Cost Analysis of Neonatal Circumcision

in a Large Health Maintenance Organization

Edgar J. Schoen,* Christopher J. Colby and Trinh T. To
From the Department of Genetics, Kaiser Permanente Medical Center and Division of Research, Kaiser Permanente Medical Care Program
(CJC), Oakland, California

Purpose: We studied the costs of newborn circumcision in relation to its health benefits later in life.
Materials and Methods: We conducted a retrospective database analysis using direct internal cost data from Kaiser
Permanente Northern Californiaa large health maintenance organizationand published cost data (including the cost of
medically indicated postneonatal circumcision). The study cohort consisted of 14,893 male infants born in 1996. One-way
sensitivity analysis was used to demonstrate the impact of selected variables in the model. Monte Carlo analysis was used
to determine the 95% confidence intervals.
Results: Postneonatal circumcision was 10 times as expensive as neonatal circumcision ($1,921 per infant vs $165 per
newborn), and was medically indicated for 9.6% of uncircumcised males. Cost benefits of circumcision resulted from
prevention of infant urinary tract infection, balanoposthitis, phimosis, HIV infection and penile cancer. Assuming initial
neonatal circumcision cost to be $200, the future health care cost offset (avoided) was calculated as $183 (range $93 to $303
in 95% of simulations).
Conclusions: Multiple lifetime medical benefits of neonatal circumcision can be achieved at little or no cost. Because
postneonatal circumcision is so expensive, its rate is the most important factor determining future cost savings from newborn
circumcision.
Key Words: circumcision; costs and cost analysis; health maintenance organizations; infant; infant, newborn

o extend our previous cost analysis of circumcision

related to preventing IUTI in the first year of life,1 we
studied lifetime net medical cost savings attributable
to neonatal circumcision at our large, group model, managed
care organization in Northern California.

METHODS
Costs were studied using published data2 6 and internal
data from Kaiser Permanente in Northern California, which
serves more than 3 million members. The analysis included
a cohort of 14,893 males born in 1996, of whom 64.9% were
known to be circumcised.1
We compared costs and complications of neonatal circumcision with its associated cost savings. We considered only
well established potential benefits, whether directly observed in KP databases or published. Cost of newborn circumcision was discounted. Rates of inpatient and outpatient
UTI in the first year of life decreased,1 as did costs (by

Submitted for publication May 20, 2005.

Presented at annual meeting of Pediatric Academic Societies, Baltimore, Maryland, April 28-May 1, 2001.
Supported by a grant from Kaiser Permanente Medical Care Program, Northern California.
Study protocol received KP Northern California institutional review board approval.
* Correspondence: Department of Genetics, Kaiser Permanente
Medical Center, 280 W. MacArthur Blvd., Oakland, California
94611-5693 (telephone: 510-752-6585; FAX: 510-752-6754; e-mail:
edgar.schoen@kp.org).

0022-5347/06/1753-1111/0
THE JOURNAL OF UROLOGY
Copyright 2006 by AMERICAN UROLOGICAL ASSOCIATION

avoiding the need for postneonatal circumcision) of treating

balanoposthitis, phimosis and paraphimosis.
We estimated a reduction in the incidence of HIV infection and penile cancer. Data on penile cancer were obtained
from a published KP study.7 Costs of care as recorded in
internal KP cost databases were used when available. Sensitivity analysis was done using simulation methods recommended by the United States Panel on Cost-Effectiveness
Analysis. Because treatment costs and outcomes for HIV
infection and penile cancer will probably change in this
century, we analyzed 2 versions of the model, one of which
eliminated estimated cost savings for these diseases.
Cost offset must be discounted back to present-day dollars because it occurs in the future. Therefore, cost offset
from UTI decrease was discounted by 6 months, balanoposthitis/phimosis by 10 years, postneonatal circumcision by
14 years, HIV by 40 years and invasive penile cancer by 65
years. Future costs were discounted at an annual rate of 3%.
Data collection. Lifetime cost of newborn circumcision
was analyzed from the perspective of a third-party payer.
(Because KP is a prepaid medical plan, no claims or charges
are generated.) Data on the cost of circumcision were derived from the KP internal automated cost database, Cost
Management Information System, which itemizes fully allocated costs by department, medical center, patient and procedure. Cost Management Information System uses data
from a separate referral database of medical utilization at
nonKP facilities. Cost of newborn circumcision during initial
neonatal hospitalization was calculated using other pub-

