560 So.2d 825 (1990)
STATE of Louisiana
v.
Robert M. HONEYMAN.
No. 89 K 1942.
Supreme Court of Louisiana.
April 30, 1990.
Dissenting Opinion May 2, 1990.
William J. Guste, Jr., Atty. Gen., Paul Carmouche, Dist. Atty., Catherine Estopinal, Asst. Dist. Atty., for applicant.
W. Eugene Golden, Hall & Golden, Shreveport, for respondent.
Dissenting Opinion of Justice Dennis May 2, 1990.
MARCUS, Justice.[*]
Robert M. Honeyman was charged by bill of information with vehicular homicide in violation of La.R.S. 14:32.1. He filed a motion to suppress the results of a blood alcohol test performed on a blood sample drawn from him after his arrest. After a hearing, the trial judge denied the motion. After a jury trial, defendant was found guilty as charged. Subsequently, he was sentenced to serve two years in prison at hard labor. He appealed his conviction and sentence to the court of appeal raising thirty-six assignments of error. The court of appeal, finding that the results of the blood alcohol test were wrongfully introduced into evidence (Assignment of Error No. 1), reversed defendant's conviction and sentence and remanded the case to the district court for a new trial. The court of appeal did not address the remaining assignments of error.[1] On the state's application, we granted certiorari to review the correctness of that decision.[2]
On October 20, 1985, defendant was involved in an automobile accident which resulted in the death of a passenger in defendant's car. The court of appeal described the accident as follows:
This one vehicle accident occurred during the early morning hours.... The defendant was driving a 1975 Cadillac with two guest passengers when the car struck the right girder of the Old Blanchard Road bridge in Caddo Parish. The impact caused the collapse of a span of the bridge. The car landed on its roof and came to rest approximately 100 feet from the initial point of impact. The front seat passenger, Reginald Stanley, died from cardio-respiratory failure when his neck was hyperflexed forward, dislocating his spine and lacerating or severing his spinal cord.
Defendant was injured in the accident and taken to a hospital. Less than three hours after the accident, a sample of defendant's blood was drawn at the hospital by a blood technician. The day after the accident, the blood technician gave defendant's blood sample to a chemist. The next day, the chemist performed blood analysis on the sample using a gas chromatograph. The analysis showed defendant's blood alcohol level to be 0.16%, which is above the level that is a condition of vehicular homicide under La.R.S. 14:32.1 A(2).[3]
The blood technician who drew the blood worked for the North Louisiana Criminalistics Laboratory (crime lab). He was trained and certified to draw blood by labs at the Veterans Administration Hospital and the Schumpert Medical Center. While working for the crime lab, he drew blood from at least 1700 people. The kit he used to draw defendant's blood was a B-D Blood Alcohol Kit Number 4990. This kit contains a preservative which inhibits any ethanol (alcohol) production in the tubes used to store defendant's blood.
The chemist who performed the blood analysis, Rebecca Collins, also was employed by the crime lab. She had a Bachelor of Science in Microbiology with a Minor in Chemistry. She was trained to use the gas chromatograph by her supervisor and a blood alcohol analyst. Additionally, she was required to read a training manual compiled by her supervisor. On March 29, 1985, the Louisiana Department of Public Safety certified her as a blood alcohol analyst. The certification was valid for five years. Prior to analyzing defendant's blood sample, she had analyzed hundreds of samples for blood alcohol determination. She participated in monthly proficiency testing, in which the crime lab analysts would test samples containing known blood alcohol content sent from the University of Tennessee Toxicology Lab. She also participated in proficiency tests on samples exchanged with crime labs in Monroe and Alexandria. The crime lab was accredited by the American Society of Crime Laboratory Directors.
