These are the most common bacterial infections during pregnancy. Although asymptomatic bacteriuria is the most common, symptomatic infection includes cystitis, or it may involve the renal calyces, pelvis, and parenchyma—pyelonephritis.
Organisms that cause urinary infections are those from the normal perineal flora. Approximately 90 percent of Escherichia coli strains that cause nonobstructive pyelonephritis have adhesins such as P- and S-fimbriae. These are cell-surface protein structures that enhance bacterial adherence and thereby, virulence (Foxman, 2010; Hooton, 2012). These adhesins promote binding to vaginal and uroepithelial cells through expression of the PapG gene that encodes the P-fimbriae tip and by production of toxins and other virulence factors (Spurbeck, 2011).
Data suggest that pregnant women have more severe sequelae from urosepsis. The T-helper cell—Th1/Th2 ratio—reversal of normal pregnancy is discussed in Chapter 4 (Immunological Functions). There are also various perturbations of cytokine expression that have been reported (Chaemsaithong, 2013). And maternal deaths have been attributed to E coli bearing Dr+ and P adhesins (Sledzińska, 2011). But even if pregnancy itself does not enhance these virulence factors, urinary stasis, vesicoureteral reflux, and diabetes predispose to symptomatic upper urinary infections (Czaja, 2009; Twickler, 1994).
In the puerperium, there are several risk factors that predispose a woman to urinary infections. Bladder sensitivity to intravesical fluid tension is often decreased as a consequence of labor trauma or conduction analgesia (Chap. 36, Perineal Care). Sensation of bladder distention can also be diminished by discomfort caused by an episiotomy, periurethral lacerations, or vaginal wall hematomas. Normal postpartum diuresis may worsen bladder overdistention, and catheterization to relieve retention commonly leads to urinary infection. Postpartum pyelonephritis is treated in the same manner as antepartum renal infections (McDonnold, 2012).
This refers to persistent, actively multiplying bacteria within the urinary tract in asymptomatic women. Its prevalence in nonpregnant women is 5 to 6 percent and depends on parity, race, and socioeconomic status (Hooton, 2000). The highest incidence is in African-American multiparas with sickle-cell trait, and the lowest incidence is in affluent white women of low parity. Asymptomatic infection is also more common in diabetics (Schneeberger, 2014). Because in most women there is recurrent or persistent bacteriuria, it frequently is discovered during prenatal care. The incidence during pregnancy is similar to that in nonpregnant women and varies from 2 to 7 percent.
Bacteriuria is typically present at the first prenatal visit. An initial positive urine culture result prompts treatment, after which, fewer than 1 percent of women develop a urinary tract infection (Whalley, 1967). A clean-voided specimen containing more than 100,000 organisms/mL is diagnostic. It may be prudent to treat when lower concentrations are identified, because pyelonephritis develops in some women despite colony counts of only 20,000 to 50,000 organisms/mL (Lucas, 1993).
If asymptomatic bacteriuria is not treated, approximately 25 percent of infected women will develop symptomatic infection during pregnancy. Eradication of bacteriuria with antimicrobial agents prevents most of these. The American Academy of Pediatrics and the American College of Obstetricians and Gynecologists (2012), as well as a U.S. Preventive Services Task Force (2008), recommend screening for bacteriuria at the first prenatal visit (Chap. 9, Intimate Partner Violence). Standard urine cultures may not be cost effective when the prevalence is low. However, less expensive screening tests such as the leukocyte esterase-nitrite dipstick are when the prevalence is 2 percent or less (Rouse, 1995). Because of a high prevalence—5 to 8 percent—at Parkland Hospital, culture screening is done in most women. Susceptibility determination is not necessary because initial treatment is empirical (Hooton, 2012). Also, a dipstick culture technique has excellent positive- and negative-predictive values (Mignini, 2009). With this, a special agar-coated dipstick is first placed into urine and then also serves as the culture plate.
In some, but not all studies, covert bacteriuria has been associated with preterm or low-birthweight infants (Kass, 1962). It is even more controversial whether eradication of bacteriuria decreases these complications. Evaluating a cohort of 25,746 mother-infant pairs, Schieve and coworkers (1994) reported urinary tract infection to be associated with increased risks for low-birthweight infants, preterm delivery, pregnancy-associated hypertension, and anemia. These findings vary from those of Gilstrap and colleagues (1981b) and Whalley (1967). In most studies, asymptomatic infection is not evaluated separately from acute renal infection (Banhidy, 2007). A Cochrane database review by Vasquez and Abalos (2011) found that benefits of treatment for asymptomatic bacteria are limited to the reduction of the incidence of pyelonephritis.
