Stakeholders

• Government - department of health, funding / expense for health care.
• The public - people who 'cannot afford to get ill' - sense of importance in their job, mothers cautious about young children, frequent travellers - tourism.
• Pharmaceutical industry.
• The news media industry - weather and health reports, and possible sponsorship associations.

As part of our research, we contacted various health agencies and officials, including one of the professors involved in the original research at Pittsburgh, establishing links between retail data and the health state of health of a place - Professor Latanya Sweeny.

----------------------------------------------------------

Several useful correlations were found between purchases and outbreaks. The research also revealed an interesting note: results were best in poorer communities. This makes sense when you think about it. In these communities, when a person gets ill they often need to go and purchase related products, but more affluent communities tend to have these kinds of supplies (e.g., orange juice, tissues, etc.) already on the shelf.

The citation appears below. We think it will be possible to put a copy of the report on-line this summer, now that the terrorist climate surrounding the project has eased.

The citation is:
L. Sweeney, Early detection of bio-terrorism using grocery purchases. Bio-ALIRT Program, Defense Advanced Research Projects Agency, Washington, DC, 2001.

--Prof Sweeney

At 12:19 AM 3/25/2004 +0100, Matty Sallin wrote:
>Hello again,
>
>Sorry, but I've got one last question: how successful was your project
>in determining a correlation between retail data and illness? I would
>love to see your study if it's online somewhere. I wanted to cite your
>project to support our idea linking retail data to city health.
>
>Thanks again,
>Matty and Ben
>
>-----Original Message-----
>From: latanya@privacy.cs.cmu.edu [mailto:latanya@privacy.cs.cmu.edu]
>Sent: 24 March 2004 15:50
>To: m.sallin@interaction-ivrea.it
>Cc: b.dove@interaction-ivrea.it
>Subject: Re: From 2 Grad Students in Italy...
>
>
>Hi there,
>
>I am delighted at your project. Yes, my earlier work shows that what
>you are seeking is in fact possible. In termsof actual deployment,
>however, it depends on the timeliness of the data you receive. What
>questions do you have?
>
>--LS