Urinary tract infections (UTIs) are a leading cause of morbidity and health care expenditures in persons of all ages. Sexually active young women are disproportionately affected, but several other populations, including elderly persons and those undergoing genitourinary instrumentation or catheterization, are also at risk. An estimated 40 percent of women report having had a UTI at some point in their lives.1 UTIs are the leading cause of gram-negative bacteremia. In the United States, these infections account for approximately 7 million office visits and more than 1 million hospitalizations, for an overall annual cost in excess of $1 billion.1,2
Recently published studies have added to the body of knowledge concerning the pathogenesis, diagnosis and management of UTIs. However, many practical issues have yet to be fully addressed. When should urine cultures be obtained? What diagnostic threshold should be used to define infection? What is the optimal duration of therapy and how should it be administered? Does bacteriuria in the elderly lead to adverse outcomes? Should trimethoprim-sulfamethoxazole (Bactrim, Septra) remain the initial therapy of choice for UTIs? This article clarifies these issues by reviewing the approach to the diagnosis and treatment of each patient group at risk for UTIs. In addition, a simple diagnostic approach to urinary tract infection in adults is presented in Figure 1.
A recent categorization of UTIs is most helpful clinically because it divides patients into groups based on clinical factors and their impact on morbidity and treatment (Table 1).3 These categories are as follows: acute uncomplicated cystitis in young women; recurrent cystitis in young women; acute uncomplicated pyelonephritis in young women; complicated UTI and its subcategories; UTI related to indwelling catheters; UTI in men; and asymptomatic bacteriuria.
|Category||Diagnostic criteria||Principal pathogens||First-line therapy||Comments|
|Acute uncomplicated cystitis||Urinalysis for pyuria and hematuria (culture not required)||Escherichia coli|
|TMP-SMX DS (Bactrim, Septra)|
|Three-day course is best Quinolones may be used in areas of TMP-SMX resistance or in patients who cannot tolerate TMP-SMX|
|Recurrent cystitis in young women||Symptoms and a urine culture with a bacterial count of more than100 CFU per mL of urine||Same as for acute uncomplicated cystitis||If the patient has more than three cystitis episodes per year, treat prophylactically with postcoital, patient-directed* or continuous daily therapy (see text)||Repeat therapy for seven to10 days based on culture results and then use prophylactic therapy|
|Acute cystitis in young men||Urine culture with a bacterial count of 1,000 to 10,000 CFU per mL of urine||Same as for acute uncomplicated cystitis||Same as for acute uncomplicated cystitis||Treat for seven to 10 days|
|Acute uncomplicated pyelonephritis||Urine culture with a bacterial count of100,000 CFU per mL of urine||Same as for acute uncomplicated cystitis||If gram-negative organism, oral fluoroquinolone||Switch from IV to oral administration when the patient is able to take medication by mouth; complete a 14-day course|
|If gram-positive organism, amoxicillin|
|If parenteral administration is required, ceftriaxone (Rocephin) or a fluoroquinolone|
|If Enterococcus species, add oral or IV amoxicillin|
|Complicated urinary tract infection||Urine culture with a bacterial count of more than 10,000 CFU per mL of urine||E. coli|
|If gram-negative organism, oral fluoroquinolone|
If Enterococcus species, ampicillin or amoxicillin with or without gentamicin(Garamycin)
|Treat for 10 to 14 days|
|Asymptomatic bacteriuria in pregnancy||Urine culture with a bacterial count of more than 10,000 CFU per mL of urine||Same as for acute uncomplicated cystitis||Amoxicillin|
|Avoid tetracyclines and fluoroquinolones|
Treat for three to seven days
|Catheter-associated urinary tract||Symptoms and a urine culture with a bacterial count of more than100 CFU per mL of urine||Depends on duration of catheterization||If gram-negative organism, a fluoroquinolone||Remove catheter if possible, and treat for seven to 10 days|
|If gram-positive organism, ampicillin or amoxicillin plus gentamicin||For patients with long-term catheters and symptoms, treat for five to seven days|
Acute Uncomplicated Cystitis in Young Women
Those most at risk for UTIs are sexually active young women. Their propensity to develop UTIs has been explained on the basis of anatomy (especially a short urethra) and certain behavioral factors, including delays in micturition, sexual activity, and the use of diaphragms and spermicides (both of which promote colonization of the periurethral area with coliform bacteria).4 Fortunately, most UTIs in this population are uncomplicated and are rarely associated with functional or anatomic abnormalities. In studies of women presenting with dysuria and increased frequency of urination, intravenous pyelography and ultrasonography have demonstrated low rates (less than 1 percent) of surgically correctable anatomic abnormalities of the urinary tract.5 Therefore, aggressive diagnostic work-ups are unwarranted in young women presenting with an uncomplicated episode of cystitis.3,6
The diagnosis of UTI was once based on a quantitative urine culture yielding greater than 100,000 colony-forming units (CFU) of bacteria per milliliter of urine, which was termed “significant bacteriuria.”7 This value was chosen because of its high specificity for the diagnosis of true infection, even in asymptomatic persons. However, several studies8–10 have established that one third or more of symptomatic women have CFU counts below this level (low-coliform-count infections) and that a bacterial count of 100 CFU per mL of urine has a high positive predictive value for cystitis in symptomatic women. Unfortunately, some clinical laboratories do not report counts of less than 10,000 CFU per mL of urine. As a result, low-coliform-count infections are not diagnosed by these laboratories.
