Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Bacterial Infections of the Urinary Tract
In 1918, Helmholz recognized the cryptogenic nature and underdiagnosis of urinary tract infection (UTI) in the newborn. His observations still hold true today. There are no specific signs of UTI in a newborn; the clinical presentation can vary, ranging from fever with or without other signs of septicemia to minimal changes such as alteration in feeding habits or poor weight gain. The diagnosis of UTI in a neonate is made only by the examination and culture of a properly obtained specimen of urine.
The reported incidence, clinical manifestations, and prognosis of UTI in neonates have varied significantly. There are at least two reasons for discrepant results obtained in studies of UTI: (1) different criteria have been used to define UTI, and (2) infants with different characteristics have been studied. Before 1960 clean-voided specimens were used almost exclusively for examination and culture of urine. It is now clear that contamination is frequent when this method is used; Schlager and coworkers observed that 16 cultures of urine obtained by bag collection from 98 healthy newborns yielded greater than 10 4 colonies/mL of urine, with organisms that were found also on periurethral skin. The only reliable methods for obtaining urine for bacteriologic study are percutaneous aspiration and urethral catheterization of bladder urine.
Bacterial infections of the kidney and urinary tract in neonates are usually acquired at or after delivery. Fungal infections develop as nosocomial infections in infants with risk factors such as prematurity and use of intravascular catheters, parenteral alimentation, and broad-spectrum antibiotics, or after prolonged or intermittent catheterization of the urinary tract. Viral infections, including rubella, herpes simplex, and cytomegalovirus infections, are responsible for in utero infection, although the organisms can be excreted in the urine for months after birth. Common bacterial infections of the urinary tract are reviewed here. For information about infection and disease of the kidney and urinary tract caused by other microorganisms, the reader is referred to the chapters on toxoplasmosis, rubella, cytomegalovirus, herpes simplex, syphilis, mycoplasmas, Candida, group B streptococci, gonorrhea, staphylococcal infection, and neonatal diarrhea (Salmonella).
Epidemiology
The incidence of UTI in infants in the first month of life varies, ranging from 0.1% to 1% in older studies. Using a national database for 2003, the rate of hospitalization for UTI was 53.6 per 100,000 population younger than 1 month. The frequency may be 10% in infants of low birth weight and 12% to 25% in infants of very low birth weight evaluated for sepsis ( Table 9-1 ). In contrast to the increased incidence of bacteriuria among females in other age groups, infection of the urinary tract in the first 3 months of life is more frequent in males.
Table 9-1
Incidence of Urinary Tract Infection in Newborn Infants: Results of Eight Studies
| Sex | Birth Weight | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Male | Female | Male | Female | |||||||
| Study | Methods Used to Obtain Urine | No. Infected/ No. Studied (%) | No. Surveyed | No. Infected | No. Surveyed | No. Infected | No. Surveyed | No. Infected | No. Surveyed | No. Infected |
| Christchurch, NZ, 1968-1969 | CVS, SPA | 14/1460 (0.95) | 757 | 11 | 703 | 3 | NS | — | NS | — |
| Göteborg, 1960-1966 | CVS | 75 ∗ /57,000 (0.14) | NS | 54 | NS | 21 | NS | 11 | NS | 64 |
| New York, 1973 7 † | CVS, SPA | 12/1042 (1.2) | 493 | 7 | 549 | 5 | 206 | 6 | 836 | 6 |
| Leeds, 1967 | CVS, SPA | 8/600 (1.3) | 309 | 7 | 291 | 1 | NS | 0 | NS | 8 |
| Oklahoma City, 1974 | SPA | 10/102 (10) | NS | NS | NS | NS | 102 | 10 | — | — |
| Lausanne, 1978 † | CVS, SPA | 43/1762 ‡ (2.4) | 1006 | 26 | 756 | 7 | 634 § | 10 | 1028 щ | 33 |
| Göteborg, 1977-1980 ¶ | CVS, SPA | 26/198 (0.81) | 1502 | 23 | 1696 | 3 | — | — | — | — |
| U.S. Army, 1975-1984 # | SPA, Ca | 320/422,328 (0.08) | 217, 116 | 162 | 205, 212 | 158 | — | — | — | — |
Ca , Catheter; CVS , clean-voided specimen; NS , not stated; SPA , suprapubic aspiration (of bladder urine).
