THE CIRCUMCISION REFERENCE LIBRARY
From the Department of Pediatrics, The University of Texas Health Science Center at Dallas, Southwestern Medical School, Dallas
Abstract. The clinical and laboratory features of urinary tract infections in 100 infants aged 5 days to 8 months are presented. Of the patients in the first three months of life 75% were boys, and of infants aged 3 to 8 months only 11% were boys; 95% of the infants were uncircumcised. Sepsis was documented in 31% of neonates, 21% of infants aged 1 to 2 months, 14% of those aged 2 to 3 months, and 5.5% of infants >3 months of age. Roentgenographic abnormalities of the urinary system were found in 45% of female and 7% of male infants. All infants responded promptly to antimicrobial therapy. The possible factors related to the predominance of male infants with urinary tract infections are discussed. Pediatrics 69:409-412, 1982; urinary tract infection, young infants, sepsis, circumcision.
There have been many published reports on urinary tract infections (UTIs) in newborn infants.1-5 These have involved infants who were either in the nursery1-3,5 or in a neonatal intensive care unit4; many urinary tract infections were detected by screening programs.2,3,5 Infections were more frequent in male infants (2.3 to 5:1).1-4 Roentgenographic abnormalities of the urinary tract were detected in more male than female infants and the case-fatality rate was 10% in two studies.1-2 This experience in neonates is markedly different from that in older infants and children in whom infection occurs almost exclusively in females, sepsis is infrequent, and the clinical illness is usually uncomplicated. To our knowledge there is no information on when in infancy this occurs.
We present data on 100 young infants who were hospitalized with their first known UTI. All infants had been considered healthy at the time of discharge from the nursery and were home for periods ranging from two days to eight months before onset of illness. The clinical and laboratory features of their illness are presented in order to emphasize several points: (1) the male predominance occurs only during the first three months of life; (2) positive blood cultures were inversely related to age and were uncommon after 9 weeks of age; (3) abnormal roentgenogram of the urinary system were found in almost half of female infants with UTI and rarely in male infants; and (4) all infants responded promptly to antibiotic therapy.
During a 59-month period from March 1976 through February 1981, 100 infants with acute urinary tract infections were admitted to hospital from the acute care clinic of Children’s Medical Center or from the emergency room of Parkland Memorial Hospital, Dallas. The age of the patients ranged from 5 days to 8 months (mean 2.1 months). There were 62 boys and 38 girls (Figure). Male infants accounted for 75% of patients in the first three months of life compared with 11% of patients who were 3 to 8 months of age. Of 41 infants who were 30 days of age or younger, 33 (81%) were boys.
The duration of illness before hospital admission was 48 hours or less in the majority of patients. Fever (63% of patients) was the most common symptom that prompted the parents to seek medical attention. Of the infants 55% were described as irritable and 38% had refused one or more feedings. Vomiting and diarrhea were reported in 36% and 31% of patients respectively.
Although many of the infants were described as “ill-appearing” at the time of the initial physical examination, there were few objective abnormal findings: 67 infants had a temperature of equal to or more than 38 C and 38 infants had temperatures of equal to or greater than 39 C. Abdominal distension and jaundice were uncommon, being present in only 8% and 7 percent of patients respectively. Three of 62 male infants (5%) were circumcised and two had primary hypospadias.
Urine was collected from all patients by suprapubic aspiration catheter. Uncentrifuged urine contained 10 or more WBCs [high power field (HPF)] in 62 (70%) of 89 speciments examined. Bacteria were visualized in 81 stained urine samples including 22 of 27 that contained less than 10 WBCs/HPF and 17 (81%) of 21 that had less than 5 WBF/HPF. Urine cultures from 96 patients grew equal to or greater100,000 colony-forming units/ml of a single bacteria species. The remaining four patients had urinary colony counts of from 40,000 to 80,000/ml; these speciments were obtained by suprapubic bladder aspiration. Escherichia coli was identified in 88 patients. The pathogens from the other 12 infants were Klebsiella pneumoniae (five infants), nonenterococcal group D Streptococcus (two infants), Pseudomonas aeruginosa (one infant), Proteus mirabilis (one infant), Staphylococcus aureus (one infant), Entertobacter cloacae (one infant), and enterococcus (one infant).
Blood cultures were obtained from 91 patients; 20 cultures (21.9%) contained a pathogen identical with that isolated from the urine. Positive blood cultures were found in 11 (31%) of 35 infants who were less than 30 days of age, in five (21%) of 24 infants aged between 1 and 2 months, in two (14%) of 14 infants age 2 to 3 months, and in one (5.5%) of 18 infants age 3 months or older. The age of one infant with a positive blood culture was unknown. There was a significant linear trend detected, suggesting a significant decrease in occurrence of sepsis with increasing age. (χ2 = 5.24, P = .022 by χ2 trend analysis6). Sepsis occurred in 13 boys and seven girls; of the infants age 3 months or less, 12 (63%) were boys. This percentage is similar to that found in male infants with UTI in this age group. There was no correlation between fever, the initial WBC count, or presence of roentgenographic abnormalities and a positive blood culture. A lumbar puncture was performed in 88 patients and all CSF cultures were sterile.
Roentgenograms were obtained during hospitalization in 86 children. In 18 patients 26 abnormalites were identified, including unilateral reflex (grades 2 through 4), six patients; bilateral reflex (grades 2 though 4), 12 patients; duplication, five patients; hydronephrosis, three patients (one with duplication ureterocele and two with ureteropelvic obstruction). Abnormalities were more common in girls (14/31, 45%) that in boys (4/55, 7%). The difference between these two groups is statistically significant (P > .01 by χ2 analysis). In three children the kidneys were not visualized by intravenous pyelography but were considered normal by ultrasound technique. Reflux (grades 2 to 4) was found in 18 infants. Twelve infants had bilateral reflux; in nine the reflux was grade 3 or 4.