1111

Vol. 175, 1111-1115, March 2006

Printed in U.S.A.
DOI:10.1016/S0022-5347(05)00399-X

1112

COST ANALYSIS OF NEWBORN CIRCUMCISION

lished cost data3 and KP physician, nurse and equipment

costs.
Of the 9,668 neonatal circumcisions performed in 1996
none led to complications observed during initial hospital
stay or severe enough to warrant rehospitalization.1 This
rarity of circumcision complications disqualified these procedures from inclusion in our automated Outpatient Summary Clinical Record (which codes the 70 most common
outpatient diagnoses), and minor complications resulting in
outpatient visits could not be determined from our databases. Therefore, we used other reported values.2 Estimated
on the basis of internal cost estimates for KP office visits,
cost of outpatient followup for complications of neonatal
circumcision was $150. On the basis of published estimates,2
we used a 0.5% complication rate. In 43% of the trials treatment cost saving was greater than $200.
Calculation of costs used in analysis. Cost savings for
early IUTI were calculated using previously determined incidence rates, ie 2.5% for uncircumcised and 0.23% for circumcised male infants.1 Conditional probability of an
inpatient stay resulting from early IUTI was 0.28 for uncircumcised and 0.10 for circumcised male infants. Our finding
of a 10-fold increased risk of IUTI in uncircumcised males
resembled other results.6
On the basis of published studies,8,9 we estimated that
balanoposthitis and phimosis developed in 3.5% of uncircumcised males. Both conditions were assumed to require
brief outpatient care costing $150.
Data from KP and from Finland10where newborn circumcision is not routinely donesuggest that the cost of
postneonatal circumcision is important for analyzing the
cost benefit of neonatal circumcision. In addition, postneonatal circumcision for personal preference is not a covered
benefit either in Finland or at KP. Finnish data from 1996
through 1998 showed that about 7.1% of uncircumcised male
infants require circumcision later in life.10
Circumcision status of KP health plan members was unavailable. Therefore, we estimated the rate of postneonatal
circumcision as 9.6% on the basis of 1,578 procedures done
at KP from 1996 through 1998, during which 16,496 (37%) of
the 44,871 males born at KP were not circumcised. The
number of births was stable in Finland, and in the KP
member population in the decade 1988 to 1998. Therefore,
we estimated the rate of postneonatal circumcision as 9.6%
in the male KP population. In addition, Finnish and KP
rates of postneonatal circumcision were similar for all age
categories and were highest in males younger than 15 years.
Uncircumcised males born at KP from 1996 through 1998
and males who underwent postneonatal circumcision during
that period do not represent the same cohort.
Between 1950 and 1975 the circumcision rate in the
United States and in KP, about 90%, was higher than it is
now, and the KP member population had fewer births because it was smaller. Thus, the actual denominator (uncircumcised males) was probably lower than our estimate.
Moreover, the 9.6% rate estimated for postneonatal circumcision was conservative because it represented only circumcision done for medical reasons. Postneonatal circumcision
done for personal preference or religious reasons was excluded because it is not a covered benefit at KP and, therefore, may have been done outside our system. Our analysis

TABLE 1. Base probabilities used in circumcision cost model

References
Schoen

Schoen et al

Kalcev,8 Holtgrave
and Pinkerton,11
and Oster12
KPNC registry
data and
Gissler10
CDC,13 Ganiats et
al15 and KPNC
registry data
CDC13 and KPNC
registry data
Schoen et al,7 and
Adami and
Trichopolous16

