1. Acute Pyelonephritis in Pregnancy

Acute pyelonephritis is one of the most common indications for antepartum hospitalization, complicating 1–2% of all pregnancies. Women with asymptomatic bacteriuria (ASB), traditionally defined as a urine culture from mid-stream collection with a single isolate of more than 100 000 colony-forming units (cfu) of a uropathogen,[1] have a 20- to 30-fold increased risk of developing pyelonephritis in pregnancy compared with women without bacteriuria.[2] In addition, treatment of ASB in pregnancy decreases the risk of subsequent pyelonephritis from approximately 20–35% to 1–4%.[3] Pyelonephritis in pregnancy occurs mostly prior to delivery, with 10–20% of cases diagnosed in the first trimester[4,5] and the majority of cases occurring in the second and third trimesters. Pyelonephritis can also occur postpartum.[6]

Numerous physiological changes predispose the pregnant woman to urinary tract infection (UTI). Ureteral and renal calyceal dilatation are evident as early as 12 weeks’ gestation, thought to be due to progesterone-induced relaxation. Ureteral peristalsis slows and the enlarging uterus compresses the ureters, particularly on the right side. Mechanical compression of the bladder and decreased detrusor tone lead to increased capacity and incomplete emptying of the bladder. All of these factors contribute to ureteral dilatation and urinary stasis. Increases in glomerular filtration with resultant elevations in urinary glucose levels and alkalization of urine also facilitate bacterial growth.

Women with urinary tract anomalies such as incompetent vesicourethral valves and renal calculi, medical conditions including diabetes mellitus, sickle cell disease or trait, and neurological problems such as paralysis from spinal cord injury, are at increased risk for acquiring pyelonephritis in pregnancy. Pyelonephritis is also more likely to occur in women of low socioeconomic status.[7,8] In a retrospective review of 242 nonpregnant women aged 18–49 years with pyelonephritis, other risk factors for the disease included frequency of sexual intercourse in the previous 30 days, recent spermicide use, recent UTI and recent incontinence.[9] The overlap of these risk factors with those described for women with cystitis supports the view that pyelonephritis usually involves ascent of infecting microbes from the bladder.[9] Another study noted that low expression of an interleukin receptor may confer a familial susceptibility to pyelonephritis.[10]

Clinical signs and symptoms of acute pyelonephritis include fever, shaking chills, flank pain, nausea and vomiting, costovertebral angle tenderness (CVAT) and, less commonly, symptoms of cystitis such as dysuria and increased frequency.[6] Urinary dipstick testing for the presence of leukocyte esterase and nitrites may be positive.[11,12] The diagnosis is confirmed by urine culture, obtained usually by midstream clean-catch, but occasionally by suprapubic aspiration or urethral catheterization. To obtain a proper clean-catch specimen, the patient should be instructed to wipe her introitus from front to back prior to micturition in order to avoid contamination with periurethral bacteria. According to the Infectious Diseases Society of America (IDSA) consensus definition of pyelonephritis, colony counts of ≥100 000 cfu/mL from clean-catch voided specimens are acceptable for use in antimicrobial treatment studies, and provide a sensitivity of 90–95%.[13] One or two bacteria per high-power field on an unspun catheterized urine specimen, or more than 20 bacteria per high-power field on spun urine, closely correlates with >100 000 cfu/mL bacteria on urine culture.[14] Pyuria or the presence of leukocyte casts are also consistent with the diagnosis.

Uropathogens responsible for pyelonephritis are taxonomically the same as those that cause ASB and cystitis. The most common uropathogen is Escherichia coli, which is cultured from 70–85% of patients.[4,6] Other Gram-negative uropathogens include Klebsiella, Enterobacter and Proteus spp. A contemporary cohort study[4] noted far fewer infections caused by Klebsiella or Enterobacter species (3%) than the 23% of cases reported historically.[6] Gram-positive organisms that are frequently identified include Enterococcus faecalis and group B streptococci. Gram-positive microbes are found more commonly as pregnancy progresses.[4]

