Breath Issues -- Partition Ratio or Blood-Breath Presumption


	Print or Email This Page January/February 1999 Breath Issues — Partition Ratio or Blood-Breath Presumption By *Peter Gerstenzang* Peter Gerstenzang, the senior partner in the Albany, New York law firm of Gerstenzang, O’Hern, Hickey & Gerstenzang, specializes in the defense of DUI cases. A former prosecutor, he was certified as a breath test operator and taught for many years at the New York State Police Academy in their Breath Test Training Program. His book, Handling the DWI Case in New York (West Publishing Company 1987) is updated annually and is a standard reference for the defense of DUI cases. He is an acclaimed lecturer for the New York State Bar Association, the judiciary, police and prosecutors. This article is abstracted from his presentation at the NACDL Ultimate in DUI Defense seminar presented in Las Vegas on September 11, 1998. Excerpted from Handling the DWI Case in New York Reprinted with Permission of West Publishing Company. This month’s .10% Solution picks up from the November issue, which introduced methods for challenging the validity of the prosecution’s breath test results. November’s column described several general grounds for an attack on the prosecution’s evidence. The focus of this column is on the discrepancy between blood alcohol content (BAC) and breath alcohol content (BrAC). All breath testing instruments operate on the presumption that there is a mathematical relationship between the concentration of alcohol present in a person’s breath and the concentration of alcohol present in their blood. The vast majority of jurisdictions base their intoxicated driver legislation on blood alcohol concentration rather than breath alcohol concentration. Accordingly, we have the anomaly of breath test instruments attempting to determine blood alcohol concentrations. It is this anomaly which creates the most basic challenge to test results. All breath testing equipment assumes that the concentration of alcohol present in the person’s blood equals the concentration of alcohol present in 2100 milliliters of air. The determination of a blood alcohol concentration requires the breath test instrument to determine the concentration of alcohol present in 2100 milliliters of air. This partition ratio of 1 to 2100 is the product of an application of the law of physics known as Henry’s Law. While William Henry was not concerned about breath testing in 1803 when he stumbled upon this law of physics, it is his principle that makes the determination of a blood alcohol concentration from a sample of breath possible. Henry’s Law states that there is a relationship between a volatile chemical substance in liquid and that same volatile chemical substance in vapor; and that relationship, whatever it may be, is directly affected by changes in temperature. To illustrate, imagine a saucepan of water sitting on a stove in someone’s kitchen. Now imagine a plastic bag being placed over the top of the saucepan and the evaporate being trapped in the plastic bag. At room temperature, the evaporate would be invisible and there would be a mathematical relationship between the concentration of water present in vapor form, and the concentration of water present in liquid form in the saucepan. Accordingly, there would be a relationship between the volatile chemical substance in liquid and the same volatile chemical substance in vapor at a particular temperature. The thrust of Henry’s Law is that this relationship is directly affected by changes in temperature. By turning the burner on beneath the saucepan and heating the water to boiling, the plastic bag would soon fill with steam and the concentration of water in vapor form would have so dramatically increased that it would be visible where the evaporate was not. Once again, there is a relationship between the volatile chemical substance in liquid and the volatile chemical substance in vapor, and that relationship has, in fact, been directly affected by a change in temperature. If the relationship between liquid and vapor is directly affected by changes in temperature, you should be able to calculate a particular mathematical relationship at a given temperature. The premise of breath testing is that Henry’s Law is applicable to people, in that the volatile chemical substance in liquid is alcohol in the blood; and the related vapor is alcohol in the lungs which has percolated out of the blood. The mathematical relationship of 1:2100 is dependent upon the presumed temperature of air leaving a “normal” person’s mouth as being 34� C. At 34� C, the partition ratio between the concentration of alcohol in the blood and the concentration of alcohol in the breath is presumed to be 1:2100. This presumption, however, is not only open to challenge, it has been rejected by the scientific community. In State v. Burling, 224 Neb. 725, 400 N.W.2d 872 (Neb. 1987), the Nebraska Supreme Court reduced by half a test result obtained from an instrument, based upon this presumptive relationship. In this