LETTERS

LETTERS

Circumcision: Benefits must outweigh the risks

Congratulations on publishing "What to tell parents aboutcircumcision" (February), as neutral a paper on newborn circumcisionas I have ever read. One quibble I have is that the authors didnot present any data showing whether the observation that uncircumcisedboys have more urinary tract infections has clinical significance.Does it matter?

A larger problem is the very neutrality of the article, whichimplies that a pediatrician's job of providing parents with abasis for informed consent ends with regurgitating undigestedfacts. Routine neonatal circumcision--a permanent, mutilatingprocedure done on a nonconsenting human being--cannot be defendedon purely medical grounds. One may or may not advocate the procedurefor religious or cultural reasons, but before pediatricians recommendcircumcision on its medical merits we must insist that it passthe same test we apply to any medical procedure advocated forlarge groups: The benefits should clearly and vastly outweighthe risks. Parents have a right to expect that we will help themsort out and weigh the medical risks and benefits.

The authors discuss some putative medical benefits of circumcisionand present a list of risks, which includes death. They clearlyshow, but do not or cannot say what is obvious from their ownarticle: We cannot tell a baby we are preparing to circumcisethat what we are about to do is perfectly safe and definitelyfor his own good.

Stephen J. Harris, MD
San Jose, CA

I appreciated your publishing the article on circumcision,though I disagree with the authors' interpretation of data atseveral points. They state that "the most certain benefitof neonatal circumcision is that it prevents later developmentof squamous cell carcinoma of the penis." The routine amputationof a normal body part at birth to prevent the (much) later developmentof a very rare form of cancer is a curious proposition. Usingthe same logic, one might suggest performing bilateral mastectomieson all newborn girls to prevent the development of breast cancer--obviouslya ludicrous proposal.

As the authors point out, "there are no medical indicationsfor neonatal circumcision." It is performed in this countryfor religious or cosmetic reasons and because of misinformation.No rate of complications is acceptable for an unnecessary procedure.It is our duty to better inform the public and to always remember,first do no harm.

Pierpaolo R. Palmieri, MD
League City, TX

While the authors do an admirable job of bringing togethera balanced review of the information available about neonatalcircumcision--a topic rife with controversy--their article hasa few inaccuracies.

The authors mention, without citation, a study showing thathuman papillomavirus is more common in men who are not circumcised.An exhaustive Medline search failed to uncover such a study. AFrench study found no difference in the rate of HPV infectionin circumcised and normal men (Aynaud O et al: Cancer 1994;74:1762).

We have no clear evidence that neonatal circumcision preventspenile inflammation. To the contrary, two studies found that whilestill in diapers, circumcised boys have more problems with penileinflammation than uncircumcised boys (Van Howe RS: Br J Urol 1997;80:776;Fergusson DM et al: Pediatrics 1988;81:537). Another study foundno difference (Herzog LW et al: Am J Dis Child 1986;140:254).All circumcised boys will have a reddened, inflamed, abnormallyexposed glans penis immediately after circumcision.

The discussion of the risk of penile cancer contains many inaccuracies.Penile cancer is rare, yet the authors cite a lifetime risk inAmerican men with normal genital anatomy as one in 600. This numberresults from manipulating national estimates using the assumptionthat only men with foreskins can develop penile cancer. A studyfrom Seattle, in which 42% of 110 men with penile cancer werecircumcised, shows that this assumption is false (Maden C et al:J Natl Cancer Inst 1993;85: 19). If circumcision prevents penilecancer, why do other first-world countries that do not circumcise,such as Finland, Norway, Denmark, and Japan, all have a lowerincidence of penile cancer than the United States?

