Each year, an organization called the Coalition to End Needless Death on our Roadways, distributes a press release to highlight a supposed "Fatal 15" states. Their use of data is misleading and dishonest.

How is it misleading?

Here's the annual procedure for the "Fatal 15." The Coalition gets data from the publicly released National Highway Traffic Safety Administration (NHTSA) statistics (see the preliminary 2005 data -pdf). They then look up which 15 states have the highest percentage of "alcohol-related" traffic deaths on that chart and put out a press release dubbing them the "Fatal 15." (Here's a screen shot of how the data is organized.)

Take a close look at that data and you'll see the problem.

1. The number they use is the combination of all those with some level of alcohol detected in their blood -- anything from .01 or more. Those on the lower end of the scale are not impaired by alcohol and so those crashes shouldn't be considered "alcohol-related."
2. Here's the big one. What they use is the number of "alcohol-related" traffic fatalities as a percentage of total traffic fatalities. That is a meaningless statistic unless you know the overall prevalence of traffic fatalities in that state. In some cases, those states ranked highest in the so-called Fatal 15 are actually the safest when you look at "alcohol-related" traffic fatalities as a percentage of the population.

Here's an example: Assume two states that have the exact same population:

• State A has 200 traffic deaths and 80 of them are alcohol related. According to END, State A would get a 40% rating.
• State B has undergone a strong safety program in their state and have only 100 traffic deaths (50 of them alcohol-related). State B would be a 50% rating.

In this example, even though State B has less overall deaths and less alcohol related deaths, State B would be considered one of the Fatal 15, worse than State A, which is, of course, ridiculous.

The numbers used are meaningless. The total number of alcohol-related deaths at .08 BAC or higher as a percentage of total population might be an interesting figure. The absolute number of alcohol related deaths compared to last year's absolute number of alcohol-related deaths might be interesting. What the Coalition to End Needless Death on our Roadways uses is dishonest.

The Coalition uses these numbers because it's easy. The percentages are already on the NHTSA chart. Once a year, they just label the top 15 states, put it on a press release and send it out as a way to drum up business and donations. It's opportunism of the worst sort -- at the expense of honest information that can be used to save lives.