The pathogenesis of ASB and risk factors for progression to symptomatic UTI, including pyelonephritis, are incompletely understood. A complex interplay between virulence factors of uropathogenic bacterial species, such as E. coli and Proteus mirabilis, and host defence mechanisms likely dictates the severity of UTIs in women. For example, E. coli from a small group of O-serotypes have characteristics epidemiologically associated with acute pyelonephritis, chronic or recurrent infection, parenchymal scarring and renal failure in the normal urinary tract.[15] The main reservoir for E. coli causing UTI is the intestinal tract, but only a small proportion of E. coli bacteria that inhabit the intestines actually cause UTI. Factors that enhance the adherence of E. coli to uroepithelial cells and thereby protect the bacteria from urinary lavage and increase their ability to multiply and invade renal tissue have been identified. These include adhesins such as P-fimbria and S-fimbria,[16–19] and haemolysin production.[20] Adhesins can bind erythrocyte membranes and inhibit serum bactericidal activity by expression of genes associated with ampicillin resistance.[21] The class of Dr adhesins has been associated with pyelonephritis in pregnancy and a high rate of preterm delivery in mice.[21–23] Individual genetic factors may also be associated with human bladder immune responses prior to the development of symptomatic UTIs.[24] Contributors to the uropathogenic potential of P. mirabilis include fimbriae and urease production.[25]

In addition to urinalysis and urine culture, laboratory evaluation often includes a complete blood cell count and serum chemistry evaluation. Evidence of haemolysis with elevated lactate dehydrogenase levels may be encountered and is attributed to endotoxin-mediated haemolysis.[26] Electrolyte abnormalities are also common. Transient renal insufficiency as demonstrated by a decrease in creatinine clearance by 50% or more occurs in some women. Previous estimates were that a quarter of women with acute pyelonephritis in pregnancy developed renal insufficiency,[6,27] but modern studies suggest that the rate may be lower (2%), possibly due to earlier presentation for treatment and intravenous fluid hydration.[4,28] Spontaneous resolution of the abnormal renal function should be expected as the acute infection clears.

Blood cultures are often obtained when pyelonephritis is suspected, although the utility of this practice has been questioned since the isolates of blood cultures rarely differ from those of urine cultures.[29–34] In a retrospective study of 391 patients with pyelonephritis in pregnancy, only 6% of patients required changes in antibacterial therapy, most commonly for persistent fever and not correlated with urine or blood culture results.[32] A change in management because of bacteraemia alone occurred in only 1% of cases.[32] The limited utility of routine cultures in the clinical management of patients suggested that patient care would not be compromised in their absence. Some authors recommend obtaining blood cultures only if the patient has a high temperature, e.g. ≥39°C, appears to have sepsis or has a major co-morbidity such as adult respiratory distress syndrome (ARDS).[32,35,36]

When acute pyelonephritis during pregnancy is diagnosed and treated in a timely fashion, the outcome for both mother and fetus is generally good. However, there are serious complications that can arise. The aim of this article is to provide an overview of the diagnosis, possible complications, and management options for women with pyelonephritis during pregnancy. This guide serves as an update of a 2001 article detailing treatment options for optimal outcomes.[37] Although the management of the disease has not changed appreciably over the last 10 years, the disturbing trend of antibacterial resistance has created new challenges in the treatment of women with this disease.

2. Complications of Acute Pyelonephritis in Pregnancy

There are both maternal and fetal complications of acute pyelonephritis in pregnancy. Anaemia is the most common complication encountered in association with the disease, occurring in approximately 25% of patients.[4] Approximately 15–20% of women with pyelonephritis will have bacteraemia.[6] Gram-negative bacteria possess endotoxins that, when released into the maternal circulation, can lead to a cascade response of cytokines, histamine and bradykinins.[38] The resulting capillary endothelial damage, diminished vascular resistance and alterations in cardiac output may lead to serious complications such as septic shock, disseminated intravascular coagulation, respiratory insufficiency or ARDS. The incidence of ARDS with pyelonephritis ranges from 1% to 8%.[39] It manifests with symptoms of dyspnoea, tachypnoea and hypoxaemia. Chest radiographs reveal pulmonary oedema. This condition is usually adequately treated with supplemental oxygen therapy and diuresis. However, it can progress to ventilator dependence. In one retrospective investigation, ARDS was diagnosed more frequently in women who had been treated with β-sympathomimetic tocolytic therapy and had received excessive intravenous hydration.[40]

Septic shock occurs as a result of endotoxaemia. Patients with septic shock require intensive care, often with pulmonary artery catheterization. Antimicrobial therapy and fluid resuscitation should be instituted at once. Dopamine may be required to maintain systolic blood pressure and adequate urine output.