Base (SD)*
Complications of newborn
circumcision
Outpatient UTI with
circumcision
Outpatient UTI without
circumcision
Proportion of UTIS with
circumcision necessitating
hospitalization
Proportion of UTIS without
circumcision necessitating
hospitalization
Balanoposthitis/phimosis

0.0050 (0.0017)

0.0350 (0.0086)

Postneonatal circumcision

0.0850 (0.0072)

HIV with circumcision

0.0015 (0.0007)

HIV without circumcision

0.0030 (0.0015)

Penile cancer with

circumcision
Penile cancer without
circumcision

0.0001 (0.00001)

0.0023 (0.0006)
0.0251 (0.0057)
0.102

(0.0220)

0.2880 (0.0185)

0.0017 (0.0003)

* Estimates are consistent with Kaiser Permanente lists.

used an 8.8% rate of postneonatal circumcision, or the mean

of the KP and Finnish data points.
Because male-to-male transmission of HIV infection is the
major mode of HIV transmission in the United States, our
analysis required estimates of the absolute risk of this mode
of transmission. Estimates of the percentage of males at risk
also were necessary. Estimating future lifetime risk of HIV
infection is necessarily imprecise. On the basis of internal
KP data, we estimated the lifetime incidence of HIV as
0.15% for circumcised and 0.3% for uncircumcised males.
To estimate the base lifetime cost of treating HIV infection, we used a study that estimated lifetime treatment cost
of HIV as $195,000,11 and extrapolated annual data from KP
to a lifetime cost of $45,200, assuming a 10-year lifespan.
Because costs of treating HIV infection will probably change
in the next 30 years, we conducted separate analyses using
a model without HIV cost offset.
Without neonatal circumcision the absolute lifetime risk
of invasive penile cancer in the United States male population has been estimated at 1:600,4 a ratio representing the
cumulative lifetime risk for an uncircumcised male. The
annual risk of invasive penile cancer in United States males
is 1:100,000. Calculated using the internal KP cost system,
the lifetime cost of treating invasive penile cancer is $7,500.
Our analysis assumed that future improvement in technology would eliminate the costs of treating this disease.
Sensitivity analysis was done using SAS software version 8 Monte Carlo analyses, in which every parameter was
varied simultaneously in a distribution centered around the
base value to account for the uncertainty of each base assumption (tables 1 and 2). One-way analysis was performed
on the 6 most influential variables, ie postneonatal costs,
postneonatal rates, inpatient UTI costs, inpatient UTI rates,
HIV costs and HIV rates. For cost estimates a lognormal
distribution was used. Distributions approximated high and
low ranges reported elsewhere1,2,6 9,12,13 and our own data.

COST ANALYSIS OF NEWBORN CIRCUMCISION

TABLE 2. Cost assumptions used in circumcision cost model
References
Jayanthi et al
KP data
Schoen et al1

Base (SD)
3

Community
experience
Learman5
Cameron et al,9
CDC13 and
KPNC data
KPNC data

Newborn circumcision
Circumcision complications
Outpatient UTI
Inpatient UTI
Balanoposthitis/phimosis
Postneonatal circumcision
HIV
Penile cancer

200.00
150.00
150.00
2,200.00
150.00

(25.00)
(25.00)
(25.00)
(500.00)
(25.00)

2,000.00 (400.00)
45,197.77 (2,250.00)
7,500.00 (1,500.00)

