INTRODUCTION

Urinary tract infections (UTIs) are the most common bacterial infection in pregnancy¹ and a frequent non-obstetrical cause for hospitalization. As a result of studies carried out several decades ago, UTIs have long been regarded as a risk factor for preterm birth and low birth weight, in addition to being associated with increased maternal morbidity. On the basis of these historical data, repeated screening for asymptomatic bacteriuria has now become standard practice in many countries worldwide, resulting in the increased and often inappropriate use of antimicrobials. In an era of increasing drug-resistant bacterial strains, it is imperative that antibiotic use be justified by solid scientific evidence. Therefore, the rationale and benefit of a screen-and-treat approach should be reviewed. Also, the risks associated with antibiotic use on neonatal outcomes and the acceptability of the intervention to expectant mothers should be taken into consideration.

The aims of the following chapter are to review the evidence published to date to assess international recommendations for clinical practice, and to highlight areas for future research.

DEFINITIONS

Note that following definitions and their criteria are generally based on expert opinion and clinical consensus.

Asymptomatic bacteriuria (ASB): ≥10⁵ colony forming units (cfu)/ml of a single pathogen or of two uropathogens in a voided urine specimen without signs or symptoms attributable to a UTI.² A second urine sample should ideally be obtained within two weeks to confirm bacteriuria, which will be absent in 10–60% of second samples.² However, for practical purposes, standard practice and most international guidelines suggest collection of a single specimen.^3,4,5,6,7,8 If three or more uropathogens are identified, the sample is described as likely contaminated.³ Dipstick testing to screen for ASB is insufficiently sensitive and should not be used as a screening method as it may lead to unnecessary treatment.^8,9

Symptomatic lower urinary tract infection or cystitis: This is a clinical diagnosis with the presence of symptoms including: dysuria, urgency, frequency, hematuria, suprapubic discomfort, and uterine contractions, and of bacteriuria 10³–10⁵ cfu/ml.⁵

Pyelonephritis is characterized by fever, chills, flank pain, nausea, vomiting, and costovertebral angle tenderness, together with bacteriuria and pyuria (>10 leucocytes per high-power field).⁵ Increased white cell count and raised inflammatory markers are also a feature.

Relapse is a recurrent UTI after therapy resulting from persistence of the pre-therapy isolate in the urinary tract.¹⁰

Reinfection is recurrent UTI with an organism originating from outside of the urinary tract–either a new bacterial strain or a strain previously isolated that has persisted in the colonizing flora of the gut or vagina.¹⁰ In clinical practice, it is often difficult to distinguish between relapse and reinfection.

Pyuria is the presence of increased numbers of polymorphonuclear leucocytes in urine and is evidence of an inflammatory response in the urinary tract.¹⁰ In pregnancy, urine is frequently contaminated by increased vaginal secretions so interpretation of the presence of urinary leucocytes in pregnancy should be made with caution.

EPIDEMIOLOGY

The prevalence of asymptomatic bacteriuria during pregnancy has been estimated at 2–15%,¹¹ although some studies from low-income countries have demonstrated prevalences exceeding 20%.³ Cystitis has been less well studied amongst the pregnant population, but its incidence is estimated at 1–2%.⁵ The incidence of pyelonephritis is 0.7–2.0% in pregnancy, which is significantly higher than in the non-pregnant population.^12,13

Risk factors for UTI in pregnancy are:^3,4,13,14

• Young age (<20 years)
• Primiparity
• Smoking
• Low socio-economic status
• Women of Afro-Caribbean or Asian descent
• Medical comorbidities: polycystic kidneys, congenital renal anomalies, sickle cell disease, history of recurrent UTI
• A history of pyelonephritis within the year preceding pregnancy has been associated with an eightfold risk of pyelonephritis during pregnancy (odds ratio (OR) 8.45 [95% CI 7.40–9.65]).¹³

Diabetes is often listed as a risk factor for UTI, but a recent cohort study from the Netherlands did not find an increased prevalence of asymptomatic bacteriuria or urinary tract infection in women with diabetes mellitus or gestational diabetes as compared to non-diabetic patients: the overall prevalence of ASB was 4.7%.¹⁵ However, high fasting glucose levels (>100 mg/dl) and proteinuria ≥2 on dipstick have been associated with UTI.¹³ Frequency of voiding has been found to be protective.³

PATHOPHYSIOLOGY

The increased risk of UTI in pregnancy relates to physiological and anatomical changes in the host, as well as to specific microbial properties.