The microbiology of uncomplicated cystitis is limited to a few pathogens. As many as 90 percent of uncomplicated cystitis episodes are caused by Escherichia coli, 10 to 20 percent are caused by coagulase-negative Staphylococcus saprophyticus and 5 percent or less are caused by other Enterobacteriaceae organisms or enterococci.3 In addition, the antimicrobial susceptibilities of these organisms are highly predictable. Up to one third of uropathogens are resistant to ampicillin and sulfonamides, but the majority are susceptible to trimethoprim-sulfamethoxazole (85 to 95 percent) and fluoroquinolones (95 percent).3,11
In view of the limited spectrum of causative organisms and their predictable susceptibility, urine cultures and susceptibility testing add little to the choice of antibiotic for the treatment of acute uncomplicated cystitis in young women. Therefore, urine cultures are no longer advocated as part of the routine work-up of these patients. Instead, these patients should undergo an abbreviated laboratory work-up in which the presence of pyuria is confirmed by traditional urinalysis (wet mount examination of spun urine), the cell-counting chamber technique or a dipstick test for leukocyte esterase.3,6
A positive leukocyte esterase test has a reported sensitivity of 75 to 90 percent in detecting pyuria associated with a UTI. Gram staining of unspun urine can be used to detect bacteriuria. In this semiquantitative test, one organism per oil immersion field correlates with 100,000 CFU per mL by culture.1 Because the procedure is time-consuming and has low sensitivity, it is not routinely performed in most clinical laboratories unless it is specifically requested. In today's office practice, the dipstick test for nitrite is used as a surrogate marker for bacteriuria. It should be noted that not all uropathogens reduce nitrates to nitrite. For example, enterococci, S. saprophyticus and Acinetobacter species do not and therefore give false-negative results. The sensitivities and specificities of the tests commonly used to diagnose UTIs are given in Table 2.12
Treatment options for uncomplicated cystitis include single-dose antibiotic therapy and three- or seven-day courses of antibiotics (Table 3). Treatment of cystitis with seven or more days of antibiotics once was the standard of therapy. Although this regimen was highly efficacious, it was associated with a certain (albeit low) frequency of side effects. Single-dose therapy appears to offer the advantages of low cost, high compliance and comparable efficacy. Studies using 3 g of amoxicillin, 400 mg of trimethoprim (Proloprim), two to three double-strength trimethoprim-sulfamethoxazole tablets, 800 mg of norfloxacin (Noroxin), 125 mg of ciprofloxacin (Cipro) or 200 mg of ofloxacin (Floxin) have confirmed that single-dose therapy is highly effective in the treatment of acute uncomplicated cystitis, with cure rates ranging from 80 to 99 percent.3
|Diagnostic group||Route of administration||Duration of therapy||Empiric options||Cost (generic)*|
|Acute uncomplicated urinary tract infections in women||Oral||Three days||Trimethoprim-sulfamethoxazole (Bactrim DS), one double-strength tablet twice daily||$ 7.50 (2.00)|
|Trimethoprim (Proloprim), 100 mg twice daily||4.00 (1.00)|
|Norfloxacin (Noroxin), 400 mg twice daily||18.50|
|Ciprofloxacin (Cipro), 250 mg twice daily||19.50|
|Lomefloxacin (Maxaquin), 400 mg per day||20.00|
|Ofloxacin (Floxin), 200 mg twice daily||21.00|
|Enoxacin (Penetrex), 200 mg twice daily||18.00|