Initial antimicrobial therapy consisted of ampicillin or amoxicillin and gentamicin in 73 patients, gentamicin alone in 25 patients, and ampicilin or amoxicillin in one patient each. Most patients became afebrile in 12 hours of receiving antimicrobial therapy, and repeat urine cultures at 24 to 72 hours after therapy were sterile in 94 of 97 infants. Culture from two infants contained nonpathogens and the third contained 4,000 colony forming units/ml of E. coli. Infants with bacteremia were usually treated with parenteral antibiotics for seven to ten days whereas in those with uncomplicated infections medication was changed to oral antibiotic therapy after clinical improvement and verification of sterile blood and CSP cultures. Fifty-eight patients, many of whom were discharged by the fifth hospital day, were treated with orally administered antibiotics at the time of discharge. Ampicillin or amoxicillin were the drugs most frequently used (32 patients) 22 infants received a sulfonamide alone or in combination with another drug (erthromycin in three infants and trimethodprim in 7 infants). Of these latter infants, 12 (55%) were less than 2 months of age.
The mean duration of hospitalization was 7.1 days (range 3 to 19 days). Two patients who had ureteropelvic obstruction underwent surgery to correct the obstruction during their hospitalization. There were no deaths.
In the present study male infants accounted for the majority of urinary tract infections in the first three months of life, but female infants predominated thereafter. This unique age susceptibility pattern cannot be explained by the number or type of patients seen in our outpatient facilities nor by changes in the method of obtaining or processing urine samples. Because only 5% of male infants in this study were circumcised, it is tempting to speculate that the uncircumcised male has an increased susceptibility to UTI. Contamination of the urine specimen during collection could not have played a role because all specimens were obtained by suprapubic bladder aspiration. To our knowledge there is no information on the incidence of UTI in circumcised as compared with noncircumcised infants. Although there are data in adult females that suggest that perineal bacteria invade the bladder by ascending the short urethra.7,8 it is generally believed that this does not occur in adult males, presumably because of the length of the urethra. Although blatant balanitis or phimosis was not described by the physicians who treated these infants, perineal hygiene was inadequate in many patients. It is conceivable that the urethra of male infants is unable to prevent ascending infection during the first months of life and that perineal bacteria are present in large enough numbers at the meatus in uncircumcised male infants to allow attachment and ascension to the bladder.
[CIRP comment: Parkland Hospital is a public hospital that serves a large urban population. At the time that this study was carried out, Parkland Hospital had a policy of refusing to do circumcisions even if requested. The hospital's client population of young male infants, therefore, would have been almost completely non-circumcised. Ginsburg and McCracken's observation that only 5 percent of male infants were noncircumcised, therefore, should not be surprising and may not indicate that noncircumcised infants get more urinary tract infections.]
Sepsis was documented in 20 of 91 infants from whom blood cultures were obtained. Approximately 85% of these positive blood cultures were infants aged more than 2 months of age and only one of 18 infants aged more than 3 months had sepsis. Although the proclivity for bloodstream infection in the very young is well known, the mechanisms accounting for this inverse relationship between sepsis and age of infants with urinary tract infection are unknown. It is possible that maturation of the immune system and development of specific local and systemic antibody to E coli antigens require several months before protection against blood stream invasion is accomplished. The data do emphasize the importance of obtaining urine and blood cultures on all young febrile infants with suspected UTI. The absence of associated meningitis in these infants, even those with bacteremia, suggests that lumbar punctures are unnecessary in infants with suspected urinary tract infections. However, it must be emphasized that two thirds of these infants were suspected of having sepsis at the time of admission to hospital. There is no other certain means of diagnosing meningitis other than by performing lumbar punctures in all febrile and ill-appearing infants, regardless of the findings on urinalysis.
As in previous reports,9,10 absence of significant pyuria did not rule out urinary tract infection. Less than 10 and 5 WBCs/HPF were found in 27% and 21% of urine samples, respectively. Bacteria were present on stained smears of most of these speciments and this proved the most reliable initial indicator of infection.
Roentgenographic abnormalities were present in 18% of patients, 78% of whom were girls. Previous reports of infants and children with UTI have demonstrated that boys are more likely than girls to have roentgenographic abnormalities, particularly obstructive lesions.1,2,11-13 The incidence of radiographic abnormalities in the girls in this study was relatively large (45%) and significantly greater than that (7%) in boys. One patient with ureteropelvic obstruction required immediate surgical intervention and three others with grade 4 reflux and hydronephrosis had surgical procedures performed on an elective basis. The majority of the other female infants had abnormalities that placed them at risk of recurrent infections and renal scarring.
Because this study was retrospective it is not possible to compare the efficacy of the two antibiotic regimens used most commonly on admission to the hospital. Suspected sepsis was the admitting diagnosis in 64 (88%) of the 73 patients who received ampicillin and gentamicin. By contrast, all 27 patients who received gentamicin alone were hospitalized with a diagnosis of uncomplicated urinary tract infection. Regardless of the initial antibiotic regimen, all patients had a satisfactory clinical response to therapy. Considering the etiologic agents of UTI, either regimen should be effective as initial antimicrobial therapy in young infants with UTI.
Dr. Joan Reisch assisted in the statistical analysis of data.
Return to CIRP library