Objective. To assess the effect of newborn circumcision on the incidence and medical costs of urinary tract infection (UTI) during the first year of life for patients in a large health maintenance organization.
Setting. Kaiser Permanente Medical Care Program of Northern California (KPNC).
Patients. The population consisted of members of KPNC. The study group consisted of a cohort of 28 812 infants delivered during 1996 at KPNC hospitals; of the 14 893 male infants in the group, 9668 (64.9%) were circumcised. A second cohort of 20 587 infants born in 1997 and monitored for 12 months was analyzed to determine incidence rates.
Design. Retrospective study of all infants consecutively delivered at 12 facilities.
Outcome Measures. Diagnosis of UTI was determined from the KPNC computerized database using the International Classification of Diseases, Ninth Revision code for inpatients and KPNC Outpatient Summary Clinical Record codes for outpatients. A sample of 52 patient charts was reviewed to confirm theInternational Classification of Diseases, Ninth Revisionand KPNC Outpatient Summary Clinical Record codes and provide additional data.
Results. Infants <1 year old who were born in 1996 had 446 UTIs (292 in females; 154 in males); 132 (86%) of the UTIs in males occurred in uncircumcised boys. The mean total cost of managing UTI was 2 times as high in males ($1111) as in females ($542). This higher total cost reflected the higher rate of hospital admission in uncircumcised males with UTIs (27.3%) compared with females (7.5%); mean age at hospitalization for UTI was 2.5 months old for uncircumcised boys and 6.5 months old for girls. In 1996, total cost of managing UTI in uncircumcised males ($155 628) was 10 times higher than for circumcised males ($15 466) despite the fact that uncircumcised males made up only 35.1% of the male patient base in 1996, reflecting the more frequent occurrence of UTI in uncircumcised males (132 episodes) than in circumcised males (22 episodes), and the larger number of hospital admissions in uncircumcised males (38) than in circumcised males (4). The incidence of UTI in the first year of life was 1:47 (2.15%) in uncircumcised males, 1:455 (.22%) in circumcised males, and 1:49 (2.05%) in females. The odds ratio of UTI in uncircumcised:circumcised males was 9.1:1.
Conclusions. Newborn circumcision results in a 9.1-fold decrease in incidence of UTI during the first year of life as well as markedly lower UTI-related medical costs and rate of hospital admissions. Newborn circumcision during the first year of life is, thus, a valuable preventive health measure, particularly in the first 3 months of life, when uncircumcised males are most likely to be hospitalized with severe UTI.
- UTI =
- urinary tract infection •
- KPNC =
- Kaiser Permanente Medical Care Program of Northern California •
- ICD-9 =
- International Classification of Diseases, Ninth Revision •
- OSCR =
- Outpatient Summary Clinical Record •
- CMIS =
- Cost Management Information System
Increased prevalence of urinary tract infection (UTI) in uncircumcised boys during the first year of life was first recognized in the 1980s.1–4 Since that time, many studies have confirmed the protective effect of newborn circumcision against UTI at all ages5 ,6 but particularly in infancy.1–4 ,7 The current study assessed the effect of newborn circumcision on the incidence, morbidity, and medical costs of UTI during the first year of life in a large health maintenance organization.
The population consisted of members of the Kaiser Permanente Medical Care Program of Northern California (KPNC). Current membership is 2.8 million people who receive their care from 32 facilities. Infants are delivered at 12 of these facilities. The study group consisted of 28 812 infants (including 14 893 male infants) delivered during 1996 at KPNC hospitals. The number of male infants who had newborn circumcision was determined using International Classification of Diseases, Ninth Revision (ICD-9) codes for infants circumcised during their neonatal hospital stay, and the KPNC Outpatient Summary Clinical Record (OSCR) codes for outpatients. Newborn circumcision by standard methods was performed before discharge, usually on the second or third day of life, after informed consent was obtained.
The diagnosis of UTI was determined from the KPNC computerized database using the ICD-9 code for inpatients and the OSCR codes for outpatients. Of the 154 UTIs observed in infant boys, a sample of 52 patient charts was reviewed to confirm the ICD-9 OSCR code and to collect additional data, including method of urine collection, culture results, and follow-up imaging studies.