|Sparfloxacin (Zagam), 400 mg as initial dose, then 200 mg per day||25.00|
|Levofloxacin (Levaquin), 250 mg per day||19.00|
|Nitrofurantoin (Macrodantin), 100 mg four times daily||17.00 (14.00)|
|Cefpodoxime (Vantin), 100 mg twice daily||17.00|
|Cefixime (Suprax), 400 mg per day||21.00|
|Amoxicillin-clavulanate potassium (Augmentin), 500 mg twice daily||17.50|
|Acute uncomplicated pyelonephritis||Oral||14 days||Trimethoprim-sulfamethoxazole DS, one double-strength tablet twice daily||36.00 (11.00)|
|Ciprofloxacin, 500 mg twice daily|
|Levofloxacin, 250 mg per day||92.00|
|Enoxacin, 400 mg twice daily||191.00|
|Sparfloxacin, 400 mg initial dose, then 200 mg per day||104.50|
|Ofloxacin, 400 mg twice daily||124.50|
|Cefpodoxime, 200 mg twice daily||111.00|
|Cefixime, 400 mg per day||97.50|
|Parenteral||Up to 3 days†||Trimethoprim-sulfamethoxazole 160/800 twice daily||32.00 (12.00) per day|
|Ceftriaxone (Rocephin), 1 g per day||41.00 per day|
|Ciprofloxacin, 400 mg twice daily||57.50 per day|
|Ofloxacin, 400 mg twice daily||53.00 per day|
|Levofloxacin, 250 mg||21.00 per day|
|Aztreonam (Azactam), 1 g three times daily||48.50 per day|
|Gentamicin (Garamycin), 3 mg per kg per day in3 divided doses every 8 hours‡§||12.50 (2.50 to3.00) per day|
|Complicated urinary tract infections||Oral||14 days||Fluoroquinolones|
|Parenteral||Up to 3 days||Ampicillin, 1 g every six hours, and gentamicin, 3 mg per kg per day‡||14.50 (5.00 to5.50) per day|
|Urinary tract infections in young men||Oral||Seven days||Trimethoprim-sulfamethoxazole, one double-strength tablet twice daily||18.00 (5.50)|
|Urinary tract infections in pregnant women||Oral||Three to seven days||Amoxicillin, 250 mg three times daily||2.00 to 4.50(1.00 to 5.00)|
|Asymptomatic bacteriuria in pregnant women||Oral||Three to seven days||Nitrofurantoin, 100 mg four times daily||12.00 to 28.00(14.00 to 32.50)|
Fosfomycin tromethamine (Monurol) can be given as a single oral 3-g sachet for the treatment of acute uncomplicated UTIs. This drug is active against E. coli, enterococci and Citrobacter, Enterobacter, Klebsiella and Serratia species. The clinical cure rate is estimated to be as high as 99 percent. Fosfomycin may be safely used in pregnancy.13
Single-dose antibiotic therapy fell into disfavor when it was observed that women had a high risk of recurrence within six weeks of the initial treatment.14,15 The risk was attributed to the failure of single-dose antibiotics to eradicate gram-negative bacteria from the rectum, the source or reservoir for ascending uropathogens.
Unlike single-dose antibiotic therapy, a three-day regimen reduces rectal carriage of gram-negative bacteria and is not associated with a high recurrence rate. Thus, three-day regimens appear to offer the optimal combination of convenience, low cost and an efficacy comparable to that of seven-day or longer regimens but with fewer side effects.11
One randomized trial16 compared three days of trimethoprim-sulfamethoxazole therapy, one double-strength tablet twice daily, with three days of treatment using the following drugs: nitrofurantoin (Macrodantin), 100 mg four times daily; cefadroxil, 500 mg twice daily; and amoxicillin, 500 mg three times daily. Trimethoprim-sulfamethoxazole was found to be the most cost-effective treatment. Three-day regimens of ciprofloxacin, 250 mg twice daily, and ofloxacin, 200 mg twice daily, were recently compared with three-day trimethoprim-sulfamethoxazole therapy.3,11 The oral fluoroquinolones produced better cure rates with less toxicity, but at a greater overall cost.