∗ Five male infants with infection and suspected or proven obstruction malformation of urinary tract not included.
† Date of published report; years of study not provided.
‡ Includes only infants <28 days of age who were admitted to neonatal intensive care unit.
§ Results reported for premature infants (<259 days of gestation)
щ Results reported for term infants (≥259 days of gestation).
¶ Results reported for infants 1 week to 2 months of age.
# Results reported for infants 1 week to 2 months of age who were hospitalized.
Infection of the urinary tract is usually sporadic, but clusters of cases, closely related in time, have been reported from nurseries in Cleveland and Baltimore. A nursery epidemic caused by Serratia marcescens was responsible for UTI and balanitis; the outbreak was caused by contamination of a solution applied to the umbilical cord.
Surveys of infants born in U.S. Army medical centers and subsequently hospitalized for UTI indicate that uncircumcised boys have substantially more UTIs than circumcised boys in the first month and in months 2 to 12 of life. In 1982, Ginsburg and McCracken observed that 95% of 62 infant boys with UTI were uncircumcised. A case-control study performed in 112 infant boys in whom suprapubic aspiration or bladder catheterization had been performed for investigation of acute illness showed that all infants with UTI were uncircumcised, compared with 32% of control subjects. Infection was associated with anatomic abnormalities in 26% of cases. In a meta-analysis of infants younger than 3 months who were evaluated for fever, rates of UTI were 7.5% for girls, 2.4% for circumcised boys, and 20.1% for uncircumcised boys. The records of more than 136,000 boys born in U.S. Army hospitals from 1980 to 1985 were reviewed through the first month of life to compare clinical courses in uncircumcised and circumcised boys. Of 35,929 uncircumcised boys, 88 (0.24%) had UTI, 33 had concomitant bacteremia, 3 had meningitis, 2 had renal failure, and 2 died. Complications followed 0.19% of 100,157 circumcisions (including 20 UTIs), and all were minor except for three episodes of hemorrhage leading to transfusion. Meta-analysis of nine published studies through 1992 yielded an overall 12-fold increased risk of infection in uncircumcised boys.
Studies published in 1996 and 1998 using case-control and cohort design support the earlier observational studies, with newer studies showing a three- to sevenfold increased risk for uncircumcised boys. An investigation by the Cochrane Neonatal Group found no randomized or quasi-randomized controlled trial of circumcision to prevent UTI. The American Academy of Pediatrics (AAP) has made a variety of statements regarding male circumcision after convening multidisciplinary workgroups and stakeholders to evaluate existing evidence. In 2012, the AAP concluded that the health benefits of newborn male circumcision outweigh the risks. Specific benefits considered were the prevention of UTI, penile cancer, and transmission of some sexually transmitted infections.
Ritual Jewish circumcision performed on the eighth day of life (when periurethral bacterial colonization has been established) seems to have attendant risk for UTI. An epidemiologic study in Israel revealed excessive UTIs in boys only in the postcircumcision period, from days 9 to 20 of life. A case-control study identified performance by a nonphysician (mohel) versus a physician as a risk factor for UTI (odds ratio, 4.34); the authors postulate the technique of hemostasis and duration of the shaft wrapping as responsible factors.
Microbiology
Escherichia coli continues to be responsible for most community-acquired UTI in infants younger than 3 months, accounting for 90% or more cases reported through the 1990s, with lower prevalence since, possibly related to widespread use of intrapartum chemoprophylaxis. Many different O serotypes of E. coli have been associated with UTI in the general population; however, UTI in neonates was associated with a limited number of O:K:H serotypes with P fimbriae, adhesive capacity, hemolysin production, and serum resistance. Serotypes of E. coli associated with diarrhea rarely cause UTI. Cultures of urine can be positive in infants with septicemia caused by group B streptococci, but primary infection of the urinary tract without septicemia is uncommon. Community-associated UTI in neonates caused by Staphylococcus aureus with or without bloodstream infection has been reported.
