Mouth Alcohol 303 mg / 100 mls
Subject with true BAC of less than 120 mg / 100 mls blows 303. A previous tests showed a BACs of 116 mg/ 100 mls. Unknown to the operator the subject had consumed beer a few minutes prior to this test. The subject's mouth was not rinsed after the beer and prior to the test.
The "Invalid Sample" error flag is caused by a spike in the apparent readings as portions of the breath are sampled. If a subject places ethanol in their mouth immediately before a subject test the mouth alcohol, on a recent model evidentiary breath tester, the instrument will, in theory, detect a sudden climb in BAC followed by a decline. This pattern is, in theory, different than a normal bvlow where the BAC rises fairly quickly over the first few seconds and then levels off as deep lung air is sampled. Approved instruments use this approach to detect and flag mouth alcohol.
The problem is that this slope detection system doesn't always work if there is already an underlying true BAC combined with mouth alcohol.
Mouth alcohol can reach equilibrium such that it does not produce a spike in the BAC being read by the instrument during the blow, because there is already an underlying level of true BAC. This video shows that there is a possibility of significant error if the qualified technician does not conduct a personal and careful 15 to 20 minute observation period looking for burps, belches, vomit, and other sources of alcohol in the mouth. The international scientific literature views the waiting period as essential. This experiment was conducted in a lab using an Intoxilyzer® 5000C.