All Content ©Copyright 2004-05 Charles L. Sifers | Site Map | Contact Information: Office Phone - 405-232-3388 E-Mail:

Breath Tests

"Real Evidence or Just Assumptions"?


In criminal law, the constitutional requirement before you can be convicted is "beyond a reasonable doubt". No person is to be convicted for DUI or APC, unless the State can prove your were under the influence of alcohol. Breath test result will be used as part of - if not the MAJOR PART - the evidence against you to prove this.

A. Evidence?

The State will present the test result as irrefutable proof of your intoxication, both in negotiations and before your jury. This is not - either at law or in fact - irrefutable. If they plan to use it as evidence, then make them prove it! When you call their hand, you will find that their evidence is grounded in a whole bunch of assumptions that they will be hard pressed to "prove".

The State will usually bring the testing officer and the maintenance supervisor of the machine to prove this test. Your testing officer will testify that he knows only how to give the test, pursuant to the appropriate Rules of the Board. He will not admit he knows more than that. He is trained to only to testify to that extent when he is in the operator's school. It is actually in their training manual! Their concern is that you will attempt to qualify him as an "expert". However, take him further.

To be certified and qualified to test your client, many of your older operators were likely first qualified on the 900 series machine. In total, to be within the standard set out by the National Safety Council and NHTSA, the training(s) SHOULD approach 40-45 hours of classroom instruction and practicum. Pushing the button and preparing the forms in a test was learned within an hour. What did he do during the remaining part of his course?

This training involved how IR works, alcohol pharmacology, it absorption and elimination, and alcohol/BAC calculations based on the Widmark Theory. In short, he knows a great deal more or he would not have passed, (IF he passed) the training course. If he squirms or begrudgingly agrees with your points, he looks like he is trying to hide some thing (which he is).

The supervisor will testify he checked the machine, changed the simulator solution and the machine was working properly. He, too, is warned in his training not to go beyond this. If your attorney obtained a Log of Tests of your test (which he should), he should make the officer tell him what he meant when he wrote "thoroughly cleaned and checked" on the log. You will discover it means he probably wiped the dust off the top of the unit!

He should make him testify as to WHAT the officer did - or did not - do to maintain or service this machine. He should listen very carefully and compare that to his instructions/requirements as a maintenance supervisor. The attorney should have already subpoenaed his supervisor's training manual and have it at his table. He can use it against the officer when he is cross-examining him.

The officer is also more knowledgeable than he would want you to know. Your attorney should .make him admit that certain safeguards exists in the machine. The officer will be pleased to do that. Then when asked what proof he has for the jury that the safeguards were working, he will not have any! He will say that the machine was accurately reading alcohol because the simulator solution was read to be within +.01. When asked from where that solution came, he will say from the approved source by the State. When asked if the documents showing that are available, he will tell your attorney that he threw them away when he did this servicing. Again, no evidence.

The State is expecting you and your jury to accept a whole bunch of assumptions as fact! My office DOES NOT allow this to occur.

B. Assumption(s)

The State is asking your jury/fact-finder to accept this machine's test result as accurate and reliable. This "acceptance" requested is based upon several assumptions, not necessarily facts which the D.A. can be prove. They want everyone to NOT question this machine. "Just believe us." However, this test's accuracy is based on assumptions which, if your attorney has the tools, can assist you in showing reasonable doubt to your jury.

What follows are several - though not ALL - examples of these "assumptions" on which the RIGHT attorney can the shine the "truth light".

1. Everyone is "average"

Probably the biggest assumption that is pushed off onto the fact-finder is that this machine assumes everyone that blows into it is average: average temperature; average blood/breath partition ratio; already absorbed and eliminating the alcohol; average elimination speed and process, etc, etc. The ENTIRE testing system in machine - the processor, the calculations, and the software - is based upon this assumption of these averages and your prosecutor does not want your jury to hear about this. No individuality or physical differences are accounted for. Remember though, if you "averaged out" a person by all those that had ever been on this planet, you would find that the person is Chinese, female, and dead.

2. The machine's temperature controls were accurate

The machine must be at a temperature to eliminate all condensation. This elimination starts, as indicated above, at the mouth piece. If it does not, the reading will be higher.

The State’s witnesses will testify that the machine is smart enough to let everybody know if that problem exists. Unfortunately, they are dead wrong. As a study by Dr. Richard Jensen - in which I was involved in the initials studies in Atlanta in 1996 - showed, you can completely remove the breath tube (and its heating tape) and the machine will show "diagnostic OK". Also, is the fan inside the machine functioning properly? This , too, can be disconnected and this "smart" machine will not know it. This fan is intended to regulate temperature. If it becomes too high, a false reading can also result.

In fact, many states require the officer giving your client his test to check to see if both the Intoxilyzer and the Simulator are at proper operating temperatures. He MIGHT recall that the simulator is supposed to be at a + .02E of 34E C. He will NOT KNOW what the proper temperature is of the Intoxilyzer OR whether it was at any specific temperature at the time of the test! They NEVER check the thermometer, or the actual temperature by independent validation, of the vapor coming off the simulator.

