December 11, 2021
In 1907 Mark Twain famously wrote, "Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark attributed to [British Prime Minister Benjamin] Disraeli would often apply with justice and force: 'There are three kinds of lies: lies, damned lies, and statistics.'" [1]
Disraeli may have been referring to the misuse of statistical data to trick others, a fallacious tactic often used to boost a weak argument. Let’s examine a tweet that an anti-circumcision organization published last month. Intact America posted a meme stating, "11.7% of pediatric surgery malpractice lawsuits involve circumcision!" Intact America warned, "Next time someone tells you there's not much risk in amputating healthy tissue from a baby boy's genitals, tell them they're wrong. #circumcision #malpractice." [2]
Note that the neither the meme nor the message provides any information about the source of the statistic. Readers should consider the lack of a source citation as a warning sign that an author may not wish viewers to examine the validity or context of a statistic or fact. [3] In this case Dan Bollinger, vice president of the Intact America board of directors, provided the source to a Circumcision Choice administrator upon request.
Analysis
The source is a study of litigation involving pediatric surgery published in 2020. [4] Consulting a legal database the researchers found a total of 4,754 pediatric surgery malpractice cases in the United States during the period 1965-2017. After eliminating cases that didn't include a care provider or health center, the researchers reviewed 170 cases that met their criteria, of which 19 involved circumcision.
The overall percentage of pediatric malpractice lawsuits involving circumcision was reported to be 11.1%. The researchers separated this statistic into lawsuits that involved surgeons and ones that involved non-surgeons. Circumcision was associated with 10.9% of malpractice lawsuits involving surgeons; the figure was 11.7% for lawsuits involving non-surgeons. Intact America cited - perhaps inadvertently - the non-surgeon figure, and erroneously applied that figure to all pediatric surgery malpractice lawsuits. While 11.7% is not that much higher than 11.1%, this error shows that Intact America continues to be sloppy with respect to its facts and figures.
This statistic - the percentage of malpractice lawsuits involving circumcision - provides no insight into the overall safety or risk of circumcision. What Intact America has done would be like saying that motor vehicle accidents cause 20% of all child deaths in the United States - more than any other cause of death - and then conclude that motor vehicles are too dangerous for child passengers. [5]
A statistic that would relate to the risk or safety of circumcision would be the percentage of circumcisions that result in a malpractice lawsuit. While Intact America may have no interest in that percentage, we have completed the work necessary to arrive at such a figure. According to the study, 8 of the 170 cases were related to allegations about informed consent, of which 3 involved circumcision. In these informed consent cases the plaintiffs alleged that parents or guardians did not "gain adequate knowledge and have a thorough understanding of the ensuing procedure." [4] While "informed consent is necessary prior to procedures from both a legal and moral standpoint," [4] such cases do not concern the actual safety or risk of a surgical procedure. Setting aside informed consent, the 16 remaining circumcision cases represent 9.9% of the 162 remaining cases studied that involve actual harm. What percentage of the 4,754 cases in the legal database involve circumcision is unknown. If one applies the 9.9% from the studied cases, then one would arrive at an approximate total of 470 cases. That figure may represent a best estimate of the total number of pediatric surgical malpractice lawsuits involving circumcision in the United States during the 1965-2017 study period.
A precise total of circumcisions performed in U.S. medical facilities each year is difficult to ascertain. AHRQ, an agency of the U.S. Department of Health & Human Services, reported that "in 2006, over 1.2 million circumcisions were performed in the hospital." [6] We will assume that the average number of pediatric circumcisions performed annually by medical professionals in the United States during the study period was 1 million. Using that estimate the total number of pediatric circumcisions during the 53-year period would be 53,000,000. The percentage of pediatric circumcisions that resulted in a malpractice lawsuit would be 0.0000887 or 0.009% - fewer than one in a million.
Conclusion
The Intact America tweet provides a classic example of the Red Herring fallacy. The percentage of pediatric surgery malpractice lawsuits that involve circumcision has no bearing on the risk or safety of infant circumcision. Ironically this study shows the opposite of what Intact America intended. Rather than showing that circumcision is a risky surgical procedure, the study demonstrates that circumcision is incredibly safe. This example provides a textbook demonstration of how a true statistic can be misused to reach a false conclusion. We call it statistical malpractice.
[1] Mark Twain; "Chapters From My Autobiography"; North American Review; 1907, p 471
[2] Intact America tweet, November 22, 2021. Intact America reposted the tweet on November 28.
[3] Circumcision Choice never asks our readers to take our word for it. We almost always provide a footnote - and a link, when available - for statistics and other medical facts that we cite.
[4] Barrie S. Rich et al; "Litigation involving pediatric surgical conditions"; Journal of Pediatric Surgery; 2020 [5] Rebecca M. Cunningham, et al; “The Major Causes of Death in Children and Adolescents in the United States”; The New England Journal of Medicine; December 20, 2018
[6] ”HCUP Facts and Figures, 2006: Statistics on Hospital-based care in the United States"; Agency for Healthcare Research and Quality; p 33
Commenti