal infections of the kidney may follow bacteremia or fungemia, spreading from distant foci of infection in the bone, skin, vasculature, or elsewhere. Conditions Affecting Pathogenesis Gender and Sexual Activity The female urethra appears to be particularly prone to colonization with colonic gram-negative bacilli because of its proximity to the anus, its short length (4 cm), and its termination beneath the labia. Sexual intercourse causes the introduction of bacteria into the bladder and is temporally associated with the onset of cystitis; it thus appears to be important in the pathogenesis of UTIs in younger women. Voiding after intercourse reduces the risk of cystitis, probably because it promotes the clearance of bacteria introduced during intercourse. Use of spermicidal compounds with a diaphragm or cervical cap or use of spermicide-coated condoms dramatically alters the normal introital bacterial flora and has been associated with marked increases in vaginal colonization with E. coli and in the risk of UTI. In males who are <50 years old and who have no history of heterosexual or homosexual rectal intercourse, UTI is exceedingly uncommon, and this diagnosis should be questioned in the absence of clear documentation. An important factor predisposing to bacteriuria in men is urethral obstruction due to prostatic hypertrophy. Insertive rectal intercourse is also associated with an increased risk of cystitis in men. Men (and women) who are infected with HIV and who have CD4+ T cell counts of <200/L are at increased risk of both bacteriuria and symptomatic UTI. Finally, lack of circumcision has been identified as a risk factor for UTI in both neonates and young men. Pregnancy UTIs are detected in 2 to 8% of pregnant women. Symptomatic upper tract infections, in particular, are unusually common during pregnancy; fully 20 to 30% of pregnant women with asymptomatic bacteriuria subsequently develop pyelonephritis. This predisposition to upper tract infection during pregnancy results from decreased ureteral tone, decreased ureteral peristalsis, and temporary incompetence of the vesicoureteral valves. Bladder catheterization during or after delivery causes additional infections. Increased incidences of low-birth-weight infants, premature delivery, and newborn mortality result from UTIs during pregnancy, particularly those infections involving the upper tract. Obstruction Any impediment to the free flow of urine—tumor, stricture, stone, or prostatic hypertrophy—results in hydronephrosis and a greatly increased frequency of UTI. Infection superimposed on urinary tract obstruction may lead to rapid destruction of renal tissue. It is of utmost importance, therefore, when infection is present, to identify and repair obstructive lesions. On the other hand, when an obstruction is minor and is not progressive or associated with infection, great caution should be exercised in attempting surgical correction. The introduction of infection in such cases may be more damaging than an uncorrected minor obstruction that does not significantly impair renal function. Neurogenic Bladder Dysfunction Interference with bladder enervation, as in spinal cord injury, tabes dorsalis, multiple sclerosis, diabetes, and other diseases, may be associated with UTI. The infection may be initiated by the use of catheters for bladder drainage and is favored by the prolonged stasis of urine in the bladder. An additional factor often operative in these cases is bone demineralization due to immobilization, which causes hypercalciuria, calculus formation, and obstructive uropathy. Vesicoureteral Reflux Defined as reflux of urine from the bladder cavity up into the ureters and sometimes into the renal pelvis, vesicoureteral reflux occurs during voiding or with elevation of pressure in the bladder. In practice, this condition is demonstrated by the finding of retrograde movement of radiopaque or radioactive material during a voiding cystourethrogram. An anatomically impaired vesicoureteral junction facilitates reflux of bacteria and thus upper tract infection. However, since a fluid connection between the bladder and the kidney always exists, even in the normal urinary system, some retrograde movement of bacteria probably takes place during infection but is not detected by radiologic techniques. Vesicoureteral reflux is common among children with anatomic abnormalities of the urinary tract as well as among children with anatomically normal but infected urinary tracts. In the latter group, reflux disappears with advancing age and is probably attributable to factors other than UTI. Long-term follow-up of children with UTI who have reflux has established that renal damage correlates with marked reflux, not with infection. The routine search for reflux would be aided by the