Bacteriuria responds to empirical treatment with any of several antimicrobial regimens listed in Table 53-1. Although selection can be based on in vitro susceptibilities, in our extensive experience, empirical oral treatment for 10 days with nitrofurantoin macrocrystals, 100 mg at bedtime, is usually effective. Lumbiganon and associates (2009) reported satisfactory results with a 7-day oral course of nitrofurantoin, 100 mg given twice daily. Single-dose antimicrobial therapy has also been used with success for bacteriuria. The important caveat is that, regardless of regimen given, the recurrence rate is approximately 30 percent. This may indicate covert upper tract infection and the need for longer therapy.
TABLE 53-1Oral Antimicrobial Agents Used for Treatment of Pregnant Women with Asymptomatic Bacteriuria ||Download (.pdf) TABLE 53-1 Oral Antimicrobial Agents Used for Treatment of Pregnant Women with Asymptomatic Bacteriuria
|Single-dose treatment |
| Amoxicillin, 3 g |
| Ampicillin, 2 g |
| Cephalosporin, 2 g |
| Nitrofurantoin, 200 mg |
| Trimethoprim-sulfamethoxazole, 320/1600 mg |
|3-day course |
| Amoxicillin, 500 mg three times daily |
| Ampicillin, 250 mg four times daily |
| Cephalosporin, 250 mg four times daily |
| Ciprofloxacin, 250 mg twice daily |
| Levofloxacin, 250 or 500 mg daily |
| Nitrofurantoin, 50 to 100 mg four times daily or 100 mg twice daily |
| Trimethoprim-sulfamethoxazole, 160/800 mg two times daily |
| Nitrofurantoin, 100 mg four times daily for 10 days |
| Nitrofurantoin, 100 mg twice daily for 5 to 7 days |
| Nitrofurantoin, 100 mg at bedtime for 10 days |
|Treatment failures |
| Nitrofurantoin, 100 mg four times daily for 21 days |
|Suppression for bacterial persistence or recurrence |
| Nitrofurantoin, 100 mg at bedtime for pregnancy remainder |
Periodic surveillance is necessary to prevent recurrent urinary infections (Schneeberger, 2012). For recurrent bacteriuria, we have had success with nitrofurantoin, 100 mg orally at bedtime for 21 days (Lucas, 1994). For women with persistent or frequent bacteriuria recurrences, suppressive therapy for the remainder of pregnancy can be given. We routinely use nitrofurantoin, 100 mg orally at bedtime. This drug may rarely cause an acute pulmonary reaction that dissipates on its withdrawal (Boggess, 1996).
Lower urinary infection during pregnancy may develop without antecedent covert bacteriuria (Harris, 1981). Cystitis is characterized by dysuria, urgency, and frequency, but with few associated systemic findings. Pyuria and bacteriuria are usually found. Microscopic hematuria is common, and occasionally there is gross hematuria from hemorrhagic cystitis (Fakhoury, 1994). Although cystitis is usually uncomplicated, the upper urinary tract may become involved by ascending infection. Almost 40 percent of pregnant women with acute pyelonephritis have preceding symptoms of lower tract infection (Gilstrap, 1981a).
Women with cystitis respond readily to any of several regimens. Most of the three-day regimens listed in Table 53-1 are usually 90-percent effective (Fihn, 2003). Single-dose therapy is less effective, and if it is used, concomitant pyelonephritis must be confidently excluded.
Lower urinary tract symptoms with pyuria accompanied by a sterile urine culture may be from urethritis caused by Chlamydia trachomatis. Mucopurulent cervicitis usually coexists, and azithromycin therapy is effective. (Chap. 65, Chlamydial Infections)
Renal infection is the most common serious medical complication of pregnancy. In a study of the 2006 Nationwide Inpatient Sample by Jolley and coworkers (2012), there were nearly 29,000 hospitalizations for acute pyelonephritis. Rates were highest for adolescents at 17.5 per 1000 and for Hispanic women at 10.1 per 1000. In another study of more than 70,000 pregnancies in a managed care organization, Gazmararian and colleagues (2002) reported that 3.5 percent of antepartum admissions were for urinary infections. The potential seriousness is underscored by the observations of Snyder and associates (2013) that pyelonephritis was the leading cause of septic shock during pregnancy. And in a 2-year audit of admissions to the Parkland Hospital Obstetrical Intensive Care Unit, 12 percent of antepartum admissions were for sepsis syndrome caused by pyelonephritis (Zeeman, 2003). There is also concern that urosepsis may be related to an increased incidence of cerebral palsy in preterm infants (Jacobsson, 2002). Fortunately, there appear to be no serious long-term maternal sequelae (Raz, 2003).
