This underscores the point that an effect may appear to be
statistically significant (though come to think of it, I'm not sure
whether the mortality decline in cancer deaths is stat sig) but
clinically meaningless.

        George M. Carter

**
http://www.cancerdecisions.com/012807.html
CANCER DEATH RATE DRAMA

The US cancer death rate declined by one-half of one percent between
2003 and 2004...and the world went wild. In fact, on January 17, 2007,
President Bush paid a rare visit to the National Institutes of Health
(NIH) in Bethesda, MD, to bask in the reflected glory of this alleged
turning point in the war on cancer.

"Progress is being made," Bush claimed, after attending a roundtable
discussion with cancer scientists. "We're spending about $28.6 billion
here at the NIH, which was doubled from 15 years ago." However, the
President failed to mention the equally significant fact that his
administration has cut the National Cancer Institute's (NCI) budget by
approximately $72 million between fiscal years 2005 and 2007.

Bush called the decline in cancer deaths "the steepest drop ever
recorded." While technically true, this statement gives the
unmistakable - and misleading - impression that the decline in the
number of deaths was both dramatic and precipitous. This is not the
case at all. As the respected weekly Cancer Letter trenchantly pointed
out, the number of cancer deaths had either risen or remained the
same, on a year by year basis, for over 70 years since record-keeping
began. Meanwhile, the widely trumpeted decline in US cancer deaths
amounted to only a few thousand - a fraction of a percentage point of
the overall toll taken by cancer in 2004. And in 2003 the fall was
even smaller, amounting to less than four hundred fewer deaths. Yet on
the announcement of that minuscule drop, also, scientists made a
self-congratulatory mountain out of a statistical molehill.

This year, the leaders of the cancer war, accompanied by legions of
enablers in the mainstream media, seemed to lose all sense of
proportion. This was "big news," they intoned, "very exciting;" we've
"finally turned the corner," it is "highly gratifying," "no fluke,"
and so on. John R. Seffrin, PhD, chief executive officer of the
American Cancer Society (ACS), said, "The hard work towards preventing
cancer, catching it early, and making treatment more effective is
paying dramatic, lifesaving dividends." Wow! That's a heavy burden of
unwarranted assumptions to place on the shoulders of such a small
change in mortality statistics.

Dr. Ahmedin Jemal, the epidemiologist who prepared the report for the
ACS, declared that the decline "is not only continuing, but the
decrease [in 2004] is much larger" [than that recorded for the
previous year, 2003].

But "much larger" is a relative term. According to the figures
released by the American Cancer Society (ACS) there were 553,888 US
deaths from cancer in 2004. This compared somewhat favorably to the
556,902 cancer deaths in 2003. 2004's figure thus represented a
decline of 3,014 deaths, just over half of one percentage point. In
2003, by comparison, there were 557,271 deaths, just 369 fewer than
2002's total.

While any drop in cancer mortality is certainly good news, we have to
ask whether the small declines recorded in these two successive years
represent some decisive turning point in the war on cancer (as nearly
everyone has unquestioningly assumed) or whether it may simply be a
statistical glitch in an otherwise unremittingly grim picture. Are we
seeing "light at the end of the tunnel," as Larry Norton, MD, of
Memorial Sloan-Kettering Cancer Center, New York, told Public
Broadcasting System's Jim Lehrer or could it be the headlight of an
approaching train?

Of course, politicians of all stripes were quick to seize the
opportunity and exploit it to the full. However, it remains to be seen
whether this tentative two-year trend (spanning the years 2002-2004)
will be sustained into the present day. While I join in appreciating
each and every extra life that is not lost to cancer, there are good
reasons for skepticism about the self-congratulatory mood that has
overtaken the oncology profession in the light of this news.

For example, one need look no further than the American Cancer
Society's own recently-released publication, 2007 Cancer Facts &
Figures, an annual review that gives a good statistical projection for
cancer incidence and mortality in the current year. For 2007, the ACS
projects a cancer death figure of 559,650. But wait a minute. Isn't
this actually 5,762 more deaths than those recorded for the year 2004,
the year that triggered the recent Presidential celebration?