The incidence of neonatal UTI as a complication of intensive care has increased sharply in recent years; intensive care–associated UTI occurs in patients with and without urinary catheters. Mean age at occurrence was 42 days in one large study. Microbiology of neonatal nosocomial UTI is dramatically different from that observed in neonatal intensive care units in the 1970s ( Table 9-2 ), with E. coli supplanted by other genera of Enterobacteriaceae, Pseudomonas, Enterococcus, Candida, and coagulase-negative staphylococci. Multiple pathogens occasionally can be present as documented by bladder aspiration and bladder catheterization. S. aureus and E. coli have been responsible for localized suppurative disease of the urinary tract in neonates, including prostatitis, orchitis, and epididymitis. Other examples of focal disease in the urinary tract include orchitis caused by Pseudomonas aeruginosa and testicular abscess caused by Salmonella enteritidis . Blood cultures frequently are positive in affected infants. In one large study of neonatal intensive care unit (NICU)-associated UTIs, 13% of infants had concordant urinary and blood pathogens; in multivariate analysis, only organism type (i.e., gram-positive cocci, such as Staphylococcus species, group B streptococcus, and Candida species) predicted concordant positive blood culture.
Table 9-2
Pathogens Responsible for Urinary Tract Infections in Neonatal Intensive Care Units
| Frequency (%) of Isolations of Each Pathogen | ||||
|---|---|---|---|---|
| Organism | 1969-1978 ∗ | 1989-1992 † | 1991-2007 ‡ | 1997-2010 § |
| Escherichia coli | 75.3 | 10.5 | 24.8 | 18 |
| Klebsiella species | 13.4 | 10.5 | 25.5 | NR |
| Enterobacter species | 1.4 | 12.3 | 14.3 | NR |
| Other gram-negative species щ | NR | NR | NR | 32 |
| Enterococcus species | 2.1 | 14 | 4.9 | 13 |
| CoNS | 1.4 | 31.6 | 4.3 | 14 |
| Other gram-positive species | NR | NR | NR | 5 |
| Candida species | — | 12.3 | 15.5 | 15 |
| Other | 6.4 | 8.8 | 8.7 | NR |
CoNS , Coagulase-negative Staphylococcus; NR , not recorded.
∗ Data from 139 patients in nurseries and intensive care nurseries.
† Data from 50 patients in neonatal intensive care units.
‡ Data from 161 patients in neonatal intensive care units.
§ Data from 984 patients in 322 neonatal intensive care units; 97% were older than 3 days.
щ Included Enterobacter, Klebsiella, Serratia, Pseudomonas, and Proteus species.
Pathogenesis
In older children and adults, most UTIs are thought to occur by the ascending route after introduction of bacteria through the urethral meatus. Less frequently, bloodborne infection of the kidney occurs. In neonates, it is frequently difficult to know whether UTI was the cause or the result of bacteremia. The predominance of males with UTI among infants younger than 3 months contrasts with the predominance of females in all other age groups. This difference may reflect increased risk of UTI in young uncircumcised boys; increased prevalence of urinary and renal anomalies in boys; transient urodynamic dysfunction; vesicoureteral reflux (VUR), which predominantly affects male infants; and the occasional UTI that complicates circumcision. In addition, bacteremia is more frequent in male infants, and hematogenous invasion of the kidney can cause UTI.
Anatomic and Physiologic Abnormalities
Anatomic or physiologic abnormalities of the urinary tract play a role in the development and consequences of UTI in some infants. Of these, obstructive uropathy is the most important. Infection often is the first indication of an abnormality. UTI was the presenting sign in half of 40 infants younger than 2 months with anomalies of the kidneys or ureters reported in 1980. Congenital obstruction of the urinary tract was diagnosed in 5 of 80 children with UTI studied in Göteborg and in 2 of 60 children studied in Leeds ; important radiologic abnormalities of the urinary tract were identified in 10 of 46 infant boys and 3 of 13 infant girls younger than 3 months from 1972 to 1982 in Christchurch, New Zealand.
