Collective DEN action on possible misunderstanding of Vartiation

[Kromkowski@aol.com]

Dear DEN:
At the very end is an AP article by L.A.Newell, about cancer maps by zip code 
for NY.  Baltimore papers and news also reported on same thing for my area.  
If you go to the website listed at the end of the article, you will see a 
link to a page about "how to read the maps." This page says: in part:
********************************
What do you mean by “expected incidence”?
Expected incidence is the number of people we would anticipate getting cancer 
in a particular ZIP Code.
How did you calculate the “expected incidence”?
Simply, the cancer rates for the entire State and the number of people in 
each age group of a ZIP Code were used to calculate the number of people who 
would be “expected” to get this type of cancer. This calculation assumes 
that people in the ZIP Code have the same risk of getting cancer, and would 
get cancer at the same rate, as people everywhere in the State.
How did you determine that a ZIP Code was higher, lower or about the same as 
expected?
The actual number of people who are diagnosed with cancer in the ZIP Code is 
compared to the “expected” number. The color of the ZIP Code on the map is 
based on how far above or below the “expected” number the actual number is. 
For example, if there are 11 people diagnosed with cancer in a ZIP Code and 
only five people were “expected” to get cancer, the ZIP Code will be dark 
purple on the map. This means the cancer incidence in that ZIP Code is more 
than 100% above what we would expect to find.
**********************

Wouldn't a control chart (I'd recommend a U) be a much better approach to 
color coding instead of this "more than 100% above what we would expect to 
find?"  In other words you'd have three colors, in control, out of control 
above and out of control below.

If the DEN agrees, wouldn't an email campaign, collectively from the DEN, be 
a specific and useful way to extend the understanding of variation.

JDKromkowski

Article:
>>ALBANY, N.Y. (AP) - State health officials have unveiled a ZIP code 
breakdown of breast cancer rates across New York that's the first of its kind 
in the nation.
Health Department officials said the breakdowns provide little insight into 
what causes the cancer that each year claims the lives of 3,500 New York 
women. The statistics are only a tool by which health experts hope to 
ultimately answer that question, officials said.
``Most of the questions that need to be asked are those we must now ask,'' 
Health Commissioner Dr. Antonia Novello said Tuesday. ``Now we really need to 
start the investigation.''
Cancer groups were generally supportive of the latest step in a $4 million 
cancer study.
``I'm actually very pleased,'' said Geri Barish, president of One in Nine, 
the Long Island Breast Cancer Action Coalition. ``You can see your community 
and what's really going on there.''
But Richard Brodsky, a Westchester County Democrat and chairman of the state 
Assembly's Environmental Conservation Committee, said that information could 
be much better.
The maps are ``completely bogus,'' he said. ``This was always about 
identifying cancer clusters in neighborhoods ... By doing it by ZIP code, 
they're hiding the truth from neighborhoods all across the state.''
The Health Department attracted some criticism when it released its first set 
of maps in December that showed a county-by-county breakdown of breast, lung 
and colorectal cancer. Critics said the county breakdown was too expansive an 
area to detect clusters of the disease that some believe can be produced by 
fouled drinking water, pesticides, industrial chemicals or other factors.
The maps are based on the number of women who developed breast cancer in each 
of New York's approximately 1,400 ZIP codes from 1993 through 1997. Among the 
areas that show higher-than-normal rates were Manhattan's Upper East Side, 
parts of Long Island's Nassau and Suffolk counties and Westchester County.
On the Net:
Health Department site: http://www.health.state.ny.us<<

[Moderator's Note:  My zip code is within a few digits of Oak Ridge, TN.  This
is where the first atomic bombs and most nuke weaponary since then has been
manufactured.  The joke is that the trees have purple leaves and the frogs
glow in the dark.

The Zip Code (postal code) and cancer correlation is a subject of extremely
heated debate around here.  So even with purportedly sound data and analysis
techniques -- many different conclusions are being derived.  I am sure that
Neal Mowery and others from this area can attest to this.

The only reason I bring this up is that motive (see SoPK interactions) can 
alter results...
========================================================================