Kansas Practice versus “Best Practice” for Breath Testing
When respected scientists, researchers, or professionals perform their task, they utilize what is commonly referred to as best practice in order to get the optimal accurate results. Best practice is the technique used and accepted as the superior method of performing something. While there may be alternative shoddy ways of doing or testing something, best practice is the most reliable and accurate method.
Before I became a criminal defense attorney, I just assumed that the government would use best practices in their cases. Alas, surely our prosecutors wouldn’t want to get something wrong or send an innocent person to jail. I was wrong. When it comes to DUI (drunk driving) charges as well as administrative driver license suspensions, the Kansas Department of Health and Environment has come up with its own protocols and methods for testing breath alcohol. Sadly, it is not consistent with Best Practices in the field of testing breath alcohol (BrAC).
One glaring discrepancy between scientific best practices versus what Kansas does, is regarding the number of tests conducted and recorded. For a test result to be deemed scientifically valid, scientists require several tests to gauge and measure reliability. In any case, there must be at least a bare minimum of 2 tests. Ignoring this basic best practice, Kansas’s protocol calls for just 1 test of a suspect’s breath. This should make the people outraged, but fairness and reliability has gone out the window, under the pretense of drinking and driving. Why does Kansas’s protocol for breath alcohol training only require one test? They don’t want more than one test because a second test will generally provide a different result, which of course shines a light on the fact that the Intoxilyzer 9000 or other breath test used is not incredibly accurate. However, with just one test, a typical juror will simply believe, “That’s what the actual BrAC (Breath Alcohol Content) was.”
Most states and jurisdiction require at least 2 tests. Furthermore, the tests are typically required to be within .02 from each other. In other words, not only do other states require at least two tests, but the results have to agree within a certain margin, usually within 0.02 grams (20 milligrams). For example, if the first breath test result was .12 and the second test was .14, (.02 grams or 20 milligrams apart), they would be considered in agreement and thus deemed reliable and valid. Different states have different allowed margins. For example, some only .01 difference. But again, Kansas would allow an infinite variance as it doesn’t even require any other test to measure itself against. This is incredibly unfair and lacks any semblance of scientific reliability, much less even close to best practice.
A second difference between Kansas’s program and best practice is a lack of regular maintenance. Yes. They do maintenance on the machines, but it is rarely actually recalibrated with the main known sample that is as close to .080 as they can get. Why is this an issue? This is an issue, because the breath machine may be measuring BrAC higher or lower than what it actually is. The Breathalyzer tries to address this by doing an external Standard Check during every breath testing sequence. Prior to testing a person, the machine actually checks itself against a known sample that is supposed to be close to .080. So long as the machine reads the known sample to be close to .080, then it proceeds forward with testing the person. However, the machine is allowed to have a margin of error even when testing the known .080 known sample. It is allowed a variance of .01. In other words, right from the beginning, the machine is allowed to test off by a margin. To make matters worse, in Kansas, the known sample that is supposed to be .080, that it is testing, is also allowed to be off by .002 according to the certificate of analysis for the known sample. That would be in addition to the other margins of errors. Therefore, without even considering many other potential errors and false positive, the breathalyzer can be off from the start such that it reads an actual BrAC of .074 to be .081. A best practice would be to require at least an annual recalibration, cleaning and full maintenance, instead of merely doing software updates over the internet. Otherwise, chambers become clogged with mold and mildew due to the bacteria filled human breath going in and out of the machine. Also fuel cells move. The lack of this needed maintenance is due to a desire to save the expense. Apparently, it’s better to send a few innocent people to jail than spend the funds maintain the machine correctly pursuant to best practice. The drift usually seems to go up over time rather than down, meaning these results start consistently overstating the BrAC.
A third difference between Kansas’s practice and best practice is a lack of review and transparency. The Intoxilyzer 9000 is a computer with a source code running it. However, the manufacturer of the machine will not provide access to the source code. In other words, we the people are just supposed to trust, without verification, the results the breathalyzer provides and the numerous errors is encounters. Sadly, as a DUI defense attorney, I have seen the machine encounter numerous errors and provide odd results, yet no one knows why or what the machine is actually doing. How our legislature and courts allow this to satisfy due process and fairness requirements is baffling.
One of the errors often encountered by the breath testing machine is “RFI.” This is called radio frequency interference. Radio waves can cause electrical or magnetic interference that disrupts the breathalyzer’s internal makeup. The machine supposedly can tell when this is happening and the testing procedure is to be restarted. However, there are times when it just magically fixes the problem itself. What is happening? We don’t get to know. We do know, through numerous testing, that the RFI readers on the breathalyzers are not very accurate or reliable.