RESULTS
Under the base case scenario, assuming a mean newborn
circumcision cost of $200, the cost offset was $183. Thus, the
lifetime cost of newborn circumcision to the provider was
$17. In absolute terms the largest contribution to the cost
offset was postneonatal circumcision. Table 3 itemizes each
cost component.
The value of each variable analyzedpostneonatal costs,
postneonatal rates, inpatient UTI costs, inpatient UTI rates,
HIV costs and HIV ratesvaried over specified ranges (tables 1 and 2), whereas all other variables remained constant
at base values. In all 6 one-way analyses total cost offset
under the most unfavorable assumptions (ie the break-even
cost) was never less than $141. Postneonatal circumcision
costs and rates were subjected to 2-way analysis. If postneonatal circumcision cost is set at $1,500 (the bottom of the
range) and the postneonatal circumcision rate is set at 7.1%,
then the lifetime health plan cost offset of newborn circumcision is $132.
Because future costs of HIV and penile cancer are speculative, additional analysis was performed that assumed
zero treatment costs for both. Under this scenario the base
lifetime health plan cost offset of newborn circumcision was
reduced to $127.
Results of our Monte Carlo analysis (10,000 simulations)
indicated that 95% of the simulated cost offsets resulting
from newborn circumcision ranged from $93 to $303 (mean
$183). Table 3 summarizes expected net lifetime costs of
circumcision using the models described (expressed as mean
cost savings) as well as variability within 95% confidence
intervals.
DISCUSSION
Despite increasing evidence that neonatal circumcision confers lifetime medical benefits, the cost-effectiveness of circumcision remains in question.1,2,4,6 9,1214 Our study
shows that much of the initial cost of neonatal circumcision
is eventually recovered when disease and the need for postneonatal circumcision are prevented. Estimated future returns (which depend on assumptions used for analysis)
range from $93 to $303 of the initial $200 cost of newborn
circumcision (base cost offset $183).
In a 1991 cost analysis of routine newborn circumcision
Ganiats et al concluded that the financial and medical advantages of newborn circumcision cancel each other out, and
that other factors should determine whether the procedure
should be performed.15 The net discounted cost of circumcision was $102, and the net discounted lifetime health cost

1113

was 14 hours of healthy life.15 We found a net cost of $17

using the base assumptions. The data of the 1991 study are
no longer current, and did not include present important
data (unavailable at the time) on preventing HIV and other
STDs, or additional data on penile cancer and IUTI.1,7,16
The cost effect of postneonatal circumcisionthe most influential cost factorwas less than ours but was based on a 4%
rate of postneonatal circumcision reported in a decades old
study.14 Our directly observed postneonatal circumcision
rate and the Finnish rate were about twice as high. If similar rates of postneonatal circumcision were used, our results would closely resemble those of Ganiats et al,15 with
whose conclusion we agree, ie that lifetime medical benefits
and parental choicenot costshould determine whether
newborn circumcision is done. However, the documented
medical benefits of neonatal circumcision now exceed those
described in 1991.1,2,6 8,12,13,16,17
Neonatal circumcision has been associated with decreased incidences of invasive penile cancer,7 foreskin
infection (balanoposthitis), permanently nonretractable
foreskin (phimosis) and poor genital hygiene.8,12 In addition, it is associated with preventive health effects against
UTI, particularly in infants,1,6 and prevents transmission
of HIV9,11 and other STDs,16 as well as penile dermatoses.18 No objective data support any detrimental effect of
circumcision on sexual function, and evidence exists to the
contrary.19
For a small expected lifetime net cost ($17) in the base
scenario a circumcised male infant can expect a 10-fold
decrease in relative risk of IUTI, including a 15-fold reduction in relative risk of serious IUTI requiring hospitalization. With time the boy can expect a 3.5% decrease in
absolute risk of balanoposthitis and phimosis, and an 8.5%
reduction in absolute risk of needing postneonatal circumcision. Uncircumcised men in the United States have a 2:1
increased risk of adult HIV infection and a 22:1 increased
risk of invasive penile cancer.7,20 Male circumcision has
been associated with a decreased risk of penile human papillomavirus infection and cervical cancer in female partners
of circumcised men.16 However, neither the cost benefit of
this advantage nor that of preventing penile dermatoses was
considered in our analysis.18 We assumed only a small increase in cost benefit.
Our study was limited by the use of only direct and
indirect cost analysis without assigning valuation to health
benefits of circumcision. Future treatment will probably
change for all conditions listed in the study. Any reduction in
future treatment costs will decrease the cost offset.
TABLE 3. Expected documented net lifetime cost savings
associated with newborn circumcision
Base (95% CI)
Initial cost
Neonatal circumcision
Complication of neonatal circumcision
Treatment cost change:
Outpatient UTI
Balanoposthitis/phimosis
Postneonatal circumcision
HIV
Penile cancer
Total lifetime cost