Recurrent pyelonephritis occurs in approximately 20% of women before delivery.[6] Issues relating to costs of care and the possibility of permanent renal damage must be considered when discussing recurrent infection. The frequency of recurrences may be reduced with careful post-treatment surveillance for recurrent infection and the use of suppressive therapy.

The risk of preterm labour and delivery attributable to pyelonephritis in pregnancy is difficult to estimate, particularly because delivery may not occur during the admission for the acute disease, and the risk factors for both pyelonephritis and preterm delivery overlap.[35] Of 368 women delivered at Parkland Hospital (Dallas, TX, USA) with a history of pyelonephritis during the pregnancy, 19 (5%) delivered at less than 37 weeks and only 4 (1%) delivered preterm during their admission for acute pyelonephritis.[4] Although most will not deliver, the majority of women with pyelonephritis in the second and third trimesters will experience uterine contractions;[41,42] thus, the threat of preterm delivery is taken seriously. Despite the presence of uterine contractions, there is often little or no cervical change. It has been suggested that the use of tocolytic therapy be reserved for those patients who do exhibit cervical change, as tocolysis may aggravate the response to endotoxaemia and predispose pregnant women to pulmonary oedema.[40]

3. Conventional Therapy

The historical approach to pyelonephritis in both nonpregnant and pregnant women has been hospitalization and treatment with parenteral antibacterials. The safety profile for the mother and fetus is of utmost importance when choosing an antimicrobial agent for treatment of UTI in pregnancy. Unfortunately, the human data on drug toxicity to mother and fetus are unusually limited, and animal studies of fetal toxicity are often difficult to extrapolate to humans.

As a general rule, amino- and carboxy-penicillin derivatives (either alone or in combination with clavulanic acid), ureidopenicillins and cephalosporins are considered to have good safety when used during pregnancy. Penicillin derivatives and cephalosporins achieve good renal parenchymal and urine concentrations shortly after administration, and have effective spectrums of coverage for common uropathogens. Ampicillin had been the mainstay of treatment for pyelonephritis in pregnancy given the low cost, minimal risk to mother and fetus, and its historical efficacy, but it has a high frequency of resistance in multiple organisms. Approximately 45–60% of E. coli strains are resistant to ampicillin.[21,32,43] As a result, first-generation cephalosporins have replaced ampicillin as first-line monotherapy for pyelonephritis during pregnancy. Cefazolin possesses the same spectrum of activity as broader-spectrum cephalosporins against the common causative microbes and is less expensive.[38]

Tetracyclines are generally contraindicated during pregnancy. These antimicrobials can chelate calcium in fetal structures, and in utero exposure may result in tooth discoloration and inhibition of bone growth.[44] Although fluoroquinolones attain high renal concentration and are a preferred treatment for UTI in nonpregnant patients, the risk of fetal arthropathy proscribes their use during pregnancy.

Aminoglycosides concentrate in renal tissue as they are reabsorbed and bound in the proximal tubular epithelium;[45] thus, they may have a distinctive place in the treatment of pyelonephritis.[46] The use of aminoglycosides should be considered carefully as there is the potential for ototoxicity with their use. To date, gentamicin has been widely used in pregnancy as first-line therapy for Gram-negative coverage with no reports of congenital complications. However, it is a Food and Drug Administration (FDA) category C drug in the US, meaning that studies have shown that the drug exerts teratogenic or embryocidal effects in animals, but there are no controlled trials in women. In addition, ototoxicity has been demonstrated following fetal exposure to related aminoglycoside compounds such as kanamycin and streptomycin.[47–49] Acute renal dysfunction associated with pyelonephritis appears to be less common[4] than the 20% rate noted previously;[6] thus, empirical treatment with gentamicin while awaiting serum creatinine measurement may be appropriate. However, if gentamicin is chosen, monitoring of serum concentrations should be performed and adjustments made in the dose administration as needed. This is partly because of concerns regarding any toxic effects such as exacerbation of renal insufficiency, but also because maternal serum concentrations near term are more likely to be subtherapeutic because of enhanced drug clearance.[50–52]

The optimal antibacterial regimen for the treatment of acute pyelonephritis in pregnancy would be characterized by the following: (i) proven effectiveness in properly conducted, prospective, randomized, double-blind trials; (ii) activity against the uropathogens likely to be responsible for the symptomatic upper UTI; (iii) ability to maintain adequate tissue and serum concentrations throughout the treatment period; (iv) lack of association with the development of resistance to antibacterials; (v) inexpensive; (vi) well tolerated; and (vii) safety for the developing fetus. There are insufficient data to recommend a specific treatment regimen for pyelonephritis in pregnancy as few studies exist comparing the efficacy of different antibacterial agents.[53–59] A Cochrane Database review found that all antibacterials studied were effective in clearing infections and decreasing the incidence of complications such as prolonged pyrexia and preterm delivery.[59] Many regimens are acceptable and appear to be effective (table I).