Increased levels of progesterone during pregnancy result in smooth muscle relaxation with reduced ureteric peristalsis and an increased bladder capacity. The kidney increases in size, the collecting system is physiologically dilated, and renal function increases by 30–50%. Mild physiological hydronephrosis may be present from 7 weeks of gestation.⁵ Furthermore, compression by the gravid uterus may predispose to reflux of urine and the presence of glycosuria increases the risk of UTI.¹¹

Micro-organisms found among pregnant women do not differ from those among non-pregnant women. The main causative organisms are the following:¹⁴

• Escherichia coli, 82%
• Streptococcus spp. (group B), 21.4%
• Other Enterobacteriaceae such as Klebsiella pneumoniae (7.6%), Proteus mirabilis (4.9%), Enterobacter and Citrobacter
• Staphylococcus spp., 6.5%
• Enterococcus spp., 5.7%
• Pseudomonas aeruginosa
• Others: Gardnerella vaginalis, Ureaplasma urealyticum

Uropathogens generally originate from the rectal flora and ascend to the bladder, or they colonize the vaginal introitus, periurethral area and urethra. Uropathogenic Gram-negative bacteria possess specific virulence factors that enhance both colonization and invasion of the urinary tract, such as the P-fimbriae of certain strains of E. coli that allow for adherence to uroepithelial cells.³

Recent advances have discovered that urine is not in fact sterile but possesses a urinary microbiota, or micro-organisms that reside within the bladder. The involvement of urinary microbiota in UTI and, specifically, in pregnancy remains to be understood.¹⁶

COMPLICATIONS OF ASB

ASB has been associated with pyelonephritis, low fetal birth weight, and preterm labor/delivery, but more recent evidence points to weaker associations than initially suggested by older research.

Pyelonephritis

Historical studies place the risk of pyelonephritis in untreated bacteriuria as high as 36%, with most studies ranging between 2.2 and 36%.³ However, these include studies published over three decades ago, with different pathogens, different socioeconomic backgrounds, and different populations. Thus, their relevance in the current healthcare context has been questioned.¹⁷

A more recent study carried out in the Netherlands in 2015, where women are not routinely screened for bacteriuria, found a significant association between untreated ASB and pyelonephritis, but overall the risk remained low at 2.4%. The risk of pyelonephritis was 0.6% in women without bacteriuria.¹⁷ The presence of ASB is therefore a risk factor for pyelonephritis, but the strength of this association appears to be lower than older studies suggest.⁴

Obstetric and neonatal outcomes of ASB

ASB has also been associated with low fetal birth weight and preterm labor, but again studies demonstrate contradictory findings and the definition of preterm birth varies between studies. Importantly, many early studies reporting preterm birth were conducted before the routine use of sonography was introduced to standardize gestational dating. In 2015, Kazemier et al. in the Netherlands did not demonstrate a significant difference in preterm birth defined as birth <34 weeks between untreated ASB and ASB-negative women (OR 0.7 [95% CI 0.2–2.8]).¹⁷ On the other hand, a recent Danish cohort concluded that women who delivered between 28 and 34 weeks were more likely to have had ASB compared to women delivering after 35 weeks (OR 2.05 [95% CI 1.36–3.09], P = 0.001).¹⁸

Two systematic reviews have explored the association between ASB and obstetric outcomes. The Canadian Task Force on Preventive Healthcare compared the relative risk of low birth weight – defined as birth weight ≤2500 g – between ASB-positive treated women and ASB-positive untreated women, with a relative risk of 0.63 [95% CI 0.45–0.90], but the overall quality of the evidence was deemed to be low. Similarly, the relative risk of preterm birth, defined as birth <37 weeks, between ASB-positive treated women and ASB-positive untreated women was 0.57 [95% CI 0.21–1.56], yet the overall quality of evidence was very low.⁴ The Cochrane review concludes that there are too few studies exploring birth weight and gestational age to draw any conclusions.³

Kazemier et al. compared neonatal outcomes between women who were ASB positive and treated with nitrofurantoin (n = 40) and women who were ASB positive and untreated or received placebo (n = 208). The outcomes these authors studied were gestational age at birth, birthweight, perinatal death, admission to neonatal intensive care, neonatal sepsis, and congenital anomalies. They did not find any significant difference between the two groups; however, their overall population size, particularly the cohort who received antibiotic therapy, was too small (n = 40) to draw any conclusions regarding the effect of antimicrobial therapy on neonatal outcomes.¹⁷

The Canadian Task Force on Preventive Healthcare did not find any difference in neonatal sepsis between ASB-positive treated women and ASB-positive untreated women [OR 0.22 (0.01–4.54)] or perinatal morbidity [OR 0.96 (0.27–3.39)].⁴

On the basis of these studies and the recent Dutch cohort, the evidence pointing to an association between ASB and poor obstetric or neonatal outcomes is weak and requires further research.