Robert S. Van Howe, MD
Michelle R. Storms, MD
Minocqua, WI

Christopher J. Cold, MD
Marshfield, WI

Frederick M. Hodges
Oxford, England

The authors reply: We thank Doctors Harris, Palmieri, and VanHowe and colleagues for their thoughtful comments. Dr. Harrisquestions the clinical significance of the circumcised male'sdecreased risk of urinary tract infections. We feel strongly thatclinical benefits indeed exist. One only needs to consider theoutpatient workup of a febrile infant, including the monetarycost of the CBC and blood culture and urinalysis/culture, theemotional cost to the parents and infant, and the physical riskto the infant who may need a urine culture obtained by catheterizationor suprapubic tap. Since circumcised males have a decreased riskfor UTI, their risk of having a febrile illness requiring a medicalevaluation should be similarly decreased. In our practice a febrile,circumcised male infant is also much less likely to undergo theinvasive procedures to obtain urine for culture.

Should the infant actually have a UTI, hospitalization withintravenous antibiotics for 48 hours is still the most commontreatment, and medical follow-up may include urinalyses and suppressiveantibiotics until diagnostic tests have been obtained. Obviously,all this diagnostic testing and treatment has a financial andemotional cost to the family as well as a financial cost to society.One study reported that the cost of hospitalization for all childrenwith UTIs younger than 15 years in New South Wales in 1994 was$1.6 million, with the cost of outpatient renal imaging proceduresan additional $5.3 million (Craig I et al: Journal of Pediatricsand Child Health 1997; 33:434). A new, unpublished study by EdgarSchoen et al on 28,812 infants shows that 86% of the UTIs in malesduring the first year of life were in uncircumcised infants, whowere younger than the circumcised infants when they were diagnosedwith a UTI and hence were more likely to be hospitalized. Themedical cost of caring for these infants was 10 times higher thanthe cost for the circumcised infants.

Dr. Harris also criticizes us for maintaining a neutral stanceon a very emotional topic. We feel our article clearly providesthe information pediatricians need to help parents "sortout and weigh the medical risks and benefits," as he suggests.Parents are ultimately responsible for every aspect of their child'scare, and we should not allow our emotional distaste for a procedureto interfere with provision of accurate medical information tofamilies.

Dr. Palmieri is concerned about the "routine amputationof a normal body part at birth to prevent the (much) later developmentof a very rare form of cancer." We never promoted the ideathat all males should be circumcised for this purpose. Rather,those parents who choose circumcision should know that there aremedical benefits. An uncircumcised boy's risk of urinary tractinfection in the first year of life is approximately 1%; the riskof minor complications from circumcision is also 1%. Therefore,physicians and parents who accurately assess the medical datacan support either decision.

Dr. Harris says that the criteria for performing a medicalprocedure should be that the procedure is "perfectly safe."Unfortunately, nothing in medicine, including immunizations, canmeet this criteria. Dr. Palmieri reiterates Dr. Harris's propositionthat "no rate of complications is acceptable for an unnecessaryprocedure." Adherence to this criterion would mean not performingcorrective surgery for conditions that affect children's appearance,such as congenital ptosis. Our care will never be "perfectlysafe," which is why we must work with parents as they makedecisions for their children.

Dr. Van Howe and colleagues claim that our article has "inaccuracies."With regard to the relationship between human papillomavirus andpenile carcinoma, we note that there are discrepancies in theliterature in this relatively new area of research. In one report,only 24% of the 63 male sexual partners of women with proven HPVinfection were themselves infected, a lower rate than would beanticipated due to the high number of circumcised partners (MaymonR et al: J Reprod Med 1995;40:31). In addition, the article byAynaud that Dr. Van Howe and colleagues cite documents that morethan 57% of the lesions identified morphologically as manifestingneoplasia were located in the foreskin, compared with only 10%on the glans. This small study of 96 men also shows that men whowere uncircumcised were likely to have a more advanced state ofpenile intraepithelial neoplasia--a condition related to HPV infection--thanthose who were circumcised.

We never stated that circumcision may prevent penile inflammation,as Dr. Van Howe and colleagues claim. Rather, we referenced Dr.Van Howe's article on the complications in circumcised males andrepeated his recommendation that parents gently retract remainingforeskin in circumcised males to prevent adhesions from developing.Thus, we agree that circumcised males may have penile inflammation.