Costs associated with UTI included both inpatient and outpatient medical costs (including outpatient pharmacy costs) and were estimated using the Cost Management Information System (CMIS) of KPNC.8 CMIS is used to estimate costs of inpatient stays in KPNC-owned hospitals and outpatient visits to KPNC medical offices (clinics), as well as costs of all laboratory and radiology tests associated with that visit or admission. Costs incurred within a KPNC emergency department or urgent care clinic are included as outpatient costs.
CMIS integrates the regional utilization databases with the general ledger of the KPNC, thus providing fully allocated costs by department, medical center, patient, or procedure. CMIS includes systems that record hospitalizations, length of stay, operating room time, outpatient visits, selected outpatient procedures, all laboratory tests, and radiology procedures for services provided within KPNC facilities. CMIS costs include the overhead and administrative costs associated with clinics and hospitals. Indirect program costs equivalent to the costs of insurance-related functions, such as membership accounting and some regional health plan administration, are excluded.
The Mann-Whitney U test9 was used to compare the age of diagnosis of UTI for uncircumcised and circumcised male infants. The 1996 data showed the number of UTI cases among the 28 812 infants born that year, but incidence figures could not be calculated because the analysis did not include 12-month follow-up for each infant. To determine valid UTI incidence, an additional cohort of 20 587 newborns from 1997 was analyzed. Each infant had been monitored for 12 months. Infants eliminated from the 1997 cohort included those infants born outside the health plan, infants transferring into the health plan, and infants leaving the health plan before 1 year of age.
Of the 14 893 male infants born during 1996, 9402 (63.1%) were circumcised before hospital discharge. An additional 266 (1.8%) male infants were circumcised as outpatients in the newborn period after discharge, yielding a total 9668 (64.9%) circumcisions among these newborn male infants.
During 1996, 446 UTI cases were diagnosed in infants <1 year old (Fig 1). Of these, 292 cases occurred in female and 154 in male infants. Of the 154 male infants with UTI, 132 (86%) were uncircumcised. Hospital admission was required for 38 (28.8%) of the 132 uncircumcised male infants, 4 (18.2%) of the 22 circumcised male infants with UTI, and 22 (7.5%) of the 292 female infants with UTI (Table 1). Mean length of stay was 2.5 days for male infants and 2.3 days for female infants.
Chart review was performed for 52 (34%) of the 154 infant boys with UTI. Of the 52 boys in this sample, 45 (87%) were uncircumcised, and 7 (13%) were circumcised. Urine was collected by catheter in 36 (69%), by bladder tap in 3 (6%), by bag specimen in 11 (21%), and by unknown means in 2 (4%). Urine culture was positive (bacterial colonies >105) in 47 (90%) of the 52 boys. In 5 (10%) of the 52 cases, positive urine culture results were not documented, and the diagnosis was made based on clinical observation and positive urinalysis results (ie, pyuria, bacteriuria, and positive leukocyte esterase test results). In 1 case, a catheterized culture specimen was taken but was not received by the laboratory; in 1 case, the specimen was received 16 hours after collection; the other 3 cases showed no evidence that any culture was taken. Of the 47 positive culture results, Escherichia coli was isolated in 36 (77%),Klebsiella in 5 (11%), Enterococcus in 4 (9%),Serratia in 1, and Pseudomonas in 1. Forty-eight (92%) of the 52 infant boys had renal sonography, 36 (69%) additionally had voiding cystourethrography, and 3 (6%) had intravenous pyelography; no imaging studies were performed for 4 patients. Urinary tract anomalies were found in 4 boys (8%): 1 uncircumcised boy had posterior urethral valves; 1 uncircumcised boy had bilateral hydronephrosis; and 1 circumcised and 1 uncircumcised boy had unilateral hydronephrosis attributable to ureteropelvic junction obstruction. Three infants, all uncircumcised, had 2 episodes of UTI during the year, and 2 uncircumcised boys had bacteremia—1 attributable to E coli and the other attributable toEnterococcus. At their parents' request, 1 uncircumcised infant was circumcised at 6 months old, and another infant was circumcised at 8 months old; 1 of these boys was circumcised after having E coli sepsis, and the other was circumcised after having 2 episodes of UTI.