200.00 (181.97225.33)
0.75
(0.700.98)
15.92 (20.4913.58)
0.66
(4.320.23)
108.00 (159.5589.67)
57.39 (109.896.43)
1.35
(1.681.09)
17.43
(103107)

Cost savings were calculated using different medical treatment and outcome scenarios ranging from most to least favorable toward procedure.

1114

COST ANALYSIS OF NEWBORN CIRCUMCISION

Our data indicate that complications of newborn circumcision are rare and usually minor. This finding confirms
conclusions of the American Academy of Pediatrics Task
Force on Circumcision.19 Because our database codes all
hospital admissions and the 70 most common outpatient
diagnoses, clinically significant complications would be discovered but some minor complications could be missed.
Without definitive medical indications some infants may
have undergone postneonatal circumcision at the insistence
of their parents. However, parents who initially refuse newborn circumcision are unlikely to demand it later, ie when it
is more complex, more expensive and more likely to involve
complications.
Benefits of reducing HIV infection in the United States
are based on 1 study of homosexual men20 but are nonetheless consistent with results of multiple studies done in less
developed countries, where the medical benefits of circumcision would be greater and where many more people are
infected with HIV than in the United States. The cost offsets
estimated in our study may not apply to underdeveloped
countries, where medical practices and financial factors differ from those in the United States.
On the basis of a meta-analysis and our own KP data, we
calculated a 10-fold higher risk of UTI among uncircumcised
newborn males in their first year of life.2,6 Another study
showed a highly significant protective effect of circumcision
but only about a 3.7 to 9.8-fold higher risk of UTI in uncircumcised males.21 That study relied on data from physician
offices in Canada, and, thus, differed from large United
States series done in large centers with easy access to laboratory analysis.1,6 Lack of such access could have led to
missed diagnoses. Moreover, Canadian circumcision rates
(43%) are lower than United States rates, and outpatient
UTI diagnoses were based not on direct observation, but on
physician billing. These factors could have caused failure to
diagnose IUTI in Canada, particularly among outpatients.
In the KP population (representative of the general population of California) postneonatal circumcision is the most
important factor affecting cost-effectiveness. In the KP experience this procedure costs about 10 times as much as
neonatal circumcision.3,15 Aside from discounting, if 10% of
uncircumcised infants need circumcision later in life, neonatal circumcision is cost neutral regardless of its other preventive health benefits. A total of 59% of the cost savings
arise from preventing postneonatal circumcision. Our 1996
to 1998 study cohort offered a reasonable basis for conservatively approximating circumcision rates because the neonatal circumcision rate at KP (64.9%) closely resembles the
65.3% United States rate reported in 1999.13

CONCLUSIONS
The initial cost of neonatal circumcision is largely defrayed by its resultant future cost savings. Thus, documented lifetime health benefits of neonatal circumcision
are achieved at little cost. We assume that small, favorable cost benefits accrue from preventing penile dermatoses and cervical cancer (ie in female partners of
uncircumcised men) but these factors were excluded from
our analysis because their main evidence was published
after completion of our study.

ACKNOWLEDGMENTS
G. Thomas Ray assisted with data analysis.