development of noninvasive tests applicable to young children, in whom the need for an effective technique is greatest. In the meantime, it appears reasonable to search for reflux in anyone with unexplained failure of renal growth or with renal scarring, because UTI per se is an insufficient explanation for these abnormalities. On the other hand, it is doubtful that all children who have recurrent UTIs but whose urinary tract appears normal on pyelography should be subjected to voiding cystoureterography merely for the detection of the rare patient with marked reflux not revealed by the intravenous pyelogram. Bacterial Virulence Factors Not all strains of E. coli are equally capable of infecting the intact urinary tract. Bacterial virulence factors markedly influence the likelihood that a given strain, once introduced into the bladder, will cause UTI. Most E. coli strains that cause symptomatic UTIs in noncatheterized patients belong to a small number of specific O, K, and H serogroups. These uropathogenic clones have accumulated a number of virulence genes that are often closely linked on the bacterial chromosome in "virulence islands." Adherence of bacteria to uroepithelial cells is a critical first step in the initiation of infection. For both E. coli and Proteus, fimbriae (hairlike proteinaceous surface appendages) mediate the attachment of bacteria to specific receptors on epithelial cells. The attachment of bacteria to uroepithelial cells initiates a number of important events in the mucosal epithelial cell, including secretion of interleukin (IL) 6 and IL-8 (with subsequent chemotaxis of leukocytes to the bladder mucosa) and induction of apoptosis and epithelial cell desquamation. Besides fimbriae, uropathogenic E. coli strains usually produce hemolysin and aerobactin (a siderophore for scavenging iron) and are resistant to the bactericidal action of human serum. Nearly all E. coli strains causing acute pyelonephritis and most of those causing acute cystitis are uropathogenic. In contrast, infections in patients with structural or functional abnormalities of the urinary tract are generally caused by bacterial strains that lack these uropathogenic properties; the implication is that these properties are not needed for infection of the compromised urinary tract. Genetic Factors Increasing evidence suggests that host genetic factors influence susceptibility to UTI. A maternal history of UTI is more often found among women who have experienced recurrent UTIs than among controls. The number and type of receptors on uroepithelial cells to which bacteria may attach are at least in part genetically determined. Many of these structures are components of blood group antigens and are present on both erythrocytes and uroepithelial cells. For example, P fimbriae mediate attachment of E. coli to P-positive erythrocytes and are found on nearly all strains causing acute uncomplicated pyelonephritis. Conversely, P blood group–negative individuals, who lack these receptors, have a decreased likelihood of pyelonephritis. It has also been demonstrated that nonsecretors of blood group antigens are at increased risk of recurrent UTI; this predisposition may relate to a different profile of genetically determined glycolipids on uroepithelial cells. Mutations in host genes integral to the immune response (interferon receptors and others) may also affect susceptibility to UTI. Localization of Infection Unfortunately, currently available methods of distinguishing renal parenchymal infection from cystitis are neither reliable nor convenient enough for routine clinical use. Fever or an elevated level of C-reactive protein often accompanies acute pyelonephritis and is found in rare cases of cystitis but also occurs in infections other than pyelonephritis. Clinical Presentation Cystitis Patients with cystitis usually report dysuria, frequency, urgency, and suprapubic pain. The urine often becomes grossly cloudy and malodorous, and it is bloody in 30% of cases. White cells and bacteria can be detected by examination of unspun urine in most cases. However, some women with cystitis have only 102 to 104 bacteria per milliliter of urine, and in these instances bacteria cannot be seen in a Gram-stained preparation of unspun urine. Physical examination generally reveals only tenderness of the urethra or the suprapubic area. If a genital lesion or a vaginal discharge is evident, especially in conjunction with <105 bacteria per milliliter on urine culture, then pathogens that may cause urethritis, vaginitis, or cervicitis, such as C. trachomatis, N. gonorrhoeae, Trichomonas, Candida, and herpes simplex virus, should be considered. Prominent systemic manifestations, such as a temperature of >38.3°C (>101°F), nausea, and vomiting, usually indicate concomitant