Other parts of the testing environment plays a crucial role in the results of a breath test results. For example the air in the room can alter results. To address this issue, the breathalyzer will test a sample of the air in the room. This process is referred to as an “Air Blank”. The result should come back at .000 before it proceeds forward with testing the known sample of .080 or the actual DUI suspect. This would be like making sure the grocery store scale is actually set at zero before weighing your bananas.
Sometimes when doing the air blank phase, the machine will detect something and provide an “AMB” result, which means to check ambient conditions. The tester is to ventilate the room and redo the test. This AMB results can be due to things like too much dust, dirt, or cleaning products floating throughout the room. One common culprit is hand soap that contains d-limonene “orange solvent.” Also, AMB can be caused by the tubing inside the device becoming contaminated, or the internal components becoming oversaturated by the standard alcohol solution of the machine. Additionally, the person being tested can cause the error due to too much perfume or even having spilled alcohol or urine on their clothing.
A best practice in such a situation would insist on figuring out what exactly is going on. Unfortunately, what often happens is that the arresting police officer simply redoes the test until it starts working again. This sub-standard way of dealing with the error can affect the reliability of the breath test result. The machine can begin providing false indicators of .000. These floating zero air blanks are indicating a .000, but are not actual true .000. It is not really starting the test out fairly. The breathalyzers often have a pre-programmed pre-determined level of acceptable contamination, and still pass the air blank test. For example, the Intoxilyzer 5000 allowed contamination of 0.009 grams (9 milligrams) before the AMBIENT FAIL error message would generate.
This allowance of contamination is a problem because the "Floating Zero Air Blank" creates a masking effect, reducing any contamination down to an arbitrary zero level. Basically, the machine is floating the known contaminate back down to .000. The manufacturer of the machine claims it takes this into effect and subtracts it. Conceivable, this could perhaps work. But the reality is, you are trying to base solid results off of skewed figures at this point. Furthermore, when a breathalyzer goes through a maintenance update to resolve their continuing AMB errors, is the new software just simply raising the allowed contamination number from .009 to something more to get rid of the AMB error? It would get rid of the machine stating the error, but the error would actually still be there, without being stated! We, the people, are not allowed to know what exactly is happening and how we are being tested when accused of a DUI.
Another source of an environmental problem is the Electrical power source. This is EMI. Various problems can occur when the breathalyzer does not have a solid uninterrupted power source. It should not share a power strip with other devices.
Another potential problem area can occur during the deprivation period. Fortunately, Kansas protocol requires that the subject be observed for 20 minutes and not allowed to put any foreign object in their mouth. If a foreign object, even water, enter the mouth during this period, the deprivation period is to be started over as it can affect the results. This is an important protocol and one which Kansas does typically enforce. I have seen breath test results suppressed by the Court because the Defendant took a drink of water 19 minutes prior to taking the breath test.
However, this protocol doesn’t always provide the protection it should. There are times when a suspect may have a little bit of chew in their mouth that goes unnoticed by law enforcement. That will skew the results and cause a false positive. Preventing this from occurring is important because the machine often cannot differentiate between the sources of ethanol. It will think it is providing a person’s actual alcohol content, when really it is reading the contaminates in the person’s mouth. Under a best practice, the breathalyzer operator should explain the issue to the person being tested, check their mouth fully, ask if anything is in their mouth, rinse out the mouth, and then start the 20 minute deprivation period.
Another instance that often gets overlooked is when the persons belches or burps during the 20 minute deprivation period.. While this is supposed to cause the deprivation period to start over, it often goes unnoticed, thus again causing a false positive. Supposedly, the machine has a “slope detector”, or Residual Alcohol Detection Systems to identify the presence of contamination from belches. However, this technology is not especially reliable. As a DUI defense lawyer, I have seen people burp, yet be able to do the breath test without error messages.
In order to provide more reliable and fair results, Kansas should adopt best practices for it’s protocol and requirements. It should require at least 2 tests. It should require at least an annual full maintenance and recalibration. It should truly ensure that the testing environment if free from contaminants as well as radio frequency. It sure demand that the air blanks are true .000 and not floating air blanks. It should make sure that the power supply is solid and not shared by other devices. It should make sure that the person’s mouth really is free from al foreign objects. Lastly, they should be transparent and allow access to the source code so that we the people can see what is really going in inside that breathalyzer computer.