Quinolones that are useful in treating complicated and uncomplicated cystitis include ciprofloxacin, norfloxacin, ofloxacin, enoxacin (Penetrex), lomefloxacin (Maxaquin), sparfloxacin (Zagam) and levofloxacin (Levaquin).11 The newer fluoroquinolone, sparfloxacin, in a dosage of 400 mg per day as the initial dose and then 200 mg per day for two days, is equivalent to three days of therapy with ofloxacin or ciprofloxacin. However, sparfloxacin can cause phototoxicity, and it has also been associated with prolongation of the QT interval.17
On the basis of cost and efficacy, trimethoprim-sulfamethoxazole remains the antibiotic of choice in the treatment of uncomplicated UTIs in young women. The use of fluoroquinolones as first-line therapy for uncomplicated UTIs should be discouraged, except in patients who cannot tolerate sulfonamides or trimethoprim, who have a high frequency of antibiotic resistance because of recent antibiotic treatment or who reside in an area in which significant resistance to trimethoprim-sulfamethoxazole has been noted. Three days is the optimal duration of treatment for uncomplicated cystitis. A seven-day course should be considered in pregnant women, diabetic women and women who have had symptoms for more than than one week and thus are at higher risk for pyelonephritis because of the delay in treatment.
Recurrent Cystitis in Young Women
Up to 20 percent of young women with acute cystitis develop recurrent UTIs. During these recurrent episodes, the causative organism should be identified by urine culture and then documented to help differentiate between relapse (infection with the same organism) and recurrence (infection with different organisms). Multiple infections caused by the same organism are, by definition, complicated UTIs and require longer courses of antibiotics and possibly further diagnostic tests (see the discussion of complicated UTIs). Fortunately, most recurrent UTIs in young women are uncomplicated infections caused by different organisms. These infections are generally not associated with underlying anatomic abnormalities and do not require further work-up of the genitourinary tract.5,11,18
Acute self-treatment with a three-day course of standard therapy.
Postcoital prophylaxis with one-half of a trimethoprim-sulfamethoxazole double-strength tablet (40/200 mg) if the UTIs have been clearly related to intercourse.
Continuous daily prophylaxis with one of these regimens for a period of six months: trimethoprim-sulfamethoxazole, one-half tablet per day (40/200 mg); nitrofurantoin, 50 to 100 mg per day; norfloxacin, 200 mg per day; cephalexin (Keflex), 250 mg per day; or trimethoprim, 100 mg per day.
Each of these regimens has been shown to decrease the morbidity of recurrent UTIs without a concomitant increase in antibiotic resistance. Long-term studies have shown antibiotic prophylaxis to be effective for up to five years with trimethoprim, trimethoprim-sulfamethoxazole or nitrofurantoin, without the emergence of drug resistance.3,19 Unfortunately, antibiotic prophylaxis does not appear to alter the natural history of recurrences because 40 to 60 percent of these women reestablish their pattern or frequency of infections within six months of stopping prophylaxis.19
A complicated UTI is one that occurs because of anatomic, functional or pharmacologic factors that predispose the patient to persistent infection, recurrent infection or treatment failure. These factors include conditions often encountered in elderly men, such as enlargement of the prostate gland, blockages and other problems necessitating the placement of indwelling urinary devices, and the presence of bacteria that are resistant to multiple antibiotics. Although antibiotic-susceptible E. coli is responsible for more than 80 percent of uncomplicated UTIs, it accounts for fewer than one third of complicated cases.1,3 Clinically, the spectrum of complicated UTIs may range from cystitis to urosepsis with septic shock.
Accurate urine culture and susceptibility information are necessary to best target and eradicate the pathogens in complicated UTIs. These infections are usually associated with high-count bacteriuria (greater than 100,000 CFU per mL of urine). Occasionally, lower quantitative counts may be encountered in patients who are undergoing diuresis or who are in renal failure. The initial empiric therapy for these patients should include an agent with a broad spectrum of activity against the expected uropathogens. Treatment most often includes a fluoroquinolone, administered orally if possible. In patients who are unable to tolerate oral medication or who require hospitalization for concomitant medical problems, appropriate initial therapy may be parenteral administration of one of the following: a third-generation cephalosporin with antipseudomonal activity such as ceftazidime (Fortaz) or cefoperazone (Cefobid), cefepime (Maxipime), aztreonam (Azactam), imipenemcilastatin (Primaxin) or the combination of an antipseudomonal penicillin (ticarcillin [Ticar], mezlocillin [Mezlin], piperacillin [Pipracil]) with an aminoglycoside.