Renal infection develops more frequently in the second trimester, and nulliparity and young age are associated risk factors (Hill, 2005). Pyelonephritis is unilateral and right-sided in more than half of cases, and it is bilateral in a fourth. There is usually a rather abrupt onset with fever, shaking chills, and aching pain in one or both lumbar regions. Anorexia, nausea, and vomiting may worsen dehydration. Tenderness usually can be elicited by percussion in one or both costovertebral angles. The urinary sediment contains many leukocytes, frequently in clumps, and numerous bacteria. Bacteremia is demonstrated in 15 to 20 percent of these women. E coli is isolated from urine or blood in 70 to 80 percent of infections, Klebsiella pneumoniae in 3 to 5 percent, Enterobacter or Proteus species in 3 to 5 percent, and gram-positive organisms, including group B Streptococcus and S aureus, in up to 10 percent of cases (Hill, 2005; Wing, 2000). The differential diagnosis includes, among others, labor, chorioamnionitis, appendicitis, placental abruption, or infarcted leiomyoma. Evidence of the sepsis syndrome is common, and this is discussed in detail in Chapter 47 (Sepsis Syndrome).
Plasma creatinine is monitored because early studies reported that 20 percent of pregnant women developed renal dysfunction. More recent findings, however, show this to be only 5 percent if aggressive fluid resuscitation is provided (Hill, 2005). Follow-up studies have demonstrated that this endotoxin-induced damage is reversible in the long term. Varying degrees of respiratory insufficiency from endotoxin-induced alveolar injury are manifest in up to 10 percent of women and may result in frank pulmonary edema (Cunningham, 1987; Sheffield, 2005; Snyder, 2013). In some cases, pulmonary injury may be so severe that it causes acute respiratory distress syndrome (ARDS) (Fig. 53-3).
A series of anterior-posterior projection chest radiographs of improving acute respiratory distress syndrome (ARDS) in a second-trimester pregnant woman with severe pyelonephritis. A. An extensive infiltrative process and complete obliteration of the diaphragm (white arrows) is seen. B. Improved aeration of lung fields bilaterally is noted as pleural disease resolves (arrows). C. Markedly improved visualization of the lungs fields with residual platelike atelectasis and normal appearance of the diaphragm.
Uterine activity from endotoxin is common and is related to fever severity (Graham, 1993). In the study by Millar and coworkers (2003), women with pyelonephritis averaged 5 contractions per hour at admission, and this decreased to 2 per hour within 6 hours of intravenous fluid and antimicrobial administration. As discussed in Chapter 47 (Acute Pulmonary Edema), b-agonist therapy for tocolysis increases the likelihood of respiratory insufficiency from permeability edema because of the sodium- and fluid-retaining properties of those agents (Lamont, 2000). The incidence of pulmonary edema in women with pyelonephritis who were given b-agonists was reported to be 8 percent—a fourfold increase over that expected (Towers, 1991).
Endotoxin-induced hemolysis is common, and approximately a third of patients with pyelonephritis develop anemia (Cox, 1991). With recovery, hemoglobin regeneration is normal because acute infection does not affect erythropoietin production (Cavenee, 1994).
One scheme for management of acute pyelonephritis is shown in Table 53-2. Although we routinely obtain urine and blood cultures, prospective trials show them to be of limited clinical utility (Wing, 2000). Intravenous hydration to ensure adequate urinary output is the cornerstone of treatment. Antimicrobials are also begun promptly with the caveat that they may initially worsen endotoxemia from bacterial lysis. Ongoing surveillance for worsening of sepsis syndrome is monitored by serial determinations of urinary output, blood pressure, pulse, temperature, and oxygen saturation. High fever should be lowered with a cooling blanket or acetaminophen. This is especially important in early pregnancy because of possible teratogenic effects of hyperthermia (Chap. 14, Risk Factors).