3. There was no RFI affecting your test

It has been long known that radio frequency interference ("RFI") affects the processors of these types of machines. It will cause the reading to be unreliable. If this feature is working properly, RFI will not interfere or cause an inaccurate test. However, was it working? There is a RFI antenna plugged to the side of the machine which runs up the breath tube. There is also one that is inside the machine. You can unplug either of these and the machine will not know it. If the breath tube has been pulled or stressed by the subjects prior to your test, how do you know it was working? The diagnostics check will not tell anyone. The supervisor does not check this with his "maintenance" as they are prohibited for opening the machine! Further, the "safeguard" is NOT on the list of items to be inspected by a maintenance supervisor.

Moreover, when was this function checked or calibrated to see if it works even if it is plugged in?? Many states do not have their own CMI, certified repairman and these machines are sent back to CMI. Some of these machines are only serviced/repaired once every two (2) years!

4. There was no Mouth Alcohol in the result

     a. Deprivation/Observation period

To insure that no mouth alcohol is present, the testing officer is required by the Rules of the Board of Tests to give a deprivation period of at least fifteen (15), proceeded by checking your mouth for foreign objects like gum, tobacco, or false teeth, before giving him a test. If properly done, this could eliminate SOME of this concern. It is assured that he will testify that he did this! If you press him, he will begrudgingly admit that these foreign objects could effect the test, too.

The officer should be asked how he knows if a mouth alcohol detection has occurred and what he is trained to do IF it does. This is ESPECIALLY important if you see on your test result a "sample rejected" printed out. This is the machine's warning to the officer that mouth alcohol HAS been detected. McBeth Sample, the Director of the Sifers' Board of Tests, says to start the deprivation period OVER again as mouth alcohol - and the REAL likelihood of a falsely high reading - is DEFINITELY present. You can be assured in almost every case that our officer/operator DID NOT RESTART THE DEPRIVATION PERIOD!!! Any test thereafter would be suspect.

     b. Slope Detector

This is their fall-back position concerning mouth alcohol. The machine reads alcohol continually as it is blown into the chamber. Since the alcohol in the lungs is at a higher concentration deeper into the lungs, the amount of alcohol being read should continue to rise until it plateaus. If it rises rapidly and then falls before the end of the blow, the slope detector activates and the machine indicates "sample rejected" or "invalid sample". In other words, mouth alcohol. The witness will tell you that this will catch any mouth alcohol, and as no such "prompt" was indicated on your client's test, the deprivation period was successful. He is wrong.

At best, the slope detector works the PART of the time (but certainly not all of the time) when the subject has mouth alcohol only, with no alveolar alcohol (deep lung) present. That is how this feature is tested (when it is tested!). It will not work most of the time when there is a combination of alcohol sources! In other words, if you silently belch during the deprivation period and you does have some alcohol in your system, the result can be an inaccurate - and falsely high - test . . . and a possible conviction!

Also, when was this last tested at all? The likely answer is it was tested ONCE when the machine came from the company and, maybe, once when the machine was installed. The regular maintenance does not check this feature. No one the State puts on will likely be able to answer this question satisfactorily!

5. The simulator verifies the machine's accuracy

The maintenance supervisor obtains a solutions from where ever he is approved from and puts it into his simulator for a new maintenance cycle. These containers have an amount of alcohol that is supposed to be in this solution. The documentation is almost always trashed. This solution is not independently tested prior to use. The supervisor pours this into the simulator which heats the solution to 34E, + .2E C. This is supposed to release the correct amount of alcohol from the liquid to be sucked out by the 5000 and tested. A recent study from Alabama's Board of Tests suggests that it must be not less than 35EC to be accurate, and any test on any machine which bases the partition ration at a 34E is inaccurate!

6. Your body’s temperature was 98.6EF

This may be the most important assumption for the State to "sell" your jury as to the machines accuracy. The machine assumes that your client is "normal" and, therefore, the gases from his/her lungs are at the expected - and what the makes of this machine contend is - normal, or 34EC.

The 5000 assumes everyone is at this body temperature in its calculation(s) of alcohol in the sample chamber so that the lungs' release of alcohol is consistent and the machine's software can be used to predict the level therein. Even a rudimentary medical education dispels that assumption. That might be the average for most people, however, this varies during any given day!!

This is significant in that the machine's calculations are based on a certain release of alcohol from the blood into the lungs. This released alcohol is greater at higher temperatures in your client and will result in an inaccurate test, i.e., for each 1EC over this average the reading will be 6 - 9% high than actually in his system!

7. Your breath alcohol is the same as his blood alcohol

This assumption is also critical to the State and the most important for them to convince you and your jury that this test is accurate! It is, however at the same time, one of their weaker points. The machines calculation(s) of alcohol in the sample and the conversion to g/210L is based on a 1:2100 breath/blood partition ratio. It can not convert your breath sample into a measurement without this assumption. Further, the simulator's function is similarly calculated at this partition ratio.