case, Dr. Norman Scholes, a research scientist and associate professor of pharmacology, testified that recent research had shown that the ratio was not fixed at 1 to 2100. Rather, it varied from one human being to another, and ranged from 1 to 1100 to a high of 1 to 3400. On cross-examination, Dr. Scholes was questioned as to the general acceptance of the 1 to 2100 relationship: Q: Now, with regard to acceptance in the medical community as to this 2100 to 1 ratio, is that still generally accepted in the medical community as the correct ratio? A: I don’t think it’s ever been accepted in the medical community as a ratio. It’s been accepted by forensic toxicologists up until the availability of new data. Q: And have forensic toxicologists adopted a new ratio or something different than 2100 to 1? A: No, they haven’t because it’s still so variable. What a lot of forensic toxicologists have recommended is abandonment of that ratio and some jurisdictions have accepted that. Q: As generally stated for forensic toxicologists in general, is the 2100 to 1 ratio still accepted by them? A: No, it’s not. Q: Is it accepted by some? A: None that I have read. Q: How would you determine the general consensus of the forensic toxicologists as to the acceptance of this ratio? A: I, generally, read the literature from the various laboratories that are internationally and nationally known. Q: So it was previously accepted by forensic toxicologists as the correct ratio? A: That’s correct. Q: And you’re saying it’s since been rejected? A: Yes. I’m saying— “rejected” is a good word. “Abandoned” is the best word. Q: And how has it been rejected or abandoned? A: Well, it’s no longer an acceptable—It’s no longer acceptable, this partition coefficient, for these types of analyses. Q: I understand that, but how did they officially reject the ratio? A: Well, aside from publishing their own results and opinions, which state that they should abandon it because it’s too variable, they have actually petitioned the federal government, and were successful in petitioning the federal government, to abandon it. Q: But they, themselves, have not abandoned it? A: Yes, they have. Q: And how did they do that? A: They just don’t do it anymore. They use it—They use the adoption of the Uniform Vehicular Code that says intoxication by law, will be so much alcohol in the expired air, or so much alcohol in the blood, but if you’re going to measure air, the intoxication would be based upon the amount of alcohol in the air. If you’re going to measure blood, intoxication would be based on the amount of alcohol in the blood. That’s what the forensic toxicologists are doing today. They’re abandoning the extrapolation air. Based on the testimony elicited, the Nebraska Supreme Court concluded: Even ignoring in this instance the effect of ingesting certain substances, the uncontroverted testimony of the pharmacologist compels the conclusion that if one’s breath-to-blood distribution ratio were at the 1:3400 level, the machine would understate his or her blood alcohol level, whereas if one’s distribution ratio were at the 1:1100 level, the machine would overstate his or her blood alcohol level. State v. Bjornsen, supra, ruled that a test result which is subject to a margin of error had to be adjusted so as to give the defendant the benefit of that margin. Thus, we are concerned only with the degree by which the machine may overstate defendant’s blood alcohol level. One having 1:1100 breath-to-blood distribution ratio has an actual blood alcohol level which is but 52.38 percent of the reading produced by the machine. Therefore, in this case the Intoxi-lyzer result must be reduced to 52.38 percent of .164 of 1 percent, that is, to .086 of 1 percent. Consequently, the Intoxilyzer test result fails to establish that defendant had “ten-hundredths of 1 percent by weight of alcohol in his blood” and therefore fails to establish a violation of � 39–669.07. State v. Burling, 400 N.W.2d at 876–77. It should be noted that the court did, in fact, affirm the defendant’s conviction based upon the other proof of intoxication elicited at the trial. 400 N.W.2d at 876–77. In State v. Babcock, 227 Neb. 649, 419 N.W.2d 527 (Neb.1988), the Supreme Court clarified its ruling in Burling, holding that each case had to be judged on its facts and that the court did not mean to enunciate a rule of law requiring reduction of all breath tests in every case. In State v. Burling, supra, this court did not intend to, nor did it, rule that as a matter of law a reading of the test results from an Intoxilyzer Model 4011AS should automatically be adjusted. In State v. Burling, supra; State v. Hvistendahl, 225 Neb. 315, 405 N.W.2d 273 (1987); and State v. Bjornsen, 201 Neb. 709, 271 N.W.2d 839 (1978), this court merely ruled upon evidentiary facts peculiar to each of those cases. Whether an adjustment is required is dependent upon the credible evidence in each case. State v. Babcock, 419 N.W.2d at 530. There is a wealth of articles published in regard to the partition