In addition, there is a paucity of research regarding the effects of acute symptomatic lower UTI in pregnancy: the majority of current research focuses on ASB, and conclusions appear to be extrapolated to include cystitis.

COMPLICATIONS OF PYELONEPHRITIS

Pyelonephritis is one of the most common non-obstetrical causes of antenatal hospitalization and is more frequent in the second and third trimesters of pregnancy, although 1 in 5 cases has been found to occur in the first trimester.^14,19 Maternal complications of pyelonephritis include anemia, septicemia, transient renal dysfunction, respiratory distress and renal abscess.^14,19

Regarding, obstetric outcomes, Wing et al. found a modest increase in preterm births and small-for-gestational-age babies amongst women with acute pyelonephritis, whereas Hill et al. did not find any significant difference.^14,19

Neonatal outcomes of mothers with acute pyelonephritis during pregnancy have not been clearly studied. A recent systematic review identified intrauterine death in 6% of cases and admission to neonatal intensive care in 9% of cases of maternal pyelonephritis.¹² However, this systematic review includes cases from 1908, with 12% of studies before 1945 when access to healthcare, diagnostic methods, therapeutic options, and standards of care was not comparable with current healthcare provision.

SCREEN-AND-TREAT FOR ASYMPTOMATIC BACTERIURIA

In a Cochrane review, Smaill and Vasquez (2019) concluded that the treatment of ASB may result in reduced incidence of pyelonephritis (RR 0.24 [95% CI 0.13–0.41]),³ and that a reduction in the incidence of pyelonephritis since the introduction of routine screening has been reported.²⁰ Antibiotic therapy for ASB may reduce the incidence of preterm birth, defined as gestational age of <37 weeks (RR 0.34 [95% CI 0.13–0.88), and of low birth weight, defined as birthweight <2500 g (average RR 0.64 [95% CI 0.45–0.93).³ However, in both cases, the strength of the evidence is rated as either low or very low. Therefore, there are insufficient data to determine reliably whether the treatment of ASB has any effect on neonatal outcomes.³

Numerous countries have adopted repeat screening for ASB in pregnancy, when there is little evidence to suggest that repeat screening is beneficial,³ and in fact two studies from the Netherlands found the overall benefit to be so low that is does not justify a screening program.^15,17

On the basis of these studies, there is no evidence to justify repeated screening in pregnancy, and the benefit of a screen-and treat-approach in early pregnancy is based on a modest benefit demonstrated by only weak evidence. The adverse consequences of this approach are often not considered.

Kazemier et al. question the health gain from a screen-and-treat approach, particularly in a low-risk population that can be treated promptly if symptoms develop. In light of the ORACLE II trial, which demonstrated a small increase in functional impairment and cerebral palsy at 7 years of age in children whose mothers had received antibiotics for preterm labor during pregnancy, the safety of antibiotic use, with its potential neonatal consequences, is questioned.¹⁷

Patient preferences, attitudes and their decision-making processes with regard to antibiotic use in the absence of symptoms need to be considered,^4,17,20, as it has been shown that almost half of pregnant women receiving antibiotics express the view that antibiotics should not be used during pregnancy and question their safety.²⁰

It is also important to highlight that inappropriate antimicrobial use drives antimicrobial resistance in the general population. Nicolle et al. argue that, in the 14 years between the publication of the original guideline in 2005 and its update in 2019, antimicrobial resistance in bacteria isolated from UTI has evolved substantially, and extended-spectrum ß-lactamase– and carbapenemase–producing Enterobacteriaceae are isolated frequently from UTI.² The inappropriate treatment of ASB has been identified as an important contributor to antimicrobial resistance.²

Thus, national guidelines are moving away from repeated screening and advocating a single urine culture in early pregnancy, between 12 and 16 weeks' gestation, justified by its modest benefit.^4,6,7

THERAPEUTIC OPTIONS

As a general recommendation, antibiotic therapy should be adjusted to the local prevalence of isolated micro-organisms and their susceptibility patterns. The resistance patterns of different strains, especially Enterobacteriaceae, vary among regions and countries, so specific antimicrobial treatment may not be suitable for all areas, especially in a world where antimicrobial resistance is rising.