Table I
figure Tab1

Examples of dosage administration regimens of antibacterial agents for the treatment of pyelonephritis in pregnancy with US FDA classification for use in pregnancy

Full size image

Intravenous antimicrobial therapy is usually continued until the patient is afebrile for 48 hours and symptoms have improved; afterward, the patient is treated with oral antibacterials. The course of oral therapy lasts for 10–14 days. After completion of therapy, a urine culture should be obtained to verify resolution of the bacteriuria. The appropriate duration of therapy for eradication of pyelonephritis has not been subjected to scientific scrutiny.

Recurrent pyelonephritis occurs in 6–8% of women.[60,61] The incidence of recurrent pyelonephritis is decreased in patients treated with antimicrobial suppression for the duration of pregnancy.[62,63] Thus, after treatment for pyelonephritis, patients should receive prophylaxis with nitrofurantoin 100 mg or cephalexin 250–500 mg orally every night during the pregnancy and for 4–6 weeks postpartum. Some practitioners choose to prescribe cotrimoxazole (sulfamethoxazole/trimethoprim), but its use is generally restricted to the second trimester due to its inhibitory effect on folate metabolism in the first trimester and the theoretical risk of inducing neonatal hyperbilirubinaemia when given close to the time of delivery. The sulfonamide moiety of cotrimoxazole competes with bilirubin for binding to plasma albumin after birth, although no cases of kernicterus after in utero exposure to sulfonamides have been reported.[64] It must be noted that prolonged use of antimicrobials such as cephalexin can predispose women to chronic vaginal candidiasis;[65] thus, symptoms should be monitored periodically. In addition to continuous antimicrobial suppression, monthly urine cultures should be obtained to screen for recurrent bacteriuria.

Because most patients with pyelonephritis are dehydrated, intravenous hydration is usually given. Urinary output is monitored carefully. Patients who are treated for pyelonephritis with the appropriate antimicrobial agent usually respond within 48 hours. If the patient fails to respond clinically by 72 hours, further evaluation should ensue for bacterial resistance to the antibacterial used, urolithiasis, perinephric abscess formation or urinary tract abnormalities, and the antibacterial agent should be changed to include an aminoglycoside. Patients with recurrent pyelonephritis or those who fail to respond quickly to aminoglycoside therapy should undergo imaging evaluation for the presence of an anatomic malformation or nephrolithiasis. This may be accomplished safely in pregnancy with ultrasonography or with intravenous pyelography (IVP). To minimize the radiation exposure to the fetus, only one exposure at 20–30 minutes should be obtained for the IVP. Magnetic resonance imaging also may be used if urinary tract obstruction is suspected.

An algorithm for the treatment of pyelonephritis during pregnancy is presented in figure 1.

Fig. 1
figure 1

Algorithm for treatment of pyelonephritis during pregnancy. IV = intravenous; PO = oral.

Full size image

4. Antimicrobial Resistance: A Growing Public Health Concern

Treatment of UTI in nonpregnant patients has been altered dramatically by changing patterns of antimicrobial resistance. Extensive use of amoxicillin after its introduction in the 1970s led to the development of widespread resistance of common uropathogens to this antibacterial. In 1999, the IDSA suggested that uncomplicated UTIs be treated with cotrimoxazole unless local resistance rates of E. coli to the drug exceeded 10–20%, in which case a fluoroquinolone should be used.[66] With the subsequent broad use of cotrimoxazole since that time, increasing levels of resistance to this antimicrobial have been reported.[67,68] Fluoroquinolones have also been impacted, with concomitantly rising resistance rates in a number of uropathogens.[67] Interestingly, E. coli resistance to nitrofurantoin is reported to be low at 1.1–2.3%,[67,69] leading some authors to suggest that it be considered the first-line treatment for uncomplicated UTI.[69]