ratio. Many experts assert that the average partition ratio is actually 1:2300. In her article, “Pharmacology and Physiology of Alcohol,” Id. Mary McMurray indicates that if this were the case, breath testing instruments would actually underestimate the blood alcohol concentration by approximately 9 percent. Id. at 6. In Edward F. Fitzgerald, Intoxication Test Evidence, 2d ed. (1995), the author cites Dr. Kurt Dubowski as asserting that the normal range of the partition ratio varies from 1 to 1100 to 1 to 3000. He cites the example of a person having a partition ratio of 1 to 1500 and an actual blood alcohol concentration of .07. That person would attain a breath test reading of .10 on a properly functioning breath test device. Conversely, a person with a BAC of .14 would attain a reading of .10 if they had a partition ratio of 1 to 3000. Edward F. Fitzgerald, Intoxication Test Evidence, 2d ed., � 28.1, at 28-3 (1995). This two volume set is an excellent resource in regard to intoxication test evidence. Fitzgerald also cites the testimony of Dr. Dubowski in Anchorage v. Serrano, 649 P2d 256 (Alaska App 1982). Here, Dr. Dubowski testified that at least 14 percent of the subjects he tested had blood-breath partition ratios of less than 2100 to 1. Accordingly, he recommended a safety factor of at least .025 be subtracted from breath test readings. Edward F. Fitzgerald, Intoxication Test Evidence, 2d ed., � 28.3, at 28-10 (1995). To the contrary, Dr. Michael P. Hlastala, who is an expert in lung physiology with the Department of Physiology and Biophysics and Medicine at the University of Washington in Seattle, asserts that the calculation of a partition ratio requires an equilibrium, which thus requires a closed container or system. Since people are not a closed container or system, it is impossible to calculate or presume a fixed partition ratio vis a vis blood/breath alcohol. In point of fact, the partition ratio varies from individual to individual. More to the point, it varies within the individual over the course of a day and is affected by a number of things — including alcohol consumption. Accordingly, this issue can be used very effectively in the right case. I emphasize the word “right” because this is not something you would want to raise in a case where the defendant is falling down drunk and the common law case is overwhelming. Insofar as the “wrong cases” are concerned, the falling down drunk is, perhaps, one of the most difficult cases to defend. Aside from corroborating the defendant’s obvious intoxication, the chemical test result is of little importance when contrasted with the physical evidence. Interestingly enough, Larry Taylor, an expert in the defense of DWI cases, suggests an all out attack on the chemical test, but for different reasons. Citing the wisdom of the great General Sun Tzu, Larry suggests using an attack on the chemical test to divert the prosecution from establishing their common law case. He quotes General Sun Tzu: If I dare ask: if the enemy is numerous, disciplined and about to advance, how should we respond to them? I would say: first, seize something that they love. . . . Lawrence Taylor, Drunk Driving Defense, 4th ed., at 8:163 (1996). Applied to the courtroom, his advice is to shift the focus of the prosecution on to defending their test result, rather than asserting their common law case. Accordingly, there will be times when even the “wrong case” will be the “right case” for a concerted attack upon the chemical test result. Body Temperature Defense The body temperature defense arises out of the partition ratio and the application of Henry’s Law. Essentially, the 1:2100 relationship depends upon the temperature of the air leaving the defendant’s mouth being 34� C. If the temperature exceeds 34� C, it will artificially and erroneously raise the test result obtained. Conversely, if the temperature of the air leaving the defendant’s mouth is less than 34� C, it will artificially and erroneously lower the test result obtained. This defense was so effectively utilized that in one jurisdiction, the police started taking the temperature of the defendant being tested. Using the Simulator To Establish the Body Temperature Defense The body temperature defense can be asserted without the benefit of an expert witness. Essentially, the breath test operator knows that the test result must be confirmed with a simulator test. A simulator is generally a portable drunk, consisting of a glass container holding a solution of alcohol and water whose concentration has been tested and whose vapor will produce a reading of .10 when blown into the breath test device. The simulator simulates a person taking a test and, therefore, the simulator solution is kept at a temperature of 34�� .2 of a degree. The operator has been trained to check the thermometer and not to utilize the simulator if the temperature exceeds the tolerance of � .2 of a degree. Most operators know that if the temperature exceeds 34 + .2 of a degree, a test run on a breath test instrument will be