Tuesday, January 29, 2008

Homogenizing the world

More catch-up in spare minutes: I'm not sure this is a good idea. I've already heard from a local producer that an attempt to sell her product (goat's milk cheese) at the local Wal-Marts fell through because the contract would have required her to provide whatever amount they requested, delivered the day it was requested, to any store they chose. Small producers can't meet those standards, and they certainly can't meet these auditing requirements (or their cost).

Wal-Mart to push for uniform standards for suppliers
By Janie Gabbett on 1/25/2008 for Meatingplace.com

Wal-Mart wants to lead an effort over the next three years with other major global retailers to create common social and environmental standards for suppliers, CEO Lee Scott told employees.

"We believe that there should be one framework of social and environmental standards for all major global retailers. And there should be one third-party auditing system for everyone," he told 7,000 Wal-Mart managers at an internal leadership meeting.

Scott said Wal-Mart is working on such a system with global retail and CIES, a consumer goods network, starting with social standards then planning to expand to environmental stipulations.

Wal-Mart supplier standards

Scott also said Wal-Mart would build specific environmental, social and quality standards into its own supplier contracts. He said these standards would apply to all suppliers who work with Wal-Mart through global procurement, who are domestic importers, or who are manufacturers of Sam's Club or Wal-Mart private brands.

"We have already started doing this, and we hope to extend the requirement to all the suppliers I mentioned within the next three to five years," he said in prepared remarks. He said the company believes suppliers can reduce the amount of energy they use to make Wal-Mart products by 20 percent.

Wal-Mart will only work with suppliers who maintain these standards, will make certification and compliance part of supplier agreements and will ask suppliers to report to them regularly.

To underscore the seriousness of Wal-Mart's commitment, Scott said, "We will favor — and in some cases even pay more i for suppliers that meet our standards and share our commitment to quality and sustainability."

Monday, January 28, 2008

The pandemic that wasn't?

I have 10 minutes, so I thought I'd play a little catch-up. Basically, this article is saying that we got a combination of lucky and prepared to avoid pandemic avian flu. Good summary -- we were lucky it wouldn't (and didn't) mutate that quickly and we were prepared with large investments in vaccine and diagnostic research. The question now, of course, is how to keep those dollars coming in the face of large-scale complacency by the public. Here's hoping this doesn't turn into a "boy who cried wolf" scenario!

Tuesday, January 22, 2008

Epi 101: Publishing bias

Why are we now finding out that antidepressant-making drug companies only publish (the better) 2/3rds of their trials? Is it because drug companies are evil, money-grubbing, and soulless? Not exactly -- it's probably also a form of publication bias. Journals don't like to publish papers that report no differences (although this is a valid, and important, scientific finding). I've recently had that complaint from a reviewer -- we didn't find a significant difference (actually, we did, just not an easy one to distinguish, but I digress), so what's the point in publishing the results?

The point is that science consists of two types of trials:
  1. trials that work, and
  2. trials that don't work.
Number 1 shows us what to do next. Number 2 shows us what not to do next. Both very useful to know. Never publishing number 2 leads to repetition of useless trials. This is one of my pet peeves.

Oh, and the drug companies probably didn't want to report less than stellar findings.

Cloned meat is safe? Or offspring . . . oh, never mind

I've put off blogging for a little while (too much writing in my real job), but I do need to comment about this:

The FDA ruling does not mean you will be eating Beta. It means that people who have spent large amounts of money to clone their best cows will be able to sell you the milk and meat of their offspring. The actual cows are worth too much to butcher.

I'm not taking a stand on this issue -- I just don't want people to be more confused than the media has already made them.

Monday, January 07, 2008

Interesting thought, hard to prove

No, despite Simon's suggestion, I didn't get a special Epiphany post up. Thought about it, but it didn't happen. Oh well.

I was intrigued by this article, though. It's an interesting idea, that violent movies could reduce crime, but there's an inherent flaw: how do you prove it? I haven't read their study, but I can guess that there are a lot of potential biases. Do you control for policing or sentencing changes, which would require using data from a regional or even local scale? Then you would need to include the presence of movie theaters, the dates violent films are shown, some sort of temporal analysis relating crime to the release of a movie. If so, your power is going to rapidly decrease (the more possible biases your study considers, the lower your degrees of freedom, the harder it is to notice anything). If not, well, any or all of those things could be biasing your results. The article hints at a possible temporal relationship:
Crime is not merely delayed until after the credits run, they say. On the Monday and Tuesday after packed weekend showings of violent films, no spike in violent crime emerges to compensate for the peaceful hours at the movies. Even a few weeks later, there is no evidence of a compensating resurgence, they say.

What did they use to measure this? Survival analysis? Whatever they did, to really get these results, would be pretty tricky stats, which the article doesn't mention.

Still, interesting thought.

Saturday, January 05, 2008

Epi 101: Treatment Bias

Here's another basic lesson in epi: if you give different treatments based on socioeconomic status, region, or willingness to pay, you will get different results. If you give the same treatments, but spend more money on one, you won't get different results. You will get more money.

That's the sort of bias this book seems to be addressing. There is no scientific basis for our bureaucratic nightmare of a health care system actually improving our health. All it does is improve the financial health of the people who own it.

Why typing is good, but can't do everything

One of my friends sent me a link to news about the Listeria outbreak in Massachusetts. I hadn't followed it too closely because it wasn't really a new story to me, but I have studied these things more than most normal people.

In this case, the bacteria that sickened at least 4 and killed 2 was linked to a milk processor by typing, a useful process that can tell us how related 2 cultures of bacteria are. That lets us go in, shut the plant down, find the culprit, spread the horror stories (one that I heard a few years back: cartons used to take waste milk to a swine herd were pressure-washed in the bottling room), and hopefully learn something.

What typing can't do is stop the outbreak before people get sick. For that, we need to rely on processors to follow S.O.P.'s and farmers to control disease within their herds. This is what my research group focuses on, modeling food safety at all the levels of production. No, it's not "bench" research, but it can be useful.

I'm guessing that the plant in this case will find a simple procedural change that led to a Listeria overgrowth. This is why modeling is important; we can predict what changes will do before people have a chance to get sick. No amount of fancy DNA technology will do that.

Please forgive the rant; I get peeved sometimes about the bias against modeling. We are important!