Fwd: NCHS Mortality Data request

[Garry Jao]

hey- county-level data at best in any time period. we would need more granular to map cleanly to districts. then, it doesn't seem that we can do what you want...

garry

---------- Forwarded message ----------
From: Rothwell, Charles J. (CDC/CCHIS/NCHS) <cj...@cdc.gov>
Date: Thu, Sep 4, 2008 at 9:31 AM
Subject: RE: NCHS Mortality Data request
To: Garry Jao <gar...@gmail.com>
Cc: "Holmes, Julia s. (CDC/CCHIS/NCHS)" <zb...@cdc.gov>, "Anderson, Robert N. (CDC/CCHIS/NCHS)" <rc...@cdc.gov>

Mr. Jao:

 

You are correct that our public use file no longer contains any geo-coding ... not even at the State level. This change in data release policy has been brought about because many States by either state law or regulation do not allow their data to be distributed in a public use manner at detailed geographic levels and thus could not allow for us to do so. Please see the information at the site below for further information on this issue. That being said ... we have never geo-coded at the district level ... the smallest geographic area we code to is the county. If you wish to make a request for county specific mortality data, you will need to provide us with a proposal. If you have questions on the content of your proposal, please contact Dr. Julia Holmes who I have cc'd.

 

Charles Rothwell

Director of Vital Statistics   

 

http://www.cdc.gov/nchs/about/major/dvs/NCHS_DataRelease.htm

---

From: Garry Jao [mailto:gar...@gmail.com]
Sent: Thursday, September 04, 2008 11:53 AM
To: Rothwell, Charles J. (CDC/CCHIS/NCHS)
Subject: NCHS Mortality Data request

Charles:

I'm currently working on a project for the non-profit http://www.watchdog.net, a website that promotes government transparency. Currently, our website helps users find their district and representative as well as the representative's voting history and district census information. I believe it  would be a useful public service to correlate mortality data with districts, ie the top five causes of death in San Francisco county are such and such.

However, as you know, the public-use mortality data on NCHS eliminates geographic information post-2005 and also limits data to counties with populations 250,000 prior to 2005.  

I understand that you have the complete geocoded dataset available, but it's tightly guarded and available only on a case by case basis. Given the situation, and that I would only be using the data in aggregate at a congressional district level, how would you advise me to proceed?

Best Regards,

Garry Jao

watchdog.net volunteer

(650) 646-8407

[Aaron Swartz]

> hey- county-level data at best in any time period. we would need more
> granular to map cleanly to districts. then, it doesn't seem that we can do
> what you want...

Even county-level data would be nice; I'm hoping to have county pages soon.