renal infection, as does costovertebral angle tenderness. However, the absence of these findings does not ensure that infection is limited to the bladder and urethra. Acute Pyelonephritis Symptoms of acute pyelonephritis generally develop rapidly over a few hours or a day and include a fever, shaking chills, nausea, vomiting, and diarrhea. Symptoms of cystitis may or may not be present. Besides fever, tachycardia, and generalized muscle tenderness, physical examination reveals marked tenderness on deep pressure in one or both costovertebral angles or on deep abdominal palpation. In some patients, signs and symptoms of gram-negative sepsis predominate. Most patients have significant leukocytosis and bacteria detectable in Gram-stained unspun urine. Leukocyte casts are present in the urine of some patients, and the detection of these casts is pathognomonic. Hematuria may be demonstrated during the acute phase of the disease; if it persists after acute manifestations of infection have subsided, a stone, a tumor, or tuberculosis should be considered. Except in individuals with papillary necrosis, abscess formation, or urinary obstruction, the manifestations of acute pyelonephritis usually respond to therapy within 48 to 72 h. However, despite the absence of symptoms, bacteriuria or pyuria may persist. In severe pyelonephritis, fever subsides more slowly and may not disappear for several days, even after appropriate antibiotic treatment has been instituted. Urethritis Approximately 30% of women with acute dysuria, frequency, and pyuria have midstream urine cultures that show either no growth or insignificant bacterial growth. Clinically, these women cannot always be readily distinguished from those with cystitis. In this situation, a distinction should be made between women infected with sexually transmitted pathogens, such as C. trachomatis, N. gonorrhoeae, or herpes simplex virus, and those with low-count E. coli or staphylococcal infection of the urethra and bladder. Chlamydial or gonococcal infection should be suspected in women with a gradual onset of illness, no hematuria, no suprapubic pain, and >7 days of symptoms. The additional history of a recent sex-partner change, especially if the patient's partner has recently had chlamydial or gonococcal urethritis, should heighten the suspicion of a sexually transmitted infection, as should the finding of mucopurulent cervicitis (Chap. 115). Gross hematuria, suprapubic pain, an abrupt onset of illness, a duration of illness of <3 days, and a history of UTIs favor the diagnosis of E. coli UTI. Catheter-Associated UTIs See also Chap. 116. Bacteriuria develops in at least 10 to 15% of hospitalized patients with indwelling urethral catheters. The risk of infection is 3 to 5% per day of catheterization. E. coli, Proteus, Pseudomonas, Klebsiella, Serratia, staphylococci, enterococci, and Candida usually cause these infections. Many infecting strains display markedly greater antimicrobial resistance than organisms that cause community-acquired UTIs. Factors associated with an increased risk of catheter-associated UTI include female sex, prolonged catheterization, severe underlying illness, disconnection of the catheter and drainage tube, other types of faulty catheter care, and lack of systemic antimicrobial therapy. Infection occurs when bacteria reach the bladder by one of two routes: by migrating through the column of urine in the catheter lumen (intraluminal route) or by moving up the mucous sheath outside the catheter (periurethral route). Hospital-acquired pathogens reach the patient's catheter or urine-collecting system on the hands of hospital personnel, in contaminated solutions or irrigants, and via contaminated instruments or disinfectants. Bacteria usually enter the catheter system at the catheter–collecting tube junction or at the drainage bag portal. The organisms then ascend intraluminally into the bladder within 24 to 72 h. Alternatively, the patient's own bowel flora may colonize the perineal skin and periurethral area and reach the bladder via the external surface of the catheter. This route is particularly common in women. Studies have demonstrated the importance of the attachment and growth of bacteria on the surfaces of the catheter in the pathogenesis of catheter-associated UTI. Such bacteria growing in biofilms on the catheter eventually produce encrustations consisting of bacteria, bacterial glycocalyces, host urinary proteins, and urinary salts. These encrustations provide a refuge for bacteria and may protect them from antimicrobial agents and phagocytes. Clinically, most catheter-associated infections cause minimal symptoms and no fever and often resolve after withdrawal of the catheter. The frequency of upper tract infection associated with catheter-induced