In 2005, the North American Urinary Tract Infection Collaborative Alliance (NAUTICA) reported on the susceptibility of outpatient uropathogen isolates to antibacterials commonly used to treat UTI.[67] 1990 urinary isolates were obtained from 41 centres in the US and Canada. E. coli represented 57.5% of all outpatient urinary isolates, with Klebsiella pneumoniae (12.4%), Enterococcus spp. (6.6%) and P. mirabilis (5.4%) making up the remaining majority. Among all of the isolates, 45.9% were resistant to ampicillin, 20.4% to cotrimoxazole, 14.3% to nitrofurantoin, 9.7% to ciprofloxacin and 8.1% to levofloxacin.[67] Although all of the characteristics of the patients from whom the samples were derived were not known to the authors, they noted that the observations from the study did not necessarily apply to young ambulatory women with uncomplicated UTIs, but rather might be more relevant to older patients with complicated infections. Of note, resistance rates were higher in the US than in Canada and varied by geographical region. A separate study corroborated that in the US, the highest prevalence of multidrug-resistant phenotypes is found on the Pacific Coast.[70]

Recent prior antibacterial treatment is one of the most important determinants for infection with a resistant microbe.[71] Although there are several important differences between recommendations for treatment of gestational pyelonephritis and antimicrobials commonly used in uncomplicated UTI, the significance of antimicrobial stewardship when selecting a treatment regimen remains vital.[72] The IDSA published guidelines for developing an institutional programme to enhance antimicrobial stewardship,[73] designed to optimize clinical outcomes and minimize unintended consequences of antimicrobial use including the development of resistant organisms. Utilization of these guidelines as well as a comprehensive infection control programme may limit the emergence and transmission of antimicrobial-resistant bacteria.[73] Because resistance patterns vary with location and antibacterial therapy is initiated empirically, knowledge of resistance patterns in the local community may influence the selection of the initial antimicrobial for treatment of pyelonephritis in pregnant patients.

4.1 Antimicrobial-Resistant Uropathogens in Antepartum Pyelonephritis

It has been suggested that infections with antimicrobial-resistant organisms may increase the risk of complications and mortality.[74] This concept is of growing importance in the management of antepartum pyelonephritis as increasing rates of antimicrobial-resistant uropathogens are recognized.[21,32,43]

Greer et al.[43] investigated the outcomes in patients with acute antepartum pyelonephritis caused by ampicillin-resistant bacteria in a secondary analysis of a prospective cohort study. After being diagnosed with pyelonephritis, every patient was presumptively treated with ampicillin and gentamicin. Fifty-one percent of the organisms identified with sufficient cfu to receive antibacterial sensitivity testing were resistant to ampicillin, 92% of which were E. coli. The patients infected with ampicillin-resistant organisms were more likely to be older and multiparous, but there were no significant differences in complications between the groups. The rates of anaemia, renal dysfunction, respiratory insufficiency and preterm birth were comparable. There were also no significant differences in length of hospital stay, days of intravenous antibacterials required, admission to extended care unit or rate of hospital readmission. The results of this study appear to contradict the notion that antimicrobial-resistant organisms cause increased morbidity; however, an important consideration is that all of these women were concomitantly treated with gentamicin, and all but one of the organisms cultured was sensitive to this drug.[43]

There is possibly a difference between microbiological resistance and clinical resistance to an antibacterial medication, as noted by Wing et al.[32] in their study of the clinical utility of blood and urine cultures in the treatment of pyelonephritis in pregnancy. In their analysis, 94% of patients had an acceptable clinical response to the initial antibacterial therapy.[32] The antibacterial sensitivity of the uropathogen did not predict clinical cure. Susceptibility testing for antimicrobials established by the National Committee for Clinical Laboratory Standards is based on serum concentrations.[75] In vitro resistance may not adequately predict therapeutic failure because the concentration of most antimicrobial agents used for UTIs is higher in the urine than in serum.[75] For this reason, we advocate that changes in antibacterial therapy should be directed by clinical response rather than culture results alone. Surveillance of microbial drug resistance at a given institution may, however, facilitate selection of the appropriate first-line antibacterial therapy.