According to statistics given in consecutive editions of Cancer Facts
& Figures, which are available online, US cancer deaths for the last
five years were as follows:

2002 557,650
2003 556,902 (a decline of 369 over previous year)
2004 553,888 (a decline of 3,014 over previous year)
2005 570,280 (an estimated increase of 16,392 over previous year)
2006 564,830 (an estimated decline of 5,450 over previous year)
2007 559,650 (an estimated decline of 5,180 over previous year)

Perhaps there is some logical explanation for the steep increase in
2005, although I have yet to see one. One is hard pressed to find
anyone who has even noticed this less than encouraging projection,
although such statistics are easily available at the ACS Web site
(www.cancer.org). I came across only one publication that pointed out
this discrepancy: the SeniorJournal.com, which tracks trends of
interest to Baby Boomers (see below). The SeniorJounal's headline
read: "Cancer Society Predicts Cancer Deaths to Increase in 2007
Despite Long Rate of Decline." This seems to me to be a more accurate
and newsworthy way of reporting the story. Yes, the absolute number of
US cancer deaths did decline a little between 2002 and 2004; however,
this year's projected death toll is already more than it was at the
end of 2003. Unfortunately, this inconvenient fact does not fit the
propaganda needs of those who are committed to the notion that we are
making slow but steady progress in the war on cancer. By and large,
the media latched uncritically onto the ACS and NCI press releases,
and ignored any inconvenient truth that might disturb the rosy picture
that these organizations systematically project.

(The issue may be further clouded by the fact that in 2006, ACS
statisticians modified their method of tallying cancer mortality. They
claim that they now utilize a new and more accurate method of
projecting mortality figures, including a larger sample that includes
86 percent of the population. One can only be thankful for improved
scientific record keeping. But I read no analyses claiming this new
method affected estimates of the overall cancer death rate.)

If the projected 2007 figures prove accurate, as those for previous
years generally have, then the alleged "turnaround" in the cancer
mortality figures between 2002 and 2004 will turn out to have been a
temporary phenomenon, a statistical hiccup that, absent the political
agendas of the various key players, would normally have passed without
notice outside of a narrow circle of biostatisticians.
Enter the Baby Boomers

There are also other reasons for skepticism. It is no secret that
cancer is primarily a disease of one's later years. According to
government (SEER) data, the median age of US cancer patients at death
is 73 years. Of course, this varies by tumor type as well as other
factors: for example, the average age at death is 69 years for breast
cancer, 71 for lung cancer, 75 for cancers of the colon and rectum,
etc. Ethnicity is also relevant: for example, African-Americans in
general die of cancer at a younger age than so-called Caucasians.

All other things being equal, however, a nation's death tally from
cancer largely depends on the age of its population. Put another way,
the larger the senior citizen segment of the population, the more
cancer deaths there will be overall.

As is well known, a great many people alive today were born in the
post-World War II era, between 1946 and 1958. They are collectively
known as the Baby Boom generation. (Some experts now contend that the
definition of this postwar generation should be expanded to include
people born through 1964.) This cohort began when GIs returned from
the War and the US economy expanded rapidly. The birth rate during
this period was more than 50 percent greater than for the preceding
generation, which was born during the Depression years. In fact, there
were over 79 million babies born into the Baby Boom generation.

In the next few years, as these Baby Boomers enter their senior years,
cancer incidence will also tend to increase. In about ten to fifteen
years, the Baby Boomers will start to die of the diseases of old age,
including, of course, cancer. The impact of such demographics will
then be solidly felt on national mortality figures for cancer, as well
as other diseases. Thus, unless something changes radically in the
sphere of cancer treatment or prevention, we can expect that there
will be a steady upturn in the overall cancer death statistics,
coinciding with the aging of the Boomer generation, and that this will
continue for a decade or longer.

The reality of rising cancer mortality figures in an aging population
will hit us hard, and it is something for which our leaders appear
unprepared, both intellectually and practically. It is likely to make
all the current talk about the decline in cancer mortality, and the
conquest of cancer, sound increasingly hollow.

Signature
--Ralph W. Moss, Ph.D.