bacteriuria is unknown. Gram-negative bacteremia, which follows catheter-associated bacteriuria in 1 to 2% of cases, is the most significant recognized complication of catheter-induced UTIs. The catheterized urinary tract has repeatedly been demonstrated to be the most common source of gram-negative bacteremia in hospitalized patients, generally accounting for 30% of cases. Catheter-associated UTIs can sometimes be prevented in patients catheterized for <2 weeks by use of a sterile closed collecting system, by attention to aseptic technique during insertion and care of the catheter, and by measures to minimize cross-infection. Other preventive approaches, including short courses of systemic antimicrobial therapy, topical application of periurethral antimicrobial ointments, use of preconnected catheter–drainage tube units, use of catheters impregnated with antimicrobial agents, and addition of antimicrobial drugs to the drainage bag, have all been protective in at least one controlled trial but are not recommended for general use. Despite precautions, the majority of patients catheterized for >2 weeks eventually develop bacteriuria. For example, because of spinal cord injury, incontinence, or other factors, some patients in hospitals or nursing homes require long-term or semipermanent bladder catheterization. Measures intended to prevent infection have been largely unsuccessful, and essentially all such chronically catheterized patients develop bacteriuria. If feasible, intermittent catheterization by a nurse or by the patient appears to reduce the incidence of bacteriuria and associated complications in such patients. Treatment should be provided when symptomatic infections arise, but treatment of asymptomatic bacteriuria in such patients has no apparent benefit. Diagnostic Testing Determination of the number and type of bacteria in the urine is an extremely important diagnostic procedure. In symptomatic patients, bacteria are usually present in the urine in large numbers (105/mL). In asymptomatic patients, two consecutive urine specimens should be examined bacteriologically before therapy is instituted, and 105 bacteria of a single species per milliliter should be demonstrable in both specimens. Since the large number of bacteria in the bladder urine is due in part to bacterial multiplication during residence in the bladder cavity, samples of urine from the ureters or renal pelvis may contain <105 bacteria per milliliter and yet indicate infection. Similarly, the presence of bacteriuria of any degree in suprapubic aspirates or of 102 bacteria per milliliter of urine obtained by catheterization usually indicates infection. In some circumstances (antibiotic treatment, high urea concentration, high osmolarity, low pH), urine inhibits bacterial multiplication, resulting in relatively low bacterial colony counts despite infection. For this reason, antiseptic solutions should not be used in washing the periurethral area before collection of the urine specimen. Water diuresis or recent voiding also reduces bacterial counts in urine. Microscopy of urine from symptomatic patients can be of great diagnostic value. Microscopic bacteriuria, which is best assessed with Gram-stained uncentrifuged urine, is found in >90% of specimens from patients whose infections are associated with colony counts of at least 105/mL, and this finding is very specific. However, bacteria cannot usually be detected microscopically in infections with lower colony counts (102 to 104/mL). The detection of bacteria by urinary microscopy thus constitutes firm evidence of infection, but the absence of microscopically detectable bacteria does not exclude the diagnosis. When carefully sought by means of chamber-count microscopy, pyuria is a highly sensitive indicator of UTI in symptomatic patients. Pyuria is demonstrated in nearly all acute bacterial UTIs, and its absence calls the diagnosis into question. The leukocyte esterase "dipstick" method is less sensitive than microscopy in identifying pyuria but is a useful alternative where microscopy is not feasible. Pyuria in the absence of bacteriuria (sterile pyuria) may indicate infection with unusual bacterial agents such as C. trachomatis, U. urealyticum, and Mycobacterium tuberculosis or with fungi. Alternatively, sterile pyuria may be demonstrated in noninfectious urologic conditions such as calculi, anatomic abnormality, nephrocalcinosis, vesicoureteral reflux, interstitial nephritis, or polycystic disease. Although many authorities have recommended that urine culture and antimicrobial susceptibility testing be performed for any patient with a suspected UTI, it may be more practical and cost-effective to manage women who have symptoms characteristic of acute uncomplicated cystitis without an initial urine culture. Two approaches to presumptive therapy have generally been used. In the first, treatment is initiated solely on the basis of a typical history and/or typical findings on physical examination. In the second, women with symptoms and signs of acute cystitis and without complicating factors are managed with urinary microscopy (or, alternatively, with a leukocyte esterase test). A positive result for pyuria and/or bacteriuria provides enough evidence of infection to indicate that urine culture and susceptibility testing can be omitted and the patient treated empirically. Urine should be cultured, however, when a woman's symptoms and urine-examination findings leave the diagnosis of cystitis in question. Pretherapy cultures and susceptibility testing are also essential in the management of all patients with suspected upper tract infections and of those with complicating factors, as in these situations any of a variety of pathogens may be involved and antibiotic therapy is best tailored to the individual organism. Treatment The following principles underlie the treatment of UTIs:   1. Except in acute uncomplicated cystitis in women, a quantitative urine culture or a comparable alternative diagnostic test should be performed to confirm infection before empirical treatment is begun. When culture results become available, antimicrobial sensitivity testing should be used to further direct therapy.   2. Factors predisposing to infection, such as obstruction and calculi, should be identified and corrected if possible.   3. Relief of clinical symptoms does not always indicate bacteriologic cure.   4. Each course of treatment should be classified after its completion as a failure (symptoms and/or bacteriuria not eradicated during therapy or in the immediate posttreatment culture) or a cure (resolution of symptoms and elimination of bacteriuria). Recurrent infections should be classified as same-strain or different-strain and as early (occurring within 2 weeks of the end of therapy) or late.   5. In general, uncomplicated infections confined to the lower urinary tract respond to short courses of therapy, while upper tract infections require longer treatment. After therapy, early recurrences due to the same strain may result from an unresolved upper tract focus of infection but often (especially after short-course therapy for cystitis) result from persistent vaginal colonization. Recurrences >2 weeks after the cessation of therapy nearly always represent reinfection with a new strain or with the previously infecting strain that has persisted in the vaginal and rectal flora.   6. Despite increasing resistance, community-acquired infections, especially initial infections, are usually due to more antibiotic-sensitive strains.   7. In patients with repeated infections, instrumentation, or recent hospitalization, the presence of antibiotic-resistant strains should be suspected. Although many antimicrobial agents reach high concentrations in urine, in vitro resistance usually predicts a substantially higher failure rate. The anatomic location of a UTI greatly influences the success or failure of a therapeutic regimen. Bladder bacteriuria (cystitis) can usually be eliminated with nearly any antimicrobial agent to which the infecting strain is sensitive; in the past, it was demonstrated that as little as a single dose of 500 mg of intramuscular kanamycin eliminated bladder bacteriuria in most cases. With upper tract infections, however, single-dose therapy fails in the majority of cases, and even a 7-day course is unsuccessful in many instances. Longer periods of treatment (2 to 6 weeks) aimed at eradicating a persistent focus of infection may be necessary in some cases. In acute uncomplicated cystitis, more than 90 to 95% of infections are due to one of two organisms: E. coli or S. saprophyticus. Although resistance patterns vary geographically and resistance has increased in many areas, most strains are sensitive to many antibiotics. In most parts of the United States, more than one-quarter of E. coli strains causing acute cystitis are resistant to amoxicillin, sulfa drugs, and cephalexin; resistance to trimethoprim (TMP) and trimethoprim-sulfamethoxazole (TMP-SMX) is now approaching these levels as well in many areas. Substantially higher rates of resistance to TMP-SMX have been documented in some other countries, as has resistance to fluoroquinolones. Many have advocated single-dose treatment for acute cystitis. The advantages of single-dose therapy include less expense, ensured compliance, fewer side effects, and perhaps less intense pressure favoring the selection of resistant organisms in the intestinal, vaginal, or perineal flora. However, more frequent recurrences develop shortly after single-dose therapy than after 3-day treatment, and single-dose