Enterococci are frequently encountered uropathogens in complicated UTIs. In areas in which vancomycin-resistant Enterococcus faecium is prevalent, the investigational agent quinupristindalfopristin (Synercid) may be useful.20
Patients with complicated UTIs require at least a 10- to 14-day course of therapy. Follow-up urine cultures should be performed within 10 to 14 days after treatment to ensure that the uropathogen has been eradicated. Recent studies have shown that patients initially placed on parenteral therapy can be switched to oral therapy within 72 hours as long as they are clinically improving and able to tolerate the oral agent, and a regimen is available that covers the identified pathogen(s).11,21
Women with acute uncomplicated pyelonephritis may present with one of the following: a mild cystitis-like illness and accompanying flank pain; a more severe illness with fever, chills, nausea, vomiting, leukocytosis and abdominal pain; or a serious gram-negative bacteremia. The microbiologic features of acute uncomplicated pyelonephritis mirror cystitis, except that S. saprophyticus is a rare cause. In most patients, uncomplicated pyelonephritis is caused by specific uropathogenic strains of E. coli possessing adhesins that permit ascending infection of the urinary tract.
The diagnosis should be confirmed by urinalysis with examination for pyuria and/or white blood cell casts and by urine culture. Urine cultures demonstrate more than 100,000 CFU per mL of urine in 80 percent of women with pyelonephritis. Blood cultures are positive in up to 20 percent of women who have this infection. With the exceptions of white cell casts on urinalysis, and bacteremia and flank pain on physical examination, none of the physical or laboratory findings are specific for pyelonephritis.3
Oral therapy should be considered in women with mild to moderate symptoms who are compliant with therapy and can tolerate oral antibiotics but do not have other significant conditions, including pregnancy and gastrointestinal upset. Since E. coli resistance to ampicillin, amoxicillin and first-generation cephalosporins exceeds 30 percent in most locales, these agents should not be used empirically for the treatment of pyelonephritis.11 Even though trimethoprim-sulfamethoxazole is often considered the treatment of choice, resistance to this drug combination may exceed 15 percent in some regions. In those instances, empiric therapy using an oral fluoroquinolone should be considered.
Patients who are too ill to take oral antibiotics or who are unable to take them should initially be treated with parenterally administered single agents, such as trimethoprim-sulfamethoxazole, a third-generation cephalosporin, aztreonam, a broad-spectrum penicillin, a quinolone or an aminoglycoside. The choice of antibiotic is largely empiric, but Gram staining of the urine may be helpful. Once these patients have improved clinically (usually by day 3), they can be switched to oral therapy based on the results of culture and sensitivity studies.11
The total duration of therapy need not exceed 14 days, regardless of the initial bacteremia. Patients with persistent symptoms after three days of appropriate antimicrobial therapy should be evaluated by renal ultrasonography or computed tomography for evidence of urinary obstruction or abscess. In the small percentage of patients who relapse after a two-week course, a repeated six-week course is usually curative.11
UTI in Men
Urinary tract infections most commonly occur in older men with prostatic disease, outlet obstruction or urinary tract instrumentation. These infections occasionally occur in young men who participate in anal sex (exposure to E. coli in the rectum), who are not circumcised (increased E. coli colonization of the glans and prepuce) or whose sexual partner is colonized with uropathogens.22
In men (unlike in women), a urine culture growing more than 1,000 CFU of a pathogen per mL of urine is the best sign of a urinary tract infection, with a sensitivity and specificity of 97 percent.23 Men with urinary tract infections should receive a minimum of seven days of antibiotic therapy (either trimethoprim-sulfamethoxazole or a fluoroquinolone). However, more extensive courses may be required in, for example, men with associated urinary tract infection and prostatitis. Consensus regarding the need for a urologic work-up in men with urinary tract infections is lacking. Among young men with acute cystitis who respond to seven days of treatment, diagnostic work-ups beyond cultures are generally unrewarding.24 Urologic evaluation should be performed routinely in adolescents and men with pyelonephritis or recurrent infections.11,25 When bacterial prostatitis is the source of a urinary tract infection, eradication usually requires antibiotic therapy for six to 12 weeks and in rare instances even longer.