5. Rationale for Hospitalization of Pregnant Patients with Pyelonephritis

Over 20 years ago, the American College of Obstetricians and Gynecologists stated in a Technical Bulletin that hospitalization of pregnant women with pyelonephritis is necessary.[76] Theoretically, this approach will circumvent serious complications associated with pyelonephritis, including respiratory insufficiency, septic shock, preterm labour and delivery, and recurrences with the possibility of permanent renal damage. However, the recommendation to hospitalize all women with pyelonephritis was not based on evidence from clinical trials. The high cost of medical care and concerns for cost containment have provided motivation to study ambulatory methods of therapy for the pregnant patient. Because pyelonephritis is one of the most common reasons for hospitalization during pregnancy, attention has been directed toward developing a safe and effective approach to outpatient treatment of acute pyelonephritis. Based on the evidence presented in this review, this approach in pregnancy has only limited utility.

6. Ambulatory Treatment of Acute Pyelonephritis

6.1 Nonpregnant Patients

As observational studies[77–79] and randomized controlled trials[80,81] have demonstrated the safety of outpatient management of acute pyelonephritis in nonpregnant patients, the hospitalization rate for these patients has decreased from almost 100% to 7–30% in recent decades.[9,82,83] Most of the evidence for ambulatory treatment stems from data collected in emergency departments and investigations have reinforced the importance of careful patient selection for ambulatory therapy. Equally as important is the need for an initial period of observation for hydration and subsequent triage.[79] A meta-analysis of the existing literature on oral therapy for pyelonephritis[84] indicates that patients who do not have diabetes or sepsis and who are not immunocompromised, can tolerate oral intake, and do not have chronic illness can be treated for upper UTI with oral agents. From these data, a regimen for the outpatient treatment of pregnant women with acute pyelonephritis may be extrapolated; however, the application of ambulatory therapy for upper UTI in pregnancy appears to be quite limited.

6.2 Pregnant Patients

There are few trials on the ambulatory management of pyelonephritis in pregnancy, with most published in the last 2 decades.[54,55,57,58] Several investigators have attempted to simulate outpatient management such that all care during the hospitalization after initial therapy could have been accomplished at home.[54,55] In one trial, 90 pregnant women were treated with oral cephalexin every 6 hours or intravenous cephalothin every 6 hours until there was a decrease in CVAT and no fever for 48 hours.[54] More than 90% of the patients were treated successfully with either treatment modality. Similarly, once-daily intravenous ceftriaxone therapy was compared with multiple-dose intravenous cefazolin therapy in a randomized, double-blind clinical trial of 178 hospitalized patients.[55] Approximately 5% of patients in both groups failed to respond to initial therapy, and similar numbers in each group also demonstrated continued positive urine cultures at follow-up evaluation.

Brooks and Garite[57] reported a 12% treatment failure rate of outpatient treatment of 21 pregnant women with pyelonephritis in pregnancy at 26 weeks or earlier. Initial evaluation was for 2 hours in an obstetric emergency room with clinic follow-up for daily intramuscular injections of ceftriaxone until both fever and CVAT resolved, then a 10-day course of oral antibacterials. The authors estimated that 50% of pregnant patients reporting to an obstetric emergency room could be treated as outpatients. A later investigation with a different approach using home health nurse visits after initial intramuscular treatment in the emergency room followed by 10 days of oral cephalexin found 90% effectiveness in 120 pregnant women at less than 24 weeks’ gestation.[58] Ten percent of the outpatients were eventually hospitalized because of sepsis, abnormal laboratory tests, deviation from study protocol and recurrent pyelonephritis while on oral therapy. Overall, the authors estimated that 5% of patients with pyelonephritis at less than 24 weeks’ gestation who are considered candidates for outpatient therapy will not be able to be discharged.[58]

These results are sharply contrasted by a similar trial conducted by the same authors in women with pregnancies after 24 weeks gestational age.[85] Over 60%, or 154 of 246 women evaluated, were ineligible for outpatient management for reasons such as preterm labour, obvious sepsis or respiratory compromise, or pre-existing maternal medical condition or fetal malformation. After two doses of intramuscular ceftriaxone 24 hours apart, all women received oral cephalexin for 10 days. Inpatients received oral antibacterials until they were afebrile for 48 hours, and then were discharged from the hospital. Nearly 30% (13/46) of planned outpatients remained hospitalized after 24 hours because of clinical sepsis, bacteraemia, excessive leukocytosis (>20 000/mm3), preterm labour and other medical complications. Six outpatients and one inpatient failed to respond to initial therapy. Although there were no differences in clinical responses or birth outcomes of inpatients or outpatients after 24 weeks if they completed their assigned protocols, because of the inability to discharge 30% of the outpatients, the authors concluded that most women with acute pyelonephritis after 24 weeks’ gestation were not candidates for ambulatory therapy[85] and, therefore, this approach is quite limited in the third trimester.