VUR is identified in many infants with UTI who are examined by radiologic techniques. VUR is frequently the result of infection but also can be a primary defect. VUR is not a prerequisite for upper tract infection (i.e., pyelonephritis); in two studies, less than half of children with pyelonephritis diagnosed by scintigraphy had VUR. Majd and coworkers found that 23 of 29 (79%) children hospitalized for UTI and found to have VUR had pyelonephritis diagnosed by scintigraphy; 39 of 65 (60%) children without VUR were also found to have pyelonephritis. VUR can be a congenital abnormality. Fetal ultrasonography showed that 30 of 107 infants with prenatally diagnosed urinary tract abnormalities had VUR, which was the only abnormality found postnatally in 10 infants. Gordon and colleagues observed that 16 of 25 infants with dilation of the fetal urinary tract had VUR, which was of grade 3 to 5 severity in 79%. In Austrian infants, 39 urinary tract abnormalities detected prenatally were compared with 46 urinary tract abnormalities found after first UTI. Obstructive lesions and multicystic dysplastic malformations of the kidneys accounted for 90% of all prenatally diagnosed malformations; VUR accounted for only 10%. By contrast, VUR accounted for 59% of abnormalities detected after the first UTI.
VUR detected prenatally has a male-to-female distribution of 6:1 (in contrast to VUR detected after UTI, when females predominate, ) may be determined developmentally by the site of the origin of the ureteral bud from the wolffian (mesonephric) duct, and in severe cases can be associated with congenital renal damage consisting of global parenchymal loss (so-called reflux nephropathy). Gunn and colleagues performed ultrasound examinations of 3228 fetuses; no renal tract abnormalities were detected before 28 weeks of gestation. Subsequently, 3856 fetuses were examined by ultrasonography after 28 weeks of gestation. Urinary tract anomalies were identified in 313 fetuses; 15 had major structural abnormalities, all of which were confirmed postnatally. In 298 (7.7%) of the fetuses, dilated renal pelvis with normal bladder was found; most of the cases resolved spontaneously, but 40 of the cases were confirmed postnatally to be due to serious abnormalities (usually obstruction or VUR). In one study, preterm infants with nosocomial UTI had a lower incidence of VUR than that noted in term infants with nosocomial UTI ; however, in another study comparing 250 neonates (mean gestation age, 39 weeks) with community-associated UTI with 51 neonates (mean gestation age, 36 weeks) with nosocomial UTI, the neonates with nosocomial infection were more likely to have VUR and abnormal renal ultrasound examination.
Isolated mild renal pyelectasis (i.e., <10 mm diameter of the collecting duct and without VUR) in fetuses is likely to be transient, unassociated with pathology or risk for UTI. In one study, 54% of such cases had resolved in the first postnatal month, and 85% of the cases of moderate or severe pelviectasis had resolved or improved over the first 2 years of life. Results of a long-term outcome study of 125 infants in the Netherlands with antenatal hydronephrosis suggested that a cutoff of less than 15 mm anterior-posterior diameter of the renal pelvis identified infants at low risk for UTI or surgical conditions and low incidence and benign course of VUR. In a 2012 Argentinean study of 196 hydronephrotic kidneys in 236 infants with mild (5-15–mm anterior-posterior pelvic diameter), isolated antenatal hydronephrosis was followed for a mean of 15 months postnatally without antibiotic prophylaxis or neonatal voiding cystourethrography (VCUG) study; 38% of patients had intrauterine total resolution, 42% had postnatal total resolution, and only 1% (3 patients) had progression. Of 98 infants with bilateral hydronephrosis, 9% had a UTI during follow-up none had VUR. Of 38 infants with unilateral hydronephrosis, 10% had a UTI; 2 had low-grade reflux. In a long-term study performed in the United Kingdom, of 425 infants with antenatally detected hydronephrosis, 284 had normal findings on neonatal ultrasound examination; the negative predictive value of normal ultrasound findings for subsequent UTI in the first year of life was 99%.
In a 5-year retrospective statewide, cohort analysis of 522 Washington state infants with birth-hospital discharge diagnosis of antenatal hydronephrosis (not quantified further) and 2610 control infants, hospitalization for UTI in the first year of life was 5% versus 1% (relative risk [RR] 11.8; 95% confidence interval [CI] 6.8 to 20.5). RR for UTI was higher, especially among girls with antenatal hydronephrosis (RR, 36.3; 95% CI, 10.6 to 124.0). In a 17-year study from a single institution in Seoul of 480 infants with antenatal and postnatal hydronephrosis without VUR, UTI developed in the first year of life in 39% of infants with obstructive uropathy, compared with 11% without obstructive uropathy ( P < .001). Higher grade of hydronephrosis and presence of hydroureteronephrosis were associated with higher incidence of UTI. Investigators from Montreal compared occurrence of UTIs and VUR in 148 children with low-grade (grade 1 and 2) and high-grade (grade 3 and 4) hydronephrosis detected prenatally. Children with anatomic abnormalities were excluded. UTIs were 3.6 times higher among the high-grade nephrosis group (11.1 infections/100 patient-years) compared with the low-grade hydronephrosis group (3.52 infections/100 patient-years); VUR was not an independent risk factor for UTI.
