THE CIRCUMCISION REFERENCE LIBRARY
To: Southern Medical Journal [unpublished]
RE: O'Brien TR, Calle EE, Poole WK: Incidence of neonatal circumcision in Atlanta, 1985-1986. South Med J 1995; 88: 411-5.
Word Count: 302
To the Editor:
I read with great interest O'Brien et al's article "Incidence of neonatal circumcision in Atlanta."1 It's implications on the controversial issue of neonatal circumcision currently going on in the United States are manifold.
The inaccuracy of reporting circumcisions on the hospital medical record face sheet, which hospitals depend on for reimbursement, has enormous impact on the validity of any study that estimates circumcision rates without inspecting individual patients. The data published by Wiswell concerning a possible association between urinary tract infection and circumcision2 deserves special scrutiny. In reviewing the charts of 217,116 boys of military families, where correct face sheets are not a factor in reimbursement, it is very likely only data from the face sheets was used to calculate circumcision rates. Applying O'Brien's 84.3% figure, which may be conservative, reveals that 32,651 (78.11%) of the 41,331 boys classified as having foreskins may have been circumcised. Depending on whether the boys with urinary tract infections were properly classified, the incidence of urinary tract infections in circumcised boys ranges from 0.09% to 0.27% while the incidence in intact boys ranges from 1.12% to 5.12%. The odds ratio varies more than ten-fold (4.18-58.04). This huge potential variation casts a large shadow of doubt over Wiswell's conclusion concerning an association between the foreskin and urinary tract infections.
In Wiswell's study there were unexplained yearly fluctuations in the incidence of urinary tract infections in circumcised boys ranging between 0.07% and 0.21%,3 and he has reported the incidence of urinary tract infections in intact boys to be as high as 4.12%.4 These wide variations are consistent with 72.8% of boys labeled as intact being misclassified. O'Brien's finding clearly demonstrate that any past or future studies that estimate circumcision rates based on chart reviews suffer from a serious methodological flaw that renders any findings derived from them suspect.
Robert S. Van Howe, M.D.
Marshfield Clinic - Lakeland Center
Return to CIRP library