Total costs for treating UTI were 10-fold higher in uncircumcised than in circumcised male infants (Table 1) despite the fact that uncircumcised infants made up only 35.1% of the patient base, reflecting the greater number of UTI cases diagnosed, as well as the higher number of hospital admissions among uncircumcised male infants. Higher total costs for male infants corresponded to the nearly fourfold increased frequency of hospital admission among male infants with UTI than among female infants, especially during the first 3 months of life. Mean outpatient cost for UTI treatment (including medication) was $239 per male and $267 per female patient. The annual cost for all episodes of UTI (292 in females and 154 in males) averaged $542 per case for females and $1111 for males. Median age at diagnosis of UTI was 2.5 months old in uncircumcised male infants, 4.5 months old in circumcised male infants, and 6.5 months old in female infants (Table 2); uncircumcised male infants (n = 132) differed significantly from female infants (n = 292; P < .001).
Mean number of pediatric outpatient visits during the first 6 months of life did not differ significantly for circumcised versus uncircumcised male infants: uncircumcised male infants had a mean of 8.6 visits, whereas circumcised male infants had a mean of 8.7 visits. Similarly, the number of total outpatient visits (including pediatric, emergency, and specialty visits) during the first 6 months of life was 9.2 for uncircumcised male infants, and 9.2 for circumcised male infants.
In the 1997 cohort of infants monitored for 12 months from birth, incidence of UTI during the first year of life was 1:47 (2.15%) among uncircumcised males, 1:455 (.22%) among circumcised males, and 1:49 (2.05%) among females. The odds ratio of UTI among uncircumcised compared with circumcised boys was 9.1 (P < .001); the 95% confidence interval was 5.2 to 15.7.
Our findings confirm the strong body of evidence attesting to the protective effect of newborn circumcision against UTI in the first year of life,1–4 ,7 an age when such infections are most severe and likely to lead to hospital admission. A meta-analysis of 9 clinical trials3 showed that, on average, during the first year of life, UTI was 12 times as likely to develop in uncircumcised male infants as in circumcised male infants. The increased risk (odds ratio) of UTI varied from 4.9 to 89 in individual studies. We found that the odds ratio of UTI in the first year was 9.1 with an 18-fold greater chance that uncircumcised male infants would be admitted to the hospital for severe infection in the first year of life, particularly during the first 3 months of life.
Chart review for 52 of the 154 male infants with UTI confirmed the high prevalence of UTI in uncircumcised boys. The most frequent method of urine collection was catheterization, and the predominant organism found was E coli. Imaging studies (usually renal sonography) were performed for almost all infants (91%) and showed that few infants (8%) had urinary tract anomalies. These chart review findings support the usefulness of our hospital and outpatient coding for UTI.
The mechanism by which the foreskin predisposes to UTI has been well elucidated.10–13 Within the first year of life, uncircumcised male infants were found to have statistically significantly higher local bacterial colony counts on the glans than did circumcised male infants, including higher numbers of uropathogenic bacteria, particularly fimbriated E coli.10–12The heavy periurethral colonization that occurs in uncircumcised male infants during the first few weeks of life gradually decreases during the first year and is scanty after 5 years old. Photoelectron microscopy shows preferential binding of uropathogens, particularly fimbriated E coli, to the mucosal surface of the foreskin compared with the squamous surface of the glans.13 E coli adhesin proteins are usually located on the distal tips of the bacterial fimbria.14 Receptivity of genitourinary mucous membranes to the bacterium depends on the binding capabilities and on distribution and density of adhesin receptors on the epithelial cell surfaces. Increased epithelial cell receptivity for uropathicE coli and recurrent UTI are linked. Newborn circumcision prevents this preputial bacterial colonization and decreases the risk of UTI particularly during the first year of life, when prevalence and risk of renal sequelae are greatest.11 15–19 Incidence of renal scarring is greatest in patients whose first infection occurs during infancy.15–18 Although UTI during the first year of life is of greatest concern, UTI is also more likely to develop in uncircumcised males than in circumcised males during the preschool years5 as well as during young adulthood.6
In addition to preventing UTI, newborn circumcision can prevent later discomfort and morbidity by decreasing the necessity for invasive procedures, such as catheterization and bladder tap, when evaluating male infants with febrile illness of unknown cause. The urine bag technique yielded no false-positive cultures in circumcised male infants but did result in false-positive urine cultures in female and uncircumcised male infants.20 Similarly, midstream urine specimens from circumcised male infants proved as reliable as suprapubic bladder aspiration for obtaining uncontaminated urine specimens.21 When evaluating a male infant with high fever, a bag specimen or midstream urine specimen yields a valid measure of UTI in circumcised male infants. Suprapubic bladder tap or catheterization is necessary to assure a valid specimen in uncircumcised male infants, although even in uncircumcised males, bag specimen was found valid in 83% of cases.