therapy does not eradicate vaginal colonization with E. coli as effectively as do longer regimens. A 3-day course of therapy with TMP-SMX, TMP , norfloxacin, ciprofloxacin, or ofloxacin appears to preserve the low rate of side effects of single-dose therapy while improving efficacy (Table 269-1); thus 3-day regimens are currently preferred for acute cystitis. In areas where TMP-SMX resistance exceeds 20%, either a fluoroquinolone or nitrofurantoin can be used (Table 269-1). Neither single-dose nor 3-day therapy should be used for women with symptoms or signs of pyelonephritis, urologic abnormalities or stones, or previous infections due to antibiotic-resistant organisms. Males with UTI often have urologic abnormalities or prostatic involvement and hence are not candidates for single-dose or 3-day therapy. For empirical therapy, they should generally receive a 7- to 14-day course of a fluoroquinolone (Table 269-1). The choice of treatment for women with acute urethritis depends on the etiologic agent involved. In chlamydial infection, azithromycin (1 g in a single oral dose) or doxycycline (100 mg twice daily by mouth for 7 days) should be used. Women with acute dysuria and frequency, negative urine cultures, and no pyuria usually do not respond to antimicrobial agents. In women, acute uncomplicated pyelonephritis without accompanying clinical evidence of calculi or urologic disease is due to E. coli in most cases. Although the optimal route and duration of therapy have not been established, a 7- to 14-day course of a fluoroquinolone, an aminoglycoside, or a third-generation cephalosporin is usually adequate. Neither ampicillin nor TMP-SMX should be used as initial therapy because >25% of strains of E. coli causing pyelonephritis are now resistant to these drugs in vitro. For at least the first few days of treatment, antibiotics should probably be given intravenously to most patients, but patients with mild symptoms can be treated for 7 to 14 days with an oral antibiotic (usually ciprofloxacin or ofloxacin), with or without an initial single parenteral dose (Table 269-1). Patients who fail to respond to treatment within 72 h or who relapse after therapy should be evaluated for unrecognized suppurative foci, calculi, or urologic disease. Complicated UTIs (those arising in a setting of catheterization, instrumentation, urologic anatomic or functional abnormalities, stones, obstruction, immunosuppression, renal disease, or diabetes) are typically due to hospital-acquired bacteria, including E. coli, Klebsiella, Proteus, Serratia, Pseudomonas, enterococci, and staphylococci. Many of the infecting strains are antibiotic-resistant. Empirical antibiotic therapy ideally provides broad-spectrum coverage against these pathogens. In patients with minimal or mild symptoms, oral therapy with a fluoroquinolone, such as ciprofloxacin or ofloxacin, can be administered until culture results and antibiotic sensitivities are known. In patients with more severe illness, including acute pyelonephritis or suspected urosepsis, hospitalization and parenteral therapy should be undertaken. Commonly used empirical regimens include imipenem alone, a penicillin or cephalosporin plus an aminoglycoside, and (when the involvement of enterococci is unlikely) ceftriaxone or ceftazidime. When information on the antimicrobial sensitivity pattern of the infecting strain becomes available, a more specific antimicrobial regimen can be selected. Therapy should generally be administered for 10 to 21 days, with the exact duration depending on the severity of the infection and the susceptibility of the infecting strain. Follow-up cultures 2 to 4 weeks after cessation of therapy should be performed to demonstrate cure. The need for treatment as well as the optimal type and duration of treatment for catheterized patients with asymptomatic bacteriuria have not been established. Removal of the catheter in conjunction with a short course of antibiotics to which the organism is susceptible probably constitutes the best course of action and nearly always eradicates bacteriuria. Treatment of asymptomatic catheter-associated bacteriuria may be of greatest benefit to elderly women, who most often develop symptoms if left untreated. If the catheter cannot be removed, antibiotic therapy usually proves to be unsuccessful and may in fact result in infection with a more resistant strain. In this situation, the bacteriuria should be ignored unless the patient develops symptoms or is at high risk of developing bacteremia. In these cases, use of systemic antibiotics or urinary bladder antiseptics may reduce the degree of bacteriuria and the likelihood of bacteremia. In pregnancy, acute cystitis can be managed with 7 days of treatment with amoxicillin, nitrofurantoin, or a cephalosporin. All