Atul Gawande: Great surgeon, not so great on cancer clusters

by Dick Clapp

Atul Gawande is well-known around Boston because of his skills as a surgeon, but also for his books and articles in the New Yorker, and his interviews with local media.  He was a recipient of one of this yearâs MacArthur grants, in recognition of his work. I got one of his books, "Complications," as a gift and read it and liked it a lot.  He's an incredibly talented writer, and he has ahumane surgeon's view of medical practice.  My father was a surgeon too, so his book resonated with me.

I have another response to Atul Gawande, though, based on an article he wrote for the New Yorker in 1999 called "The Cancer-Cluster Myth."  (PDF here)

That article, which has gotten quite a bit of play since it was written, mentions a Los Alamos, New Mexico brain cancer cluster (7 cases in a neighborhood), and Dover Township, NJ (30% higher than expected childhood cancer) as examples of cluster alarms.  There are concerns about cancer clusters in the news every year, and there is now a coalition of citizens around the U.S. focused on improving public health responses, especially to reports of childhood cancer clusters.  Dr. Gawandeâs article asserts that cancer clusters almost never are caused by any discernible environmental exposure.  He quotes Raymond Neutra and Alan Bender, from the California and Minnesota Health Departments, among others. These two argue that clusters are usually statistical artifacts with no real cause, and that they are a big waste of personnel and resources when they are investigated by health departments.

The article is not helpful, I think, and my experience with the Woburn, MA and Tom's River, NJ childhood cancer clusters leads me to believe that real exposures in communities sometimes lead to real, discernible effects.  Gawande's reference to Dover Township having a 30% excess is really misleading.  That's just an average over a 19-year period, and there was a nearly 3-fold excess in the middle of that interval.  But more to the point, it was really Tom's River, within Dover Township, where the excess childhood cancer was occurring, and this was the location of two Superfund sites, both of which were polluting the groundwater and drinking water. (see the state's Department of Health page on Dover Township ) The New Jersey Health Dept. cancer registry reported that there was a statistically significant 9-fold excess of leukemia and eleven-fold excess of brain cancer in young girls who lived in Tom's River between 1979 and 1995.  This led to an ambitious case-control study that revealed a significant excess risk of leukemia in children of mothers who had access to contaminated drinking water during their pregnancy. The case-control study report came out in 2000, so Dr. Gawande didn't have access to it when he wrote his New Yorker piece, but I would say it contradicted his point.

Also, the seven cases of brain cancer in Los Alamos may have been what got the initial headlines, but there was a much longer-lasting and statistically significant excess of thyroid cancer in Los Alamos. This led to a study by the New Mexico Health Dept. which traced at least some of the cases to exposures to ionizing radiation in workers at the Los Alamos National Lab.  One of our BUSPH doctoral students tried to track down additional evidence of radioactive iodine releases to the community (there were some), but was not able to pry loose all the information that might explain the thyroid cancer excess.  In any event, this was not a pointless exercise in response to an hysterical citizen alarm - it was a real thyroid cancer cluster with a probable cause.

All this leads me to conclude that Atul Gawande is a fine wrter, and undoubtedly a gifted surgeon, but not a public health leader on the issue of cancer clusters and environmental exposures.  Public health departments and public health practitioners still need to respond to citizens concerns and develop better tools and approaches to investigating cancer clusters when there is a suspicious exposure that might be the cause.

Dick Clapp is a professor at Boston University School of Public Health, a member of the Scientific Knowledge and Public Policy Planning Committee, and Co-Chair of Greater Boston Physicians for Social Responsibility.  He was Director of the Massachusetts Cancer Registry from 1980-1989 and has been involved in numerous cancer cluster investigations.

More like this

(This piece appears today at Science-Based Medicine and is re-posted here today because I like it and I'm lazy. --PalMD) A couple of years ago, a number of us raised concerns about an "investigative reporter" at a Detroit television station.  At the time I noted that investigative reporters serve…
This post is part of The Pump Handle’s new “Public Health Classics” series exploring some of the classic studies and reports that have shaped the field of public health. Links to past posts in the series are available here. If you have a favorite Public Health Classic to recommend, let us know in…
By Dick Clapp  Rachel Carson has been in the news quite a lot recently, first as the object of a diatribe by a U.S. Senator, and also in a series of news stories commemorating what would have been her 100th birthday last week.  Tim Lambert at Deltoid has addressed the false allegations about…
By Dick Clapp  Late last month, there was a series of news stories about the drop in cancer deaths reported in 2004 as compared to 2003.  The Washington Post story ran under the headline âCancer Deaths Decline for Second Straight Year,â and the New York Times headline read âSecond Drop in Cancer…

I know at least one masters level environmental health course (and there are probably more) that uses Dr. Guwande's 1999 article to explain the pitfalls of cancer cluster investigations. I recommend that Dr. Clapp's follow-up here be provided to students to supplement the Guwande "myth" article.