The cost analysis included in this study represents the actual costs to KPNC of diagnosing and managing UTIs during the first year of life and is not a cost-benefit study. We found that the total costs for UTI were greater in male than in female infants during the first year, although almost twice as many cases of UTI occurred in females. The increased costs in male infants resulted from the greater necessity for hospital admission early in life (27.6% of males vs only 7.5% of females with UTI were hospitalized). Similarly, the 10-fold higher cost in uncircumcised compared with circumcised males resulted from the greater number of cases (132 in uncircumcised males vs 22 in circumcised males) as well as more hospital admissions in uncircumcised male infants (38 in uncircumcised males vs 4 in circumcised male infants).
To determine the cost-benefit of newborn circumcision, the costs of newborn circumcision and its complications must be balanced against all the medical benefits of the procedure over a lifetime. This analysis would include not only the costs of UTI at all ages but also the costs of circumcision performed later in life (because of recurrent balanoposthitis and phimosis or of personal preference), when the procedure is more risky and expensive22 and savings result from prevention of diseases for which evidence shows protective effects of circumcision. These diseases include cancer of the penis and certain sexually transmitted diseases, particularly syphilis, chancroid, and human immunodeficiency virus infection.22–28 A cost-benefit analysis is beyond the scope of this report, which is limited simply to prevalence, morbidity, and costs of UTI in the first year of life.
In this study of 28 812 infants born during 1996 in a large health maintenance organization, 446 cases of UTI were diagnosed in infants <1 year of age. Of these cases, 292 occurred in female and 154 in male infants. Of the male infants with UTI, 132 (86%) were uncircumcised, although they represented only one third of all male infants born during 1996. Incidence of UTI in our study population was 2.05% (1:49) in female infants, 2.15% (1:47) in uncircumcised male infants, and .22% (1:445) in circumcised male infants.
We found a 10-fold increased UTI cost in uncircumcised male infants compared with circumcised male infants, reflecting the larger number of hospitalized cases as well as greater total number of UTI cases in uncircumcised male infants. UTI was 9.1 times as common and cost >10 times as much in uncircumcised compared with circumcised male infants. The findings reflected that uncircumcised male infants had UTI at an earlier age and were almost 4 times as likely to be admitted to the hospital as female infants were, and 18 times as likely to be admitted to the hospital as uncircumcised males were.
Uncircumcised male infants during the first year of life have a higher incidence of UTI, increased risk of hospitalization (particularly in the first 3 months of life), and greater costs of UTI than do circumcised male infants or female infants.
The research was supported by the Kaiser Foundation Research Institute, Northern California Region, Community Service Program.
Geoffrey Machin, MD, Sheldon Orloff, MD, and Michael Oehrli, MPA, CTR, provided helpful comments and suggestions regarding the study.
The Medical Editing Department, Kaiser Foundation Research Institute, provided editorial assistance.
- Received October 13, 1998.
- Accepted August 27, 1999.
Reprint requests to (E.J.S.) Department of Genetics, Regional Perinatal Screening Program, Kaiser Permanente Medical Center, 280 W MacArthur Blvd, Oakland, CA 94611. E-mail:
This report was presented at the annual meetings of the Western Society for Pediatric Research and the Ambulatory Pediatric Association, Region IX and X; January 27–31, 1999; Carmel, CA; and published in J Invest Med. 1999;47:26A.
- Wiswell TE,
- Enzenauer RW,
- Holton ME,
- Cornish JD,
- Hankins CT
- ↵Shaw KN, Gorelick M, McGowan KL, Yakscoe NM, Schwartz JS. Prevalence of urinary tract infection in febrile young children in the emergency department. Pediatrics. 1998;102(2). URL: http://www.pediatrics.org/cgi/content/full/102/2/e16
- Selby JV,
- Ray GT,
- Zhang D,
- Colby CJ
- ↵Hamburg M. Statistical Analysis for Decision Making. 4th ed. New York, NY: Harcourt Brace Jovanovich; 1987:565–567
- ↵Fussell EN, Kaack MB, Cherry R, Roberts JA. Adherence of bacteria to human foreskins. J Urol. 1988;140:997–1001. Comment, page 1001
- Wiswell TE,
- Tencer HL,
- Welch CA,
- Chamberlain JL
- Schoen EJ
- Wiswell TE
- Copyright © 2000 American Academy of Pediatrics