Finally, data can always be twisted or misinterpreted, makingthe epidemiology of penile carcinoma a hotly debated topic. Instating that the Maden article shows that "42% of the 110men with penile cancer were circumcised," Dr. Van Howe andcolleagues fail to acknowledge that only 20% of the 110 men withpenile carcinoma were circumcised at birth--a very important distinctionsince later circumcision is known to be less protective againstpenile carcinoma than circumcision performed in infancy. Moreimportant, Maden's study is significantly flawed by the inclusionof patients with carcinoma in situ, which differs markedly fromthe more aggressive squamous cell carcinoma.

Jane E. Anderson, MD
Karl A. Anderson, MD
San Francisco, CA

Office hydration not easy

"Managing acute diarrhea: What every pediatrician needsto know" (February) is outstanding. It definitely made merethink the criteria I use for diagnosing dehydration, and theauthors make a great argument for oral rehydration and for continuingto feed children who have diarrhea.

My only criticism is that the authors make office oral andintravenous hydration seem easier to accomplish than it is. Wehave tried and had to give it up because the time and manpowerrequired were beyond the capabilities of our very busy practice.I believe that most practicing pediatricians would agree.

Stanley Karp, MD
Cinnaminson, NJ

The authors reply: We thank Dr. Karp for his kind words. Ourexperience does not support his observation that office-basedrehydration therapy is difficult to accomplish. We know of manypractices in the Boston area and elsewhere that successfully carryout both oral and intravenous rehydration. As we point out, prerequisitesfor success include adequate office space, nursing staff, andadministrative support. If office-based rehydration can be billedproperly, it may even offset the extra work it requires. Moreover,office-based rehydration is certainly a cost-effective alternativeto emergency department referrals. We believe that most pediatricpractices can and should carry out effective rehydration therapyin the office.

Jack Lasche, MD
Christopher Duggan, MD
Boston, MA

Placing lice resistance in perspective

"Lice: Resistance and treatment" (November) is anexcellent overview for practitioners. Recognizing considerablepublic anxiety about the possible increase of resistance to availablepediculicides, the authors make an effort to place the resistanceproblem in perspective. Most important, they note that "incorrectuse of pediculicides accounts for most clinical failure."Indeed, products must be used precisely as directed on their labelslest treatment failure be attributed incorrectly to resistance.Failure of parents and school officials to differentiate nonviablenits or dandruff from the viable eggs near the surface of thescalp undoubtedly accounts for numerous "treatment failures"

The authors' statements regarding resistance in lice must beinterpreted with caution. Resistance to available agents has beendocumented in Israel, the United Kingdom, and the Czech Republic,but its extent in the United States is unknown. The use of insecticidesand pediculicides is markedly different here. Readers should alsonote that methods for testing for resistance in lice are muchcruder than the antibiotic resistance testing clinicians are usedto interpreting in their practice. It is very difficult to collectlice for testing and impossible to preserve them for more thana brief period. Methods for resistance testing are not standardizedand are still evolving.

Donald Goldmann, MD
Boston, MA

The author replies: As Dr. Goldmann emphasizes, the methodsused to study the susceptibility of lice to pediculicides arevariable, imperfect, and poorly standardized. Based on "clinicallysignificant resistance," however, other countries have usedthese imperfect methods to study the killing effect of pediculicideson lice removed from humans. In several countries, it has beendemonstrated that it now takes 500 times as much pediculicideto kill 90% of removed lice as it took 10 years ago.

In the US, many experienced pediatricians, family physicians,nurse clinicians, dermatologists, and public health officialsnow describe the failure of the same pediculicides that worked10 years ago. Studies to determine that these clinical observationstranslate to confirmatory laboratory data have not been publishedin this country, as Dr. Goldmann states. Consequently, statementsabout resistance in this country are based only on observationsof "clinically significant resistance."

P. Joan Chesney, MD
Memphis, TN