This partition ratio average is closer to 1:2300 according to former Director of Oklahoma’s Board of Tests and internationally known researcher, Dr. Kurt Dubowski. As such, the machine's function at 1:2100 is in your favor. However, 99% of the population has a ratio ranging from 1:1555 to 1:3005. What is yours? You don’t know. It is assumed by the machine, though, that you are "average".

If however, you are at the low end of this range, the result of the test will be higher than actually in your blood! For instance under the older .10% laws, if his sample measured and calculated at 1:2100 by that machine is .10%, but his ratio is actually 1:1800, his actual blood alcohol level was only .085%! For a multiple offender, this is the difference between a felony DUI with a jail potential of up to ten (10) years OR only a misdemeanor DWI with a maximum jail sentence of six (6) months!!

Moreover, with any new aggravated DUI statute (.15% or higher etc), you might be required to complete in-patient treatment or other countermeasure for a problem that you DO NOT HAVE!

What is MORE frightening is how this ratio varies and vacillates during the absorption of the alcohol into your client's body. As the alcohol rises and before it "peaks" in an individual, the ratio will vary constantly. Further, the ratio is HIGHER during absorption versus at peak or elimination from the body. This can lead to a GROSS over - or under - calculation by the 5000 of your client's alcohol level at different times in his drinking experience.

8. Any test variance is acceptable and the test is still reliable

In Oklahoma, we have a two (2) sample test rule with a variance between the two (2) test results that it considers "okay". Our state uses a + .03 variance. That means, at the "magic" level of .08, that this machine gives a test result - which could cost you not only your freedom, but your livelihood, your self respect, and place a mark on you for life - that could be as much as 40% to 60% inaccurate (.05 to .11) as to precisely what amount of alcohol you ACTUALLY had in your system at the time of test!!

Apply some common sense to this. Apply this same "okay" to the altimeter of the plane you just flew in? What if you find that it is "off" by being only 40% higher, OR maybe lower, than the pilot's gauge reads? Want to fly in that same plane again? The list goes on and on and is limited ONLY by your imagination.

9. The machine does not mis-read other substances as alcohol

a. Alcohol specific

First, understand that this machine does not measure alcohol at all. It measures "shadows" caused by unknown molecules in the light beam in the chamber and predicts alcohol from that. It does not know whether this "shadow" was caused by alcohol or some substance molecularly similar!

b. Other substances are not mis-read

This assumption, promulgated by law enforcement and the various makes of IR breath test machines, is BS and they know it. This machine does convert "shadows", caused by various chemicals that may end up in that sample chamber, into an alcohol reading! Why would the company add filters to remove these substances if it did not?? In fact, the newer machines have five (5) filters in the wheel which remove acetaldehyde (which is the first by-product of ethanol metabolism) and toluene. OKLAHOMA ONLY HAS THE BASIC THREE (3) FILTERS. THEY CHOSE TO NOT INCLUDE THEM IN OUR NEW MACHINES!!! Therefore, if you are a painter or work around solvents, your reading could be part alcohol and part something else, and incorrectly high.

In fact, a test could be failed with no alcohol in the person's system. I have had two (2) clients where this was the case. Based on our calculations, neither could have had any measurable alcohol in their system at the test time but registered greater than .08%. In both instances, each was exposed for several days to various solvents which my experts showed that the 5000 mis-read as alcohol.

10. The warranty of the machine

Lastly, have you considered the warranty offered by the company that makes it? It expressly disclaims that it is warranted for any particular purpose. It does not even warrant that it tests alcohol in gases!

III.  The NEW Machine that is coming - The Intoxilyzer 8000

By staying vigilant in keeping current - and on the cutting edge of the machines used in evidence against our clients - we  became aware a few years ago (before it was ever marketed by the company) of a new device made by CMI called the "8000".  This is CMI's new toy and is being used in several states in the Nation now.  Allegedly, this device has more "protections" against false positives and other errors that are typical with the Intoxilyzer 5000.  It uses an ADDITION filtering system to remove false readings caused by OTHER interferents.  (Go to the "BREATH TEST:  Evidence or Assumption" link on the Breath Testing Page of this site for more on this problem with breath testing's reliability).

We have become aware that the State of Oklahoma will begin using this machine at the end of 2006 or the beginning of 2007.  So as to be AHEAD of the State and it's witnesses, our attorneys will be attending a cerftification course in September of 2006 on the Intoxilyzer 8000.  When it gets here, we'll be ready to defend our clients unlike ANY OTHER LAW FIRM - just like we did and have be doing on 5000 breath test cases since 1996 - on cases with this new machine, the Intoxilzer 8000.


Back to Top

Home     Crime-Punishment     Breath Test     Choosing an Attorney    


Who Is Charles Sifers    


Free Evaluation     Find Your Courthouse     DUI: Constitutional Exception?


Success Stories         Helpful Links


Site Map - Thank YOU