Abbreviations and Acronyms

CDC Centers for Disease Control and
Prevention
HIV human immunodeficiency virus
IUTI infant urinary tract infection
KP Kaiser Permanente
KPNC Kaiser Permanente Northern California
UTI urinary tract infection

REFERENCES
1. Schoen, E. J., Colby, C. J. and Ray, G. T.: Newborn circumcision
decreases incidence and costs of urinary tract infections during the first year of life. Pediatrics, 105: 789, 2000
2. Schoen, E.: Ed Schoen, MD on Circumcision: Timely Information for Parents and Professionals from Americas #1 Expert
on Circumcision. Berkeley, California: RDR Books, 2005
3. Jayanthi, V. R., Burns, J. E. and Koff, S. A.: Postneonatal
circumcision with local anesthesia: a cost-effective alternative. J Urol, 161: 1301, 1999
4. Kochen, M. and McCurdy, S.: Circumcision and the risk of
cancer of the penis. A life-table analysis. Am J Dis Child, 134:
484, 1980
5. Learman, L. A.: Neonatal circumcision: a dispassionate analysis. Clin Obstet Gynecol, 42: 849, 1999
6. Wiswell, T. E.: John K. Lattimer Lecture. Prepuce presence
portends prevalence of potentially perilous periurethral
pathogens. J Urol, 148: 739, 1992
7. Schoen, E. J., Oehrli, M., Colby, C. J. and Machin, G.: The
highly protective effect of newborn circumcision against invasive penile cancer. Pediatrics, 105: E36, 2000
8. Kalcev, B.: Circumcision and personal hygiene in school boys.
Med Off, 112: 171, 1964
9. Cameron, D. W., Simonsen, J. N., DCosta, L. J., Ronald, A. R.,
Maitha, G. M., Gakinya, M. N. et al: Female to male transmission of human immunodeficiency virus type 1: risk factors
for seroconversion in men. Lancet, 2(8660): 403, 1989
10. Gissler, M.: Written communication. National Research and
Development Center for Welfare and Health (STAKES), Helsinki, Finland, February 7, 2000
11. Holtgrave, D. R. and Pinkerton, S. D.: Updates of cost of illness
and quality of life estimates for use in economic evaluations
of HIV prevention programs. J Acquir Immune Defic Syndr
Hum Retrovirol, 16: 54, 1997
12. Oster, J.: Further fate of the foreskin. Incidence of preputial
adhesions, phimosis, and smegma among Danish schoolboys.
Arch Dis Child, 43: 200, 1968
13. United States Centers for Disease Control and Prevention. National Center for Health Statistics. Trends in circumcision
among newborns. Available at: http://www.cdc.gov/nchs/
products/pubs/pubd/hestats/circumcisions/circumcisions.htm.
Accessed August 31, 2001.
14. Kaplan, G. W.: Circumcisionan overview. Curr Probl Pediatr,
7: 1, 1977
15. Ganiats, T. G., Humphrey, J. B., Taras, H. L. and Kaplan, R. M.:
Routine neonatal circumcision: a cost-utility analysis. Med
Decis Making, 11: 282, 1991
16. Adami, H. O. and Trichopoulos, D.: Cervical cancer and the
elusive male factor. N Engl J Med, 346: 1160, 2002
17. Circumcision policy statement. American Academy of Pediatrics. Task Force on Circumcision. Pediatrics, 103: 686, 1999

COST ANALYSIS OF NEWBORN CIRCUMCISION

18. Mallon, E., Hawkins, D., Dinneen, M., Francis, N., Fearfield, L.,
Newson, R. et al: Circumcision and genital dermatoses. Arch
Dermatol, 136: 350, 2000
19. Bleustein, C. B., Fogarty, J. D., Eckholdt, H., Arezzo, J. C. and
Melman, A.: Effect of neonatal circumcision on penile neurologic sensation. Urology, 65: 773, 2005

1115

20. Kreiss, J. K. and Hopkins, S. G.: The association between circumcision status and human immunodeficiency virus infection among homosexual men. J Infect Dis, 168: 1404, 1993
21. To, T., Agha, M., Dick, P. T. and Feldman, W.: Cohort study on
circumcision of newborn boys and subsequent risk of urinarytract infection. Lancet, 352(9143): 1813, 1998