TABLE 53-2Management of the Pregnant Woman with Acute Pyelonephritis ||Download (.pdf) TABLE 53-2 Management of the Pregnant Woman with Acute Pyelonephritis
|Hospitalize patient |
|Obtain urine and blood cultures |
|Evaluate hemogram, serum creatinine, and electrolytes |
|Monitor vital signs frequently, including urinary output—consider indwelling catheter |
|Establish urinary output ≥ 50 mL/hr with intravenous crystalloid solution |
|Administer intravenous antimicrobial therapy (see text) |
|Obtain chest radiograph if there is dyspnea or tachypnea |
|Repeat hematology and chemistry studies in 48 hours |
|Change to oral antimicrobials when afebrile |
|Discharge when afebrile 24 hours, consider antimicrobial therapy for 7 to 10 days |
|Repeat urine culture 1 to 2 weeks after antimicrobial therapy completed |
Antimicrobial therapy usually is empirical, and ampicillin plus gentamicin; cefazolin or ceftriaxone; or an extended-spectrum antibiotic were all 95-percent effective in randomized trials (Sanchez-Ramos, 1995; Wing, 1998, 2000). Fewer than half of E coli strains are sensitive to ampicillin in vitro, but cephalosporins and gentamicin generally have excellent activity. Serum creatinine is monitored if nephrotoxic drugs are given. Initial treatment at Parkland Hospital is ampicillin plus gentamicin. Some recommend suitable substitutes if bacterial studies show in vitro resistance. With any of the regimens discussed, response is usually prompt, and 95 percent of women are afebrile by 72 hours (Hill, 2005; Sheffield, 2005; Wing, 2000). After discharge, most recommend oral therapy for a total of 7 to 14 days (Hooton, 2012).
Generally, intravenous hydration and antimicrobial therapy are followed by stepwise defervescence of approximately 1°F per day. With persistent spiking fever or lack of clinical improvement by 48 to 72 hours, urinary tract obstruction or another complication or both are considered. Renal sonography is recommended to search for obstruction manifest by abnormal ureteral or pyelocaliceal dilatation (Seidman, 1998). Although most women with continuing infection have no evidence of obstruction, some are found to have calculi. Although renal sonography will detect hydronephrosis, stones are not always seen in pregnancy (Butler, 2000; Maikranz, 1987). If stones are strongly suspected despite a nondiagnostic sonographic examination, a plain abdominal radiograph will identify nearly 90 percent. Another option is the modified one-shot intravenous pyelogram—a single radiograph obtained 30 minutes after contrast injection—which usually provides adequate imaging (Butler, 2000).
In some women, MR imaging may disclose the cause of persistent infection (Spencer, 2004). Even without urinary obstruction, persistent infection can be due to an intrarenal or perinephric abscess or phlegmon (Cox, 1988; Rafi, 2012). Obstruction relief is important, and one method is cystoscopic placement of a double-J ureteral stent (Rodriguez, 1988). Because these stents are usually left in place until after delivery, they frequently become encrusted and require replacement. We have found that percutaneous nephrostomy is preferable because the stents are more easily replaced. Finally, surgical removal of stones may be required in some women (Pregnancy after Renal Transplantation).
Outpatient Management of Pyelonephritis
Outpatient management is an option for nonpregnant women with uncomplicated pyelonephritis (Hooton, 2012). Wing and associates (1999) have described outpatient management in 92 pregnant women who were first given in-hospital intramuscular ceftriaxone, two 1-g doses 24 hours apart. At this point, one third of the group was considered suitable for outpatient therapy, and these women were randomized either to discharge and oral antimicrobials or to continued hospitalization with intravenous therapy. A third of the outpatient management group was unable to adhere to the treatment regimen and was admitted. These findings suggest that outpatient management is applicable to very few pregnant women.
Recurrent infection—either covert or symptomatic—is common and develops in 30 to 40 percent of women following completion of treatment for pyelonephritis (Cunningham, 1973). Unless other measures are taken to ensure urine sterility, nitrofurantoin, 100 mg orally at bedtime given for the remainder of the pregnancy, reduces bacteriuria recurrence (Van Dorsten, 1987).
Vesicoureteral reflux in early childhood can cause recurrent urinary tract infections, and thus, subsequent chronic interstitial nephritis was attributed to chronic pyelonephritis. Moreover, it was also found that high-pressure sterile reflux impaired normal renal growth. Combined, this leads to patchy interstitial scarring, tubular atrophy, and loss of nephron mass and is termed reflux nephropathy. In adults, long-term complications include hypertension, which may be severe if there is demonstrable renal damage (Diamond, 2012; Köhler, 2003).
Perhaps half of women with reflux nephropathy were treated during childhood for renal infections. Of these, many also had surgical correction of reflux as children, and these commonly have bacteriuria when pregnant (Mor, 2003). In the other half of women with reflux nephropathy, there is no clear history of recurrent cystitis, acute pyelonephritis, or obstructive disease (Diamond, 2012). Reports describing 939 pregnancies in 379 women with reflux nephropathy indicate that impaired renal function and bilateral renal scarring were associated with increased maternal complications (El-Khatib, 1994; Jungers, 1996; Köhler, 2003). Chronic renal disease and pregnancy outcome is discussed further on Pregnancy and Chronic Renal Disease.