In a long-term study in Toronto of 260 infants with a prenatal diagnosis of hydronephrosis, 25 also had VUR (grade 3 or higher in 73%), received antibiotic prophylaxis, and did not have surgical correction during 4 years of follow-up. Breakthrough infection occurred in only four patients. Improvement was seen in most of the children with VUR, and there was no difference in renal growth in children who had resolved versus unresolved VUR or high-grade versus low-grade VUR. In a 2013 systematic review to evaluate the value of continuous antibiotic prophylaxis in reducing UTIs in this population, only 21 citations were included, of which 76% were of moderate or low quality. Pooled UTI rates in children younger than 2 years with low-grade hydronephrosis were similar regardless of receiving antibiotic prophylaxis (2.2%) or not (2.8%). In children with high-grade hydronephrosis, patients receiving prophylaxis had a significantly lower UTI rate (14.6%) compared with those not receiving prophylaxis (28.9%) ( P < .01). Postnatal management of prenatal hydronephrosis is controversial. Although experts favor evaluation of even mild cases by postnatal ultrasonography, and some favor VCUG, recent evidence supports a broad trend to use VCUG and antibiotic prophylaxis judiciously.
The outcome of interest for strategies to identify obstructive uropathy, hydronephrosis, VUR, and UTI is lessening renal damage. Except for obstructive uropathies, the relative impacts and benefits of special management of hydronephrosis and VUR are increasingly controversial. A National Institutes of Health multicenter, randomized, placebo-controlled, double-blind study has been designed to determine preventive effect of antimicrobial prophylaxis on recurrent UTI and renal scarring in children with primary VUR (i.e., not due to increased bladder pressure, as from a neurogenic bladder, outlet obstruction, or other vesicular anomalies). Chesney and colleagues commented on background knowledge and rationale for the study design. A few relevant points are as follows: (1) The percentage of patients with recurrent UTIs and VUR who develop renal scarring is small. (2) Many children, even those with high-grade VUR, do not develop scars. (3) Patients without VUR who have recurrent UTI can develop scars. (4) Older studies (with poorer outcomes) included patients with secondary VUR (i.e., high-pressure VUR) and genetic renal syndromes. (5) Trials of antimicrobial prophylaxis or surgery or both for VUR generally are underpowered and do not include a nonintervention arm. (6) In a Cochrane analysis using 10 trials involving 964 children, the authors could not conclude that identification and treatment of children with VUR conferred a long-term benefit.
Bacterial Virulence
Bacterial virulence factors are likely to play an important role in the pathogenesis of UTIs. Strains of E. coli causing UTI are a selected sample of the fecal flora. Pyelonephritic isolates belong to a restricted number of serotypes, are resistant to the bactericidal effect of serum, attach to and invade uroepithelial cells, and produce hemolysins. Pili on the bacterial cell surface that adhere to specific receptors on epithelial cells may play a role in development of UTI. Some of these features of pyelonephritic strains of E. coli have been shown in UTIs in newborns.
The increased rate of UTIs in uncircumcised boys is likely to be associated with periurethral bacterial flora. During the first 6 months of life, uncircumcised boys have significantly higher total urethral bacterial colony counts and more frequent isolation and higher colony counts of uropathogenic organisms, such as E. coli, Klebsiella / Enterobacter species, Proteus, and Pseudomonas . With increasing age, the foreskin is more easily retracted, and penile hygiene improves; by 12 months of age, the excessive periurethral flora and UTIs in uncircumcised boys almost disappear.
Natural defenses in the urinary tract include antibacterial properties of urine, antiadherence mechanisms, mechanical effects of urinary flow and micturition, presence of phagocytic cells, antibacterial properties of the urinary tract mucosa, and immune mechanisms. There is scant knowledge about these mechanisms in the newborn.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here