Between 10 and 20 percent of patients who are hospitalized receive an indwelling Foley catheter. Once this catheter is in place, the risk of bacteriuria is approximately 5 percent per day. With long-term catheterization, bacteriuria is inevitable. Catheter-associated urinary tract infections account for 40 percent of all nosocomial infections and are the most common source of gram-negative bacteremia in hospitalized patients.26
The diagnosis of catheter-associated urinary tract infection can be made when the urine culture shows 100 or more CFU per mL of urine from a catheterized patient. The microbiology of catheter-associated urinary tract infections includes E. coli and Proteus, Enterococcus, Pseudomonas, Enterobacter, Serratia and Candida species. The bacterial distribution reflects the nosocomial origin of the infections because so many of the uropathogens are acquired exogenously via manipulation of the catheter and drainage device. Bacteriuria is often polymicrobic, especially in patients with long-term indwelling urinary catheters.
Symptomatic bacteriuria in a patient with an indwelling Foley catheter should be treated with antibiotics that cover potential nosocomial uropathogens. Patients with mild to moderate infections may be treated with one of the oral quinolones, usually for 10 to 14 days. Parenteral antibiotic therapy may be necessary in patients with severe infections or patients who are unable to tolerate oral medications. The recommended duration of therapy for severe infections is 14 to 21 days. Treatment is not recommended for catheterized patients who have asymptomatic bacteriuria, with the following exceptions: patients who are immunosuppressed after organ transplantation, patients at risk for bacterial endocarditis and patients who are about to undergo urinary tract instrumentation.26
Bacteriuria is almost inevitable with long-term catheterization, and prevention strategies have largely been unsuccessful. In such patients, catheters should be changed periodically to prevent the formation of concretions and obstruction that can lead to infection. Prophylactic systemic antibiotics have been shown to delay the onset of bacteriuria in catheterized patients, but this strategy may lead to increased bacterial resistance.26 Prophylactic antibiotic therapy has been successful in reducing the frequency of bacteriuria only in patients who can be weaned from indwelling catheters to intermittent catheterization.
Asymptomatic bacteriuria is defined as the presence of more than 100,000 CFU per mL of voided urine in persons with no symptoms of urinary tract infection. The largest patient population at risk for asymptomatic bacteriuria is the elderly. Up to 40 percent of elderly men and women may have bacteriuria without symptoms. Although early studies noted an association between bacteriuria and excess mortality, more recent studies have failed to demonstrate any such link.27 In fact, aggressively screening elderly persons for asymptomatic bacteriuria and subsequent treatment of the infection has not been found to reduce either infectious complications or mortality. Consequently, this approach currently is not recommended.
Three groups of patients with asymptomatic bacteriuria have been shown to benefit from treatment: (1) pregnant women, (2) patients with renal transplants and (3) patients who are about to undergo genitourinary tract procedures.3 Between 2 and 10 percent of pregnancies are complicated by UTIs; if left untreated, 25 to 30 percent of these women develop pyelonephritis.28,29 Pregnancies that are complicated by pyelonephritis have been associated with low-birth-weight infants and prematurity. Thus, pregnant women should be screened for bacteriuria by urine culture at 12 to 16 weeks of gestation. The presence of 100,000 CFU of bacteria per mL of urine is considered significant.
Pregnant women with asymptomatic bacteriuria should be treated with a three- to seven-day course of antibiotics, and the urine should subsequently be cultured to ensure cure and the avoidance of relapse.29 Although amoxicillin is frequently suggested as the agent of choice, E. coli is now commonly resistant to ampicillin, amoxicillin and cephalexin. Thus, treatment should be based on the results of susceptibility tests. Nitrofurantoin or trimethoprim-sulfamethoxazole may also be used; however, caution should be exercised in the third trimester because the sulfonamides compete with bilirubin binding in the newborn.
Symptomatic urinary tract infections complicate 1 to 2 percent of pregnancies, usually in women with persistent bacteriuria.28,29 Most pregnant women with pyelonephritis should be hospitalized. Initially, these patients should receive intravenous antibiotic therapy. They should complete a 14-day course of acute antibiotic therapy followed by nightly suppressive therapy until delivery. Recent studies have shown that selected pregnant women with pyelonephritis can be treated with either outpatient intramuscularly administered ceftriaxone (Rocephin) or orally administered cephalexin.28 Ceftriaxone, a third-generation parenterally administered cephalosporin, is a suitable agent for inpatient treatment. Tetracyclines and fluoroquinolones should be avoided in pregnancy.