artificially high. Once you have the operator committed to the fact that the simulator simulates a person taking the test, and that an increase in temperature of as little as .3 of a degree will result in an erroneously high test result, you can then plug in the fact that the same principle applies to a defendant whose air sample exceeds 34�� .2 of a degree. You then establish that the police failed to take the defendant’s temperature and have no idea whether the test result they obtained is accurate or not. The problem with this defense is that an expert would testify that the artificial increase is approximately 6.5 percent per degree centigrade. In an article entitled Variability of the Blood: Breath Alcohol Ratio in Vivo, Professor A. W. Jones discusses the physiological variables which affect the determination of a blood alcohol concentration: The factors accounting for this variability include the physiology and dynamics of ethanol uptake into tissue and the kinetics of ethanol metabolism. Physiological factors influencing the breath alcohol concentration, in particular the variation in temperature of expired breath, contribute to the observed variations in the blood: breath alcohol ratio. A change of 1� C in the temperature of expired breath will change the breath alcohol concentration and hence the apparent blood: breath ratio by 6.5 percent. A.W. Jones, “Variability of the Blood: Breath Alcohol Ratio in Vivo,” Journal of Studies on Alcohol, vol. 39, No. 11, 1978. In Edward F. Fitzgerald, Intoxication Test Evidence, 2d ed. (1995), the author points out that the average human body temperature of 98.6� F is just that, an average which can vary by several degrees over the course of a day. He also points that different people have “different” normal temperatures: A rise in temperature of only 2� F results in an “apparent” 10% increase in the BAC due to the increased volatility of the alcohol. A corresponding drop may be produced in “apparent” BAC on a breath test with a heavy dose of aspirin, which tends to reduce body temperature. Edward F. Fitzgerald, Intoxication Test Evidence, 2d ed., � 28:1, at 28-3 (1995). Unless you have a marginal test result such as a .10 BAC, you are probably better off without an expert witness. Here, the failure of the police department to thoroughly train their operators can prove fatal to the state’s case. Where the defendant had a .15 BAC, a change of 7, 10 or even 20 percent is not going to affect the outcome. The vast majority of police officers, however, are not in a position to testify as to how much that effect will be, but only that there will be an effect. Mouth Alcohol Defense Mouth alcohol is the most common cause of error in breath testing. It is also one of the most effective defenses that can be asserted to challenge a test result. Essentially, breath test instruments assume the integrity of the sample of air being delivered to the instrument for testing. They presume that the air being trapped is the deep lung “alveolar” air that has traditionally been assumed to contain the alcohol concentration most reflective of that contained in the blood. This deep lung or alveolar air is captured by having the defendant blow into the instrument for a period sufficient to exhaust the defendant’s lungs. The last air blown into the instrument is deemed to contain the alveolar air sample requisite for an accurate test result. In the Breathalyzer, the process depends upon the operator observing the manner in which the defendant blows. The operator is supposed to note whether the defendant blows consistently and continuously to the point of evacuating his or her lungs. Further, the operator is presumed to have the ability to note and detect the presence of mouth alcohol resulting from a belch or regurgitation of alcohol bearing stomach contents. This is the purpose behind the 20-minute waiting or observation period. In 20 minutes, any alcohol that has found its way into the defendant’s mouth should have dissipated. Police officers routinely testify that they are able to fill out paperwork and perform routine administrative chores while keeping an eye on the defendant to ensure that the air sample is not adulterated by alcohol coming into the mouth via a belch or regurgitation. Anomalously, the defendant is rarely, if ever, informed of this danger. Accordingly, if the defendant has suffered the detriment of having been raised not to burp or belch in a noticeable manner, there is every likelihood that this can and will occur without being noticed by the breath test operator. Since the only effect of the presence of alcohol in the mouth is the erroneous elevation of the test result obtained, the vigilance of the observing officer is highly suspect. The defendant is never informed of this danger because of the fear that the defendant will commence a course of burping and belching every 20 minutes and, thereby, prevent the performance of a valid test. Police witnesses attempt to minimize the impact of mouth alcohol on the validity of a test. While they acknowledge the vulnerability of the instrumentation to mouth alcohol, they try to assert that the impact on the numerical reading is minimal. In that regard, it is important to know how this actually works. Twenty-one hundred milliliters of air is a relatively large volume. It is the rough equivalent of that contained in a two-liter soda bottle. There are no breath test devices with sample chambers big enough to accommodate this volume of air. Accordingly, blood alcohol concentrations are determined by taking a small proportionate sample of air and multiplying the result obtained in order to determine the alcohol present in 2100 milliliters of air. For example, the Breathalyzer 900 and 900A have a sample chamber which traps 56.5 milliliters of air and tests 52.5 milliliters of air. The result obtained is multiplied times 40 to determine the concentration of air in 2100 milliliters which is presumed to be the blood alcohol concentration. Seen in this light, any contamination of the breath sample is subject to being multiplied times 40 and can have a very dramatic effect. Our office uses an Alco-sensor which is a portable breath test device to demonstrate this effect to clients. The client will blow into the instrument initially and we will obtain a .00 reading (at least most of the time with most of the clients). We will then have the client ingest a tiny amount of Scope mouthwash, coat their mouths, and spit out the residue. We will then retest with the Alco-sensor and routinely obtain high readings. Demonstrations of the significant impact of mouth alcohol are a routine part of the training provided to breath test operators. It is a part of the training that virtually every operator remembers and reminding the officer of that training can be very effective in getting the officer to reconsider his or her assertion of the minimal effect of mouth alcohol. How You Blow Does Make A Difference Like so many other “facts” in law enforcement, the presumed physiology of the lungs and the consistency of the air contained therein is just not so. As mentioned earlier, Dr. Michael P. Hlastala has emerged as a beacon of light in this darkness of presumption. He has written articles and given lectures in regard to the physiology of the lungs in relation to breath testing. In an article entitled “The Alcohol Breath Test — a Review,” Dr. Hlastala points out that breath testing is based upon an understanding of pulmonary physiology which is outdated. Essentially, breath testing assumes that alcohol emerges from the blood, through an exchange taking place in the alveoli, in a manner similar to the oxygen and carbon dioxide exchange. He points out that recent physiological studies have shown that: [A]lcohol is exchanged entirely within the conducting airways via diffusion from the bronchial circulation. . . . The airway alcohol exchange process is diffusion (airway tissue) and perfusion (bronchial circulation) limited. The dynamics of airway alcohol exchange results in a positively sloped exhaled alveolar plateau that contributes to considerable breathing pattern-dependent variation in measured breath alcohol concentration measurement. (Hlastala, “The Alcohol Breath Test — a Review,” J. Appl. Physiol. 84(2): 401-408, [1998]). Succinctly — how you blow does make a difference. For example, a change in your breathing pattern immediately before delivering a breath sample can affect the result obtained: Hyperventilation for 20 s[econds] before performing the ABT causes an 11 percent reduction in BrAC. Three deep breaths before the sample breath reduce BrAC by 4 percent. After breath holding for 15 s. before exhalation, the BrAC increases by 12 percent (for a minimum exhalation) and 6 percent (for a maximum exhalation). A 30 s. breath hold before exhalation increases BrAC by 16 percent. These effects are caused by altering the ventilation (hyper-or hypo-ventilation) passing over the airway mucosa. Such data support the theory of airway surface interaction of alcohol as the mechanism causing BrAC to change during exhalation. Id. at 403 (footnotes omitted). Interestingly, a person with increased lung volume can produce an erroneously elevated test result by blowing for a prolonged period of time. The reason for this is that the breath has a longer period of contact with the mucosa lining the respiratory tract. Mucous has a high concentration of water and water will contain alcohol in a concentration closer to that of the blood. This is added and can have a dramatic effect since the blood has an alcohol level that is purported to be 2,000 times greater than that contained in the breath. Edward F. Fitzgerald, Intoxica-tion Test Evidence, 2d ed., �32:13, at 32-14-32-15 (1995). Slope Detectors In the newer computerized devices, so-called “slope detectors” have been installed which are designed to determine whether an air sample has been adulterated by mouth alcohol. Essentially, these devices monitor the concentration of alcohol in the breath sample as the person is blowing into the instrument. The presumption of a slope detector is that a valid breath sample will gradually increase as the air closest to the alveolar cells is blown into the instrument. This deep lung air is deemed to contain the highest and/or most accurate concentration of alcohol. The air in the mouth and throat is deemed to contain a lower and less accurate concentration. Accordingly, there should be a gradual and increasing slope of blood alcohol concentration as the person blows. The theory of the slope detector is that if there is mouth alcohol, the person blowing will start off with a high blood alcohol concentration that will slope downwards as the person continues to blow and the initial air containing mouth alcohol is expired. When the instrument detects a slope that is not in conformity with its expectation for alveolar air, it declares the sample invalid and voids the test. The problem with this is that a number of experts have determined that the slope detector doesn’t work. In Richard E. Erwin, Defense of Drunk Driving Cases, 3d ed., � 21-06, at 21-49 (1996), Dr. Harvey Cohen discusses the failure of breath testing devices to detect mouth alcohol. Erwin also reports a study performed by the Minnesota Bureau of Criminal Apprehension in which alcohol-free subjects produced readings ranging from .11 to .24. Again, these test results were obtained with sufficient samples and with a slope detector operational. Similar results were obtained by Rick Swope, who is an adjunct professor for the University of North Florida, Jacksonville, and for the Institute of Police Technology and Management, as well as lead DUI instructor at the Broward Criminal Justice Institute, Fort Lauderdale, Florida. The results of his study are set forth in the September 1995, DWI Journal, in an article entitled “An Intoxilyzer 5000 Primer.” Here, Swope conducted over 100 tests with respect to the detection of mouth alcohol and the Intoxilyzer 5000R machine. Individuals who had consumed no alcohol were given one ounce of alcohol to rinse out their mouths. After spitting out the alcohol, the participants waited for periods ranging from 5 to 38 minutes, prior to providing a breath sample. In most cases, the Intoxilyzer 5000R machine gave “valid” readings indicating that the person had consumed alcohol. More importantly, in most cases, no warnings of mouth alcohol were present. Similarly, the Intoxilyzer 5000R machine was unable to distinguish mouth alcohol on those participants who were wearing dentures and who had waited for periods of up to 50 minutes prior to giving a breath sample. According to Swope: The unreliability of this detector and the insufficient training given breath operators today lead to disturbing conclusions. Id. at page 8. In the Drinking/Driving Law Letter of May 10, 1996, Clark Boardman Callaghan, Vol. 15, No. 10, Dr. Hlastala discusses slope detectors and their vulnerability to mouth alcohol: If mouth alcohol is present in small quantities, it will add to the normal exhaled profile of alcohol coming from the blood and alveoli. The result is that a small amount of alcohol (and its decreasing concentration) will be added to the normal exhaled profile (and its increasing concentration) with a net effect of producing a constant (or slightly increasing) alcohol concentration. The slope detector will not work correctly when alcohol is present both in the blood and the mouth. The measured BrAC will be higher than it should be. However, the slope detector will not detect the contribution of mouth alcohol. . . . The slope detector, as used on current infrared alcohol breath testing machines, does not carry out the intended functions. The fact that different exhalation volume results in different BrAC values demonstrates that the “alveolar plateau” function has no purpose. There is never any guarantee of “alveolar air.” In fact, the physiology of the lungs guarantees that alveolar air cannot be sampled (for alcohol concentration) as it was when in the alveoli. The observation that the mouth alcohol function only works when there is no alcohol in the blood stream argues that mouth alcohol may be contributing to high BrAC values up to a 0.03 to 0.04 error (in the increasing direction) on any given test. Id. at 156-57. Conclusion This article sets forth only a few of the things that can be used to impeach a breath test result. The literature in regard to this subject presents a vast array of things that can and do go wrong with breath testing. That array, however, should be critically evaluated to determine those challenges that are most congruent with your theory of the defense and with the common law facts of your case. .10% Solution is devoted to providing practical information on the defense of DUI/DWI cases. Authors and ideas are welcome. Please contact: .10% Solution Jim Bell or Steve Oberman 110 W Summitt Hill Dr Knoxville TN 37902 Phone (423) 637-2900 Fax (423) 971-4298
	National Association of Criminal Defense Lawyers (NACDL) 1660 L St., NW, 12th Floor, Washington, DC 20036 (202) 872-8600 • Fax (202) 872-8690 • assist@nacdl.org