A regularly-raised question is whether breath testers are accurate enough to be reliable. There are really two underlying questions to this debate—in what ways are breath testers reliable, and what could be relied upon in their stead?
The law already understands the limitation of the breath tester: it is, after all, not measuring actual blood alcohol concentration (BAC) in the bloodstream but rather, the implied BAC through the medium of air passing through lungs (see Mclachlan v Mackey, 1994). If something interferes with that reading, then that particular result becomes questionable. Such interference has been found to be creatable by for example; certain types of breads, vigorous exercise and even higher acetone levels in some people who suffer from diabetes (see “Breath Analyzer Accuracy” by David J Hanson, PhD). We need to remember that the breath tester is actually measuring levels of chemicals in the breath, and using that to give a reading of the levels of alcohol in the bloodstream. The devices use the breath to indicate what’s happening in the blood—they do not directly take or measure blood samples and therefore the current technology can only ever be claimed to be “indicative.”
Naturally, the higher the quality of the device that is used, the more likely it is to be able to take potential interference into account and thus provide a more accurate result. One of the very best machines on the market, and one designed for law enforcement use such as the Dräger Alcotest 7110 uses two different sensors to derive its readings from, and is the size of a small suitcase. However, it also costs over $10,000—hardly a reasonable purchase for the average family! Even these machines provide only indicators of intoxication, and so when people are stopped for suspected drunk driving, they will often be required to undertake at least one field breath test, plus another breath test in a booze bus, a blood test or both. Next time there’s a drunk driving ad on TV, pay attention to the words as the person blows over .05 and you’ll hear the police officer say “I suspect that you are above .05”—it’s for this very reason. Even the police can’t claim that a single breath test is accurate. And it doesn’t matter what type of test you are required to take, a failure to follow operating procedures can be problematic—even blood tests can be tainted from failure to follow correct procedures (see for example “40 ways to beat a DUI”).
As you can see, a single breath test shouldn’t be relied upon as conclusive as to whether you are safe to drive or not. Indeed, knowing your own physiology is important to understand what may be affecting your breath test readings at any one time. However, what could be relied upon if you do not use a breath tester? There are three possible choices: guesswork, calculation or waiting to be stopped by the police. Guesswork is the method of choice for many people, who use the rather ethereal answer to the question “How drunk do I feel?” as a way of deciding whether they are fit to drive home after a night at the local pub. A less subjective (yet no less problematic) method is the calculation method, where the person tries to consider the factors that would impact their sobriety to arrive at a decision about their possible intoxication. Some people drive home knowing they might be drunk, hoping they won’t get pulled over for a police test even though this would be their only means of knowing if they’re too drunk to drive or not.
How can most people have any indication of their level of intoxication? The assumption that people “can tell” is challenged every single day as hundreds of news reports from across the globe attest. One of the more interesting ones was from the “Ely Standard”, a UK newspaper that reported how someone assumed they were okay, but they weren’t. Why? Because they had a home remedy for a sore throat, and that remedy was shots of spirit. You can’t just assume people “know” how drunk they are without any way of testing it. Naturally, in an ideal world, people would not drink and drive at all. But people do have a drink and then drive, and relying on guesswork or attempts at calculation will not provide a reliable way of knowing if you really are going to be safe. You are better off having an independent assessment rather than a subjective (dare we say “Beer Goggled”) opinion of your level of intoxication.
So when you consider whether or not to purchase a breath tester for a new driver in the family, a loved one or yourself, remember that they’re a great way to help you determine whether you’re safe to drive or not, but they are only able to indicate your BAC and always err on the side of caution. If the breath tester indicates you’re okay but you’re not sure, trust yourself. If you think you’re okay but the breath tester says otherwise, trust your tester!

References
Ely Standard Newspaper, “Drunk driver had been using strong alcohol as an anaesthetic for mouth pain”, 3rd August 2009
Hanson, D. (2007), “Breath Analyzer Accuracy”, http://www2.potsdam.edu/hansondj/DrivingIssues/1055505643.html downloaded 9th September 2009.
Peek, D. (1998), “Intoxication and the Criminal Lawyer” in The Law Society of South Australia Bulletin, April 1998.
Ramsell, D.J. (2009) “40 Ways to Beat a DUI” on Findlaw, http://dui.findlaw.com/articles/7510.html downloaded 9th September 2009.