There are no trials describing outpatient oral therapy for acute pyelonephritis in pregnancy.

6.3 Recommendations for Ambulatory Treatment of Pyelonephritis in Pregnancy

Cumulatively, the results of the existing trials are only encouraging for the ambulatory treatment of acute pyelonephritis in pregnancy in the first and early second trimesters of pregnancy. Outpatient treatment of acute pyelonephritis in pregnancy beyond 24 weeks appears to have limited utility, and it would appear most prudent to admit women for conventional treatment with intravenous hydration and antibacterials if their gestational ages exceed this threshold.

Clearly, careful selection of appropriate candidates for outpatient therapy of acute pyelonephritis in pregnancy is required. Patients should be less than 24 weeks pregnant, relatively healthy and able to comply with outpatient therapy. They should not have excessive fever (>38°C), severe nausea and vomiting, recurrent upper urinary tract disease, signs of sepsis including tachypnoea, tachycardia, rigors or hypotension, immunocompromise, significant medical conditions such as diabetes or previous renal disease, history of substance abuse, concurrent preterm labour or uncertain diagnosis. Liberalizing the criteria for outpatient therapy, in the absence of supporting data, may lead to serious detriment.

A period of observation is required, during which the patient should be hydrated and antibacterial therapy with intramuscular ceftriaxone initiated. Laboratory studies (complete blood cell count, serum chemistry panel including serum blood urea nitrogen and creatinine, and urine culture) should be obtained. Although urine culture may not aid in the immediate clinical management, approximately 6–10% of patients will not be successfully treated in an ambulatory setting and the results of the cultures may guide subsequent therapy. Surveillance of resistance rates to empirically chosen antibacterial therapies is mandatory. The period of observation should be long enough to review laboratory results and to assess the ability to tolerate oral intake, and the patient’s clinical response to therapy. If women with gestational ages beyond 24 weeks are to be managed on an ambulatory basis, even longer periods of observation with continuous fetal heart rate and uterine activity monitoring are required to ensure both maternal and fetal stability before discharge from the observation unit, although more recent data would indicate that the risk of preterm delivery associated with acute pyelonephritis is lower than previously estimated.[4]

Emphasis must be placed on close outpatient follow-up evaluation until both clinical and microbiological cure are obtained. Follow-up evaluation within 24 hours may take place in an office, clinic or emergency room. Alternatively, a home health nursing visit may be performed. At this evaluation, an additional dose of ceftriaxone should be given and an assessment made of the clinical condition. After the second dose, the antibacterial may be changed to oral cephalexin 500 mg four times a day (or a similar substitute) for 10 days, or daily intramuscular administration could be continued until all symptoms have resolved. The daily intramuscular administration regimen does not exceed 5 days. If the change to oral therapy is made, another outpatient clinical evaluation should be made within the 24 hours after the change. After completion of the daily ceftriaxone, oral treatment for 7–10 days is continued with cephalexin or with another antibacterial to which the causative organism is susceptible.

Clinic follow-up evaluation within 2 weeks after acute therapy should also be required. A urine culture should also be obtained as a ‘test of cure’. Throughout the remainder of the pregnancy, the patient should be followed for symptoms of recurrent infection and monthly urine cultures should be performed until delivery. Antimicrobial prophylaxis should be initiated in all patients for the remainder of the pregnancy and should be continued until 4–6 weeks postpartum.

7. Conclusions

The standard approach to the treatment of acute pyelonephritis in pregnancy is hospitalization and administration of intravenous hydration, antipyretics and parenteral antimicrobial therapy. There is insufficient data to recommend one antibacterial regimen at the current time, although consideration should be given to the antimicrobial resistance patterns at a given hospital. As more microorganisms become resistant to antimicrobial therapies, clinicians will be faced with additional treatment challenges. There is a small body of evidence to support the ambulatory treatment of pregnant women at less than 24 weeks gestational age with acute pyelonephritis provided these women are relatively healthy and do not manifest signs or symptoms of respiratory insufficiency or sepsis. Ambulatory treatment of acute pyelonephritis in pregnancy beyond 24 weeks appears to be limited in its applicability and is therefore not recommended.