The stage of pregnancy, the availability, the cost, and local practice patterns have to be taken in account, as well as individual considerations (allergy, glucose-6-phosphate dehydrogenase [G6PD]-deficiency, etc.). Antibiotics in pregnancy are not without risk, as demonstrated by the ORACLE II study discussed above. Apart from direct medication-related side effects, there is also the issue of the effect of antibiotics on the diversity of infant microbiota, with loss of diversity associated with increased body weight, insulin resistance, dyslipidemia, and a host of other disorders later in life.²¹

Antimicrobial therapy for asymptomatic bacteriuria

The recommended regimens are beta-lactams (semi-synthetic penicillins—for example, ampicillin, co-amoxicillin, or oral cephalosporin), nitrofurantoin, or fosfomycin. Several trials have shown their effectiveness in treating asymptomatic bacteriuria, with no statistically significant difference in terms of safety and effectiveness among different regimens. Nitrofurantoin should be avoided at term and per-partum because of a potential risk of neonatal haemolysis.^{22,23,24,25,26,27}

A 2015 Cochrane review assessed duration of treatment for ASB during pregnancy, comparing four interventions: a single-dose antibiotic regimen, short-course regimens defined as 4–7 days, a long course defined as 14 days, or continuous. Outcomes were maternal test of cure (negative urine culture), recurrent ASB, reinfection, pyelonephritis, preterm birth, and low birth weight. The review concluded with high-quality evidence that there are better cure rates with short-course treatments (4–7 days), even when assessing the outcome of low birth weight. No difference was noted in terms of preterm birth or incidence of pyelonephritis.²⁸

There is insufficient evidence to repeat urine cultures after antibiotic treatment for an initial ASB episode, nor following retreatment. Data are lacking to support antimicrobial prophylaxis or suppressive treatment to prevent recurrence of ASB for the remainder of the pregnancy.²

Antimicrobial therapy for symptomatic lower urinary tract infection in pregnancy

Antimicrobial therapy is usually introduced first empirically after urinary complaints. Following a Cochrane review in 2011, no particular regimen was associated with a better outcome in terms of cure rates of cystitis.²⁹ However data are too scarce in terms of equivalence or superiority studies to draw a reliable conclusion. The most important aspect is that empirical treatment should be reviewed after culture results. Suggested options are beta-lactam antibiotics, nitrofurantoin, and fosfomycin. The two last regimens are to be considered strongly in women known to be colonized with extended-spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae. The optimal duration of treatment is unclear, though a short course should be favored (see section above).

Antimicrobial therapy for pyelonephritis

There are limited trial data available for the management of antenatal pyelonephritis. It is usually recommended to admit the patient initially for work-up and close surveillance. Urine culture before antibiotic initiation is key for management. Blood cultures do not offer any additional information compared to pretreatment urine culture and therefore are not considered mandatory.³⁰

Treatment usually consists of parenteral antibiotics, such as the combination of ampicillin or amoxicillin with an aminoglycoside or a third-generation cephalosporin. Carbapenems or monobactam (e.g., aztreonam) should be favored for patients with a documented past history of ESBL infection or colonization. Outpatient oral treatment after 24–48 hours of initial observation is possible.

Antimicrobials should be adapted to the urine culture results. Oral options are beta-lactams and trimethoprim-sulfamethoxazole if in the second trimester. Nitrofurantoin and fosfomycin are not recommended for pyelonephritis treatment, given their low serum levels and poor tissue penetration. Further work-ups such as imaging to search for perinephric abscess or urinary tract obstruction are warranted in the absence of a prompt clinical response.

For patients presenting with recurrent pyelonephritis, suppressive treatment can be introduced although data are limited and further research is needed.³¹

PROPHYLACTIC MEASURES

Recurrent urinary tract infections (RUTIs)

Varying definitions of RUTI exist, but RUTI in pregnancy has been defined as a UTI in a pregnant woman with a history of one or more UTIs before or during pregnancy.^3,32

Antibiotic prophylaxis during pregnancy for women with RUTI is not routinely recommended.

A Cochrane Review in 2015 showed, with low-quality evidence, that suppressive low-dose oral nitrofurantoin in combination with increased clinical follow-up and surveillance was not superior to increased surveillance alone in the prevention of recurrence of UTI. There was no difference observed in the incidence of preterm birth in this group.^32,33

Non-pharmacological measures

Alternate measures such as cranberry juice, supplementary ascorbic acid, or behavioral modifications have been described, but all these approaches rely on poor evidence showing little clinical effect.³⁴

PRACTICE RECOMMENDATIONS