pregnant women should be screened for asymptomatic bacteriuria during the first trimester and, if bacteriuric, should be treated with one of the regimens listed in Table 269-1. After treatment, a culture should be performed to ensure cure, and cultures should be repeated monthly thereafter until delivery. Acute pyelonephritis in pregnancy should be managed with hospitalization and parenteral antibiotic therapy, generally with a cephalosporin or an extended-spectrum penicillin. Continuous low-dose prophylaxis with nitrofurantoin should be given to women who have recurrent infections during pregnancy. Asymptomatic bacteriuria in noncatheterized patients is common, especially among elderly patients, but has not been linked to adverse outcomes in most circumstances other than pregnancy. Thus antimicrobial therapy is unnecessary and may in fact promote the emergence of resistant strains in most patients with asymptomatic bacteriuria. High-risk patients with neutropenia, renal transplants, obstruction, or other complicating conditions may require treatment when asymptomatic bacteriuria occurs. Seven days of therapy with an oral agent to which the organism is sensitive should be given initially. If bacteriuria persists, it can be monitored without further treatment in most patients. Longer-term therapy (4 to 6 weeks) may be necessary in high-risk patients with persistent asymptomatic bacteriuria. Urologic Evaluation Very few women with recurrent UTIs have correctable lesions discovered at cystoscopy or upon intravenous pyelography, and these procedures should not be undertaken routinely in such cases. Urologic evaluation should be performed in selected instances—namely, in women with relapsing infection, a history of childhood infections, stones or painless hematuria, or recurrent pyelonephritis. Most males with UTI should be considered to have complicated infection and thus should be evaluated urologically. Possible exceptions include young men who have cystitis associated with sexual activity, who are uncircumcised, or who have AIDS. Men or women presenting with acute infection and signs or symptoms suggestive of an obstruction or stones should undergo prompt urologic evaluation, generally by means of ultrasound. Prognosis In patients with uncomplicated cystitis or pyelonephritis, treatment ordinarily results in complete resolution of symptoms. Lower tract infections in women are of concern mainly because they cause discomfort, morbidity, loss of time from work, and substantial health care costs. Cystitis may also result in upper tract infection or in bacteremia (especially during instrumentation), but little evidence suggests that renal impairment follows. When repeated episodes of cystitis occur, they are more commonly reinfections rather than relapses. Acute uncomplicated pyelonephritis in adults rarely progresses to renal functional impairment and chronic renal disease. Repeated upper tract infections often represent relapse rather than reinfection, and a vigorous search for renal calculi or an underlying urologic abnormality should be undertaken. If neither is found, 6 weeks of chemotherapy may be useful in eradicating an unresolved focus of infection. Repeated symptomatic UTIs in children and in adults with obstructive uropathy, neurogenic bladder, structural renal disease, or diabetes progress to chronic renal disease with unusual frequency. Asymptomatic bacteriuria in these groups as well as in adults without urologic disease or obstruction predisposes to increased numbers of episodes of symptomatic infection but does not result in renal impairment in most instances. Prevention Women who experience frequent symptomatic UTIs (3 per year on average) are candidates for long-term administration of low-dose antibiotics directed at preventing recurrences. Such women should be advised to avoid spermicide use and to void soon after intercourse. Daily or thrice-weekly administration of a single dose of TMP-SMX (80/400 mg), TMP alone (100 mg), or nitrofurantoin (50 mg) has been particularly effective. Norfloxacin and other fluoroquinolones have also been used for prophylaxis. Prophylaxis should be initiated only after bacteriuria has been eradicated with a full-dose treatment regimen. The same prophylactic regimens can be used after sexual intercourse to prevent episodes of symptomatic infection in women in whom UTIs are temporally related to intercourse. Other patients for whom prophylaxis appears to have some merit include men with chronic prostatitis; patients undergoing prostatectomy, both during the operation and in the postoperative period; and pregnant women with asymptomatic bacteriuria. All pregnant women should be screened for bacteriuria in the first trimester and should be treated if bacteriuria is demonstrated.