1 figure omitted
The publication of three recent studies showing that circumcision of adult, African, heterosexual men reduces their risk for acquiring human immunodeficiency virus infection and other sexually transmitted infections1- 4 has stimulated interest in the practice of routine newborn male circumcision (NMC) and the benefits it might confer for HIV prevention. In the United States, rates of in-hospital NMC increased from 48.3% during 1988-1991 to 61.1% during 1997-2000.5 To monitor trends in in-hospital NMC during 1999-2010, CDC used three independent data sources (the National Hospital Discharge Survey [NHDS] from the National Center for Health Statistics, the Nationwide Inpatient Sample [NIS] from the Agency for Healthcare Research and Quality, and the Charge Data Master [CDM] from SDIHealth) to estimate rates of NMC.* Each system collects discharge data on inpatient hospitalization.
NHDS uses an 8% sample of short-stay hospitals (hospitals with an average length of stay for patients of less than 30 days) or those whose specialty is general medical or surgical (including children's hospitals) regardless of length of stay, through a three-stage stratified, clustered design from 50 states to generate weighted national inpatient hospitalization estimates. NHDS collects a random sample of discharge records from hospitals sampled based on strata formed by geographic region, primary sampling unit, and service status and specialty group, then on annual discharge volume within strata. NHDS data are cross-sectional, recorded in International Classification of Disease, Ninth Revision (ICD-9) codes, and available for public use with a 2-year lag.†
NIS uses a sample that approximates 20% of U.S. community hospitals, defined by the American Hospital Association to be all nonfederal, short-term, general, and other specialty hospitals, excluding hospital units of institutions, through a five-stage stratified design currently from 42 states to generate weighted national inpatient hospitalization estimates. NIS collects 100% of discharge records from hospitals sampled based on geographic region, ownership, location, teaching status, and bed-size category. NIS data are cross-sectional, recorded in ICD-9 codes, and available with a 2-year lag.‡
CDM is a convenience sample of health-care reimbursement claims from a 20% sample of U.S. short-stay, acute-care, and nonfederal hospitals from 48 states and the District of Columbia. CDM data are recorded in ICD-9 and Current Procedural Terminology codes and are available with a 2-month lag.§
All three data sources underestimate the actual rate of NMC because none of the datasets include NMC performed in the community. Rates of NMC through the first 28 days of life were calculated for the most recent 10 years of available data from each data source (i.e., 1999-2008 data from NHDS and NIS, and 2001-2010 data from CDM), and a Poisson regression model was used to calculate the average annual percentage change (AAPC). The changes in incidence estimated from the three data sources were compared using the trends homogeneity test.
For the period 1999-2010, the weighted analysis yielded 11,789,000 (59.1%; 95% confidence interval [CI] = 59.1%-59.2%) of 19,933,000 and 12,347,096 (57.8%; CI = 57.8%-57.8%) of 21,359,690 newborn males circumcised in the United States from NHDS and NIS, respectively. Of 2,339,760 newborn males recorded in CDM, 1,306,466 (55.8%; CI = 55.7%-55.9%) were circumcised.
Incidence of NMC decreased from 62.5% in 1999 to 56.9% in 2008 in NHDS (AAPC = -1.4%; p<0.001), from 63.5% in 1999 to 56.3% in 2008 in NIS (AAPC = -1.2%; p<0.001), and from 58.4% in 2001 to 54.7% in 2010 in CDM (AAPC = -0.75%; p<0.001).
When compared using the trends homogeneity test, the decreases in incidence were statistically different (p<0.01) for the 8 years of commonly available data (2001-2008); however, the maximum difference in absolute incidence did not exceed 5.9 percentage points for any given year.
Many factors likely influence rates of NMC. A recent study found that, after controlling for other factors, hospitals in states in which Medicaid covers routine male circumcision had circumcision rates that were 24 percentage points higher than hospitals in states without such coverage.6 As of 2009, Medicaid provided coverage for NMC in 33 states. The procedure was not covered by Medicaid in 15 states, and two states had variable coverage dependent on the enrollment plan (Sarah Clark, MPH, University of Michigan, personal communication, 2011).
Xinjian Zhang, PhD, Sanjyot Shinde, PhD, Peter H Kilmarx, MD, Robert T Chen, MD, Div of HIV/AIDS Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention; Shanna Cox, MSPH, Lee Warner, PhD, Div of Reproductive Health, National Center for Chronic Disease Prevention and Health Promotion; Maria Owings, PhD, Div of Healthcare Statistics, National Center for Health Statistics; Charbel El Bcheraoui, PhD, EIS Officer, CDC. Corresponding contributor: Charbel El Bcheraoui, firstname.lastname@example.org, 404-693-2038.
*The NMC rate is the number of newborn males circumcised within 28 days of birth in a hospital divided by the number of newborn males discharged from a hospital.
†Additional information available at http://www.cdc.gov/nchs/nhds.htm.
‡Additional information available at http://www.hcup-us.ahrq.gov/nisoverview.jsp.
§Additional information available at http://www.sdihealth.com/data_warehousing/expertds.aspx.
Trends in In-Hospital Newborn Male Circumcision—United States, 1999-2010. JAMA. 2011;306(15):1651. doi: