More than twenty years ago (in 1986) Edward J. Imwinkelreid, Professor of Law at the University of California, Berkely, CA,( 1) in an article on science and law made the following observation: "Lawyers, judges, and juries are weighing the scientific merits of theories and techniques they may find strange and confusing. Today, science is being distorted and the legal system is suffering for it."

This is still true today (October 15, 2007 ) where the science involved in prosecuting drivers for allegedly driving under the influence of alcohol (DUI) is seriously distorted. This review will identify some of the faulty scientific distortions and legal difficulties present in today's approach to the drinking-driver problem. In the United States the advent of the automobile following the end of the prohibition era so increased traffic fatalities that statutory control of the drinking driver became a necessity. Mason and Dubowski (2) present a very good resume of the events leading up to intervention by the federal government in determining that any individual having a blood alcohol concentration (BAC) of 0.15 % W/V (0.15 g/dL) or higher was guilty of driving under the influence of alcohol. The federal statute implementing the proscribed BAC essentially made it mandatory for the states. Failure of the states to enact the DUI legislation would result in the loss of federal funds for repair and construction funds for its highways. Several months prior to the enactment of the federal legislation the electorate of the state of Ohio had rejected enactment of a state DUI law. The Ohio legislature had to enact emergency legislation to meet the August deadline and, thus, avoid loss of its highway funds.

It is important to recognize the full meaning and extent of that initial proscription. The determination of a BAC is a measurement of the concentration of a chemical, alcohol (alcohol as used herein means ethanol), in the blood of the driver at a time when the driver is suspected of being impaired. Before going into the analytical problems it is necessary to recognize the difficulties arising from the lack of definitions of two critical items: blood and impaired. The statutes do not define blood. Blood may be venus (in the aorta or in other conduits leading to the heart) or arterial (in the veins or other conduits carrying the blood away from the heart to all parts of the body) or capillary (in the minute vessels connecting the arterial and venous conduits. The alcohol concentrations measured at the same specific time for arterial, venous and capillary blood may vary significantly. Arterial blood will have a higher alcohol concentration because the alcohol has not been distributed throughout the body. Collection of arterial blood samples are more difficult and more dangerous than collection of venous or capillary blood samples and are not done routinely for forensic DUI. When a micro-procedure for the analysis of blood is used, capillary blood is commonly obtained by the "finger-stick" procedure. Obviously, in addition to the variable content of alcohol in arterial, venous or capillary blood, there are added pre-analytical variables due to the different sites and collection procedures.

Standard clinical laboratory procedures routinely begin with the collection of venous blood samples. Forensic ethanol determinations, likewise, begin with a venipuncture and the collection of a blood sample, but as the result of a long-standing clinical laboratory nomenclature practice a serious problem arises. For many clinical analytes, blood is obtained without anti-coagulants, is allowed to clot, and is, then, centrifuged to obtain serum. Proper reporting of the test result should indicate that it was a serum concentration not a blood concentration. For many common clinical analytes such as glucose, cholesterol, urea and others, the difference between the serum and blood concentration is minimal; however, that is not true for alcohol. Serum alcohol concentrations may be up to 20% higher than blood alcohol concentrations. Serum alcohol concentrations (SAC) will always be higher than BAC because alcohol does not readily penetrate the red cell membrane. Since the red blood cells are removed by centrifugation of the clotted blood, the volume(or space) occupied by red cells will be taken up by fluid containing alcohol. Thus an inaccurate higher BAC occurs. Many laboratory reports do not make the distinction between serum and blood and, thus, report an inaccurately high BAC. Many times the police will use a sample collected for the hospital laboratory (one collected without anti-coagulant) and thus get an inaccurate high test result. Unfortunately, some laboratory directors do not recognize this inaccurate procedure. A Pennsylvania court order (3) was necessary to prohibit a PA laboratory director from reporting serum alcohol concentrations as blood alcohol concentrations. Additional evidence of this distortion is that many laboratories have used a ratio from 1.09 to 1.35 to convert a measured serum alcohol concentration to a presumed blood alcohol concentration. The use of a ratio (presumably derived from a series of tests and averaged) results in a post-analytical variable error because the original analytical result has been converted into a number the accuracy of which is not known.

This difference between a BAC and a SAC becomes more important when the problem reaches the courtroom and the judges have to decide which is correct. Frajola (4) reported two egregious examples of pseudo science given to jurors to ponder. As an expert witness for the defense in an Indiana case, he brought a test tube rack of samples to show to the jury. One test tube contained an un-clotted sample of blood (an anti-coagulant was present in the tube when the blood was collected). A second tube contained plasma obtained by centrifugation of un-clotted blood. A third sample contained clotted blood. The fourth sample had serum obtained by centrifugation of clotted blood. Because the red blood cells forming the clot do not contain alcohol, examination of the clotted tube would easily explain why serum AC is greater than blood AC. The evidence had previously shown that serum had been analyzed, but blood alcohol concentration had been reported. The judge would not allow the jury to see the difference as would have been evident had the jurors seen the four tubes. The second courtroom experience is more difficult to understand. A West Virginia judge in the absence of the jurors listened to the prosecutor's expert, the Ph.D. clinical chemist in charge of the laboratory, describe how the test sample was analyzed in a DuPont Automatic Analyzer. He agreed that the blood was obtained without an anti-coagulant, that the DuPont instructions clearly stated that the result was a serum alcohol concentration. He argued that the result was correctly reported as a blood alcohol concentration. The defense expert witness explained that a sample centrifuged without anti-coagulant present would result in a serum sample. Because alcohol does not readily penetrate the red cell membrane, a serum sample would have a higher alcohol concentration than a sample with red cells present. The judge, faced with two conflicting opinions, decided that only the numerical result without mention of the units of concentration and without any discussion of whether the test was done on serum or blood would be given to the jury. Imwinkelreid was certainly correct when he noted that lawyers, judges, and juries were weighing scientific techniques that they find strange and confusing.

Another major distortion of science in forensic alcohol determination results when a numerical concentration for the alcohol, (a measure of the amount of a chemical in the blood) is equated to the effect of that chemical upon the individual. Every person regardless of age, sex, previous driving experience, acute or chronic adaptation to alcohol, and biochemical individuality is deemed an impaired driver based only upon the content of alcohol in the blood.. The phenomenon of acute alcoholic adaptation was demonstrated as long ago as 1919 by Mellanby (5). He noted that in episodes of alcoholic consumption three phases were recognizable: absorption, peak, and elimination. A graph of the BAC versus time as shown in Figure 1 reveals that the same BAC occurs twice , once during the absorption phase and again during the elimination phase. The effects of alcohol during absorption were more severe than during elimination although the BAC was the same. The effect of a given concentration of alcohol may vary considerably. The difference was presumed to be due to the time interval between the two measurements allowing for acute adaptation to higher levels of alcohol concentration and for short term effects of the higher BAC between measurements.

Figure 1 The Mellanby Effect

Several examples of chronic adaptation to alcohol have been reported, but the example reported by Hammond et al (6) in 1973 is a very striking example of chronic adaptation. A 23-year old, white female was admitted to the emergency room of the Colorado Medical Center, Denver, CO in a coma after an automobile accident. Her BAC was 0.78 g% which is well over the generally accepted fatal concentration of 0.45 g%. She was treated with intravenous fluids and other procedures for a few hours and was discharged after 11 hours as fully coherent, completely normal neurologically, and with no evidence of intoxication. Her BAC at discharge was 0.25 g%! The validity of the gas chromatographic BAC determination was established by simultaneous assays of standards and controls. During her treatment she related that she consumed one bottle of bourbon in a two hour period without food since the previous night. She, also, stated that she had a history of alcohol abuse since she was 13 years old, and that she had been discharged from an alcohol rehabilitation center after two months of treatment. Hammond commented that the lack of correlation between BAC and the patient's clinical condition indicated a high degree of central nervous system tolerance to alcohol.

It is important to recognize the meaning of being impaired due to alcohol. The BAC is not a measure of the driver's impairment. Dubowski (7) has published a table identifying various clinical signs and symptoms associated with seven stages of alcoholic influence and a BAC range for each stage. Each stage overlaps the following stage and the signs increase in severity. The 0.01 to 0.05 g% stage is the sub-clinical stage with no apparent influence by alcohol but increasing to slight changes in behavior detected by special tests. Euphoria, the second stage, from 0.03 to 0.12 g% shows a maximum effect as a loss of efficiency in finer performance tests. The third stage, excitement, from 0.09 to 0.25 g% lists drowsiness and sensory motor in-coordination and impaired balance as its signs. Confusion, the next stage from 0.18 to 0.30 g% is said to add apathy and lethargy to the previous conditions. Finally, stupor, coma, and death are the last three stages with the latter occurring at 0.45 g/dL. Please note, firstly, that the overlap between stages is not insignificant, and secondly, that a given BAC can produce differing clinical signs and symptoms in different individuals. E.P.M. Widmark,(8) the internationally recognized scientist honored for his achievements in alcohol research, described his studies of individual variations in tolerance to alcohol in his book. In 1932 he wrote: "It is well known that different persons are affected to different degrees by the same alcohol consumption. For this reason it has been questioned whether the alcohol concentration of the blood can be considered a measure of intoxication."

Dubowski's tabulation mentioned above is means of showing that different persons are affected to different degrees by the same alcohol consumption. The BAC, alone, does not define intoxication nor impaired driving. In general, the statutes require that the driver should have normal control of his vehicle and not exhibit any signs of being under the influence of alcohol. The officer's observations indicative of impaired driving include speeding, or not stopping on a red light, or weaving in the lane, or some other minor infraction of the state or local regulations. After the driver has been stopped, the officer conducts what are known as field sobriety tests such as walking heel-to-toe, standing on one leg, or the finger-to-nose test. During the performance of the tests, the officer is carefully watching for any misstep, or erratic performance. The driver is marked pass or fail not on a percentage of correct procedures but only whether the driver exceeded a certain number set for each test. It is doubtful whether many teachers grade their test papers in such a fashion.

Roger Williams(8), Professor of Biochemistry at Louisiana State University, in his book, "Biochemical Individuality" had as his main thesis that people differed from one another. He identified anatomic, physiologic, and biochemical variations among individuals and showed how nutritional and environmental experiences accounted for these differences. Frajola (9) studied enzyme patterns in families with identical twins and noted that the patterns for the siblings and parents differed noticeably from one another, but that the patterns for the identical twins were strikingly similar. Reed (10) in his report, "The Myth of the Average Response to Alcohol" reviewed studies of seven responses to alcohol (changes in pre-baseline to post-baseline) for heart rate, systolicand diastolic blood pressure, automobile driving ability, Romberg balance, finger-to-nose and speech clarity. He indicated that the variations were sufficient to deem the average response a myth. One of the studies he reviewed was that of Laves (11) who examined the reports on 5000 German motorists who were arrested for DUI. Of the 5000 more than 1800 had BAC's of 0.10 g/dL or more. Ninety percent of those with a BAC from 0.10 to 0.15 g/dL passed the speech clarity test, 80 % passed the finger to nose test, and 50 % passed the Romberg test. At the higher BAC's from 0.25 to 0.30 g/dL the passing percentages were 70 %, 50%, and 10% respectively. A summary of the above data supports the conclusion that it is erroneous to believe that alcohol affects all individuals similarly or that all individuals are equally affected to the degree that they are incapable of safe driving. Is it reasonable to expect that a truck driver who has driven thousands of miles per year for several years will be affected by alcohol to the same extent as the driver who has hardly driven his new car a few hundred miles? How does one explain why a truck driver whose blood alcohol concentration is 0.02 g/dL is guilty of DUI when driving his truck, but is not guilty if he is driving his automobile? How does one explain that in 1969 0.15 g/dL was the proscribed limit for the BAC, that it was reduced to 0.10 g/dL a few years later , and that it was reduced to 0.08 g/dL more recently without supporting data that the previous limits were erroneous?

Edward R. Murrow, the famous author and radio news analyst, once observed that a great battle was being fought against ignorance, intolerance and indifference. Today's battle with highway safety can be similarly characterized. Ignorance of the serious effects of alcohol upon the mental and motor functions of the driver results in drunk drivers. Intolerance to alcohol produces varying degrees of driver impairment. Indifference to the ever-rising costs and the human suffering caused by drunk drivers has created serious problems, but the problems have been compounded because faulty science has been used to obtain the desired goal.

1. Edward J. Imwinkelreid, "Science Takes the Stand: The Growing Misuse of Expert Testimony" The Sciences, Nov./Dec. 1986, pp.20-25.
2. M. F. and K. M. Dubowski, "Breath-Alcohol Analysis: Uses, Methods, and Some Forensic Problems---Review and Opinion" J. Forensic Sciences, 21, 9-41( 1976).
3. Commonwealth v. Bartolacci, 598 A 2d 287 (Pennsylvania Super. 1991.
4. Walter J. Frajola, "Blood Alcohol Testing in the Clinical Laboratory: Problems and Suggested Remedies" Clin. Chem. 39/3, 373-379 (1993).
5. E. Mellanby,"Alcohol: Its Absorption into and Disappearance from the Blood Under Different Conditions" Medical Research Committee, Special Report Series. No. 31 (1919).
6. K. B. Hammond, B.H. Rumack, and D.O. Rodgerson, "Blood Alcohol--A Report of Unusually High Levels in a Living Patient" J.A.M.A. 226 (1), 63-64 (1973).
7. K. Dubowski, "Stages of Acute Alcoholic Influence/Intoxication".
8. Roger J. Williams, "Biochemical Individuality" John Wiley & Sons, Inc. Hoboken, N.J. (1956).
9. W. J. Frajola, E. Meyer-Arendt, and J. Waltz, "Serum Enzymes and Biochemical Individuality" Fed. Proc. 19, No. 1, PT. 1 (1960).

Those wishing to respond, question or add relative information are urged to contact me by email.