On Fri, Mar 07, 2003 at 12:10:35PM -0800, Bill Stewart wrote: > > Doing the technical part of detecting alcohol vapor is cool, >[...]
Actually, that's not even really a solved problem yet, but that's not well-known outside of people who litigate drunk driving cases for a living. This article - <http://www.forensic-evidence.com/site/Biol_Evid/Breath_Tests.html> does a pretty good job of explaining how the breath testers start by measuring one thing (alcohol in exhaled air) to reach conclusions about something entirely different (alcohol content of the blood). For those who don't like to read links, in a nutshell - The extrapolation of blood alcohol content (BAC) from breath air depends on a few assumptions - that the machine will be able to read deep lung air (where the alcohol concentration in the air will be equal to the alcohol concentration in the blood, due to Henry's Law) and not mouth air (which may be contaminated by residual alcohol liquid or vapors in the mouth, deposited there by drinking, burping, or vomiting, which aren't unusual in people who've been partying) - and that there's a standard conversion ratio (called the partition ratio) which is used to calculate the BAC. That conversion ratio depends on a number of factors (like body size, gender, body temperature, hematocrit density) which differ from person to person, and can differ so widely (between 60% and 150% of the standard) that it's very difficult to get an accurate result without taking the other factors into consideration. Breath testing is the least accurate of the three widely used methods - blood, breath, or urine - though it's the most popular because it's cheap and less invasive than blood or urine sampling. When law enforcement needs a defensible BAC measurement (in death-related DUI cases, for example) they use a blood test (taken by force, if necessary); they use breath tests for misdemeanor cases where they're not as concerned about the outcome. Given the relatively poor quality of the results achieved in a controlled setting for breath testing, I think it's very unlikely that the device described works well enough to achieve anything positive. -- Greg Broiles [EMAIL PROTECTED]