The chemist described the gas chromatograph and how she runs it.[4] She routinely inspects and maintains it. She always follows the crime lab's standard procedures for running it. Before testing evidence samples such as defendant's, she calibrates and tests the gas chromatograph by injecting reference samples of known blood alcohol content certified by the manufacturer. First, she injects a sample containing an alcohol content of 0.40% and the machine is told that this is a 0.40% blood alcohol sample. She reinjects the 0.40% sample to make sure that she will get the correct reading. Next, she injects a 0.0% sample (water) to insure that there is no contamination from the internal standard or carryover from the 0.40% sample. She then injects a 0.10% sample because that is a significant level in criminal cases. As a final step in the calibration procedure, she injects a blood control sample which has 0.15% blood alcohol. The calibration procedure spot checks the chemicals used in the calibration because the results of the different reference samples confirm each other. She must get the correct answer for each of the samples. After the calibration is complete, she is ready to test evidence samples such as defendant's to determine their blood alcohol level. The evidence samples are tested in batch lots usually once or twice weekly. After half of a batch is analyzed, the 0.15% blood control sample is reanalyzed. After the rest of the batch is analyzed, the 0.15% blood control sample is again reanalyzed to make sure that nothing has gone wrong during the analysis of the blood samples. Each evidence sample is analyzed in duplicate. The readings on the duplicate samples must be within 0.005% of each other. The chemist ran duplicate samples of defendant's blood and obtained results of 0.164% and 0.165%. She took the lower figure, 0.164%, and rounded down to 0.16%. She stated that she was certain that the result of 0.16% was correct. Both the chemist and her supervisor testified that they would definitely be aware of any malfunction in the gas chromatograph. The supervisor said, "One of the nice things about gas chromatograph analysis is if it's working it is working well. If it doesn't work you don't get an answer, wrong or right."
The chemist's supervisor, Jimmy Barnhill, was qualified as an expert in blood analysis and the effects of alcohol on driving. During his testimony, he used the Widmark Formula to calculate the amount of alcohol in defendant's body when the sample was drawn based upon defendant's weight (180 pounds), sex, and blood alcohol level. According to the calculation, defendant had 3.9 ounces of pure alcohol in his body when the sample was drawn. That corresponds to the amount of pure alcohol in about 8.1 twelve-ounce cans of beer. Mr. Barnhill also calculated that at the time of the accident, defendant's blood alcohol level probably was between 0.186% and 0.20%. He then testified about the effects of a blood alcohol level of 0.16%:
At a level of .16 a person would be noticeably intoxicated.... [T]hey would be able to walk possibly, they might be able to talk and to some extent carry on a conversation. But I think a sober person would be ... able to know that that person was intoxicated, had been drinking alcohol. At ... this level a person's vision would be impaired. There [are] two types of vision that we depend on when we are driving. One of them is our peripheral vision....
When you are driving down the road you depend on that vision extensively. You make decisions, albeit unconscious in some cases, based on your ability to use your peripheral vision.... When your blood alcohol level is elevated your peripheral vision is greatly impaired and the higher it gets the more impaired it gets. It would be like wearing blinders so that you could not have the benefit of being able to see things out to the side.
Of course your ability at night to recover from a dazzling light is greatly diminished as your blood alcohol level increases. At very high levels double vision sets in where you actually see two objects when you are looking at something.... The ability to make judgments is impaired. Judgments about how fast you are going, judgments about how far it is from this point to that point, when do I need to put my brakes on....
All judgment is impaired. Your ability to get your foot from the gas pedal to the brake pedal, that time is increased, the reaction time is increased. All of these things that you need to drive safely are greatly impaired at the level of .16.
The court of appeal relied on State v. Rowell, 517 So.2d 799 (La.1988), to find that the results of the blood alcohol test should not have been admitted into evidence. In Rowell, we found that the results of a blood alcohol test using a gas chromatograph were inadmissible against a defendant charged with operating a motor vehicle while under the influence of alcohol in violation of La.R.S. 14:98. We stated in Rowell, 517 So.2d at 800:
In order for the state to avail itself of the statutory presumption of a defendant's intoxication arising from a chemical analysis of his blood under La.R.S. 32:662, it must show that the state has promulgated detailed procedures which will insure the integrity and reliability of the chemical test, including provisions for repair, maintenance, inspection, cleaning, certification, and chemical accuracy. It must also show that the state has strictly complied with the promulgated procedures.
In Rowell, we decided that the state had failed to meet its burden of proving that the regulations insured the integrity and reliability of the blood alcohol test.[5] Subsequent to the court of appeal's decision, we stated in State v. McElroy, 553 So.2d 456, 458 n. 1 (La. 1989), that "Rowell is not pertinent when the state does not rely on the statutory presumption of intoxication." The statutory presumption of intoxication is contained in La.R.S. 32:662 A(1)(c).[6] In the instant case, the trial judge did not mention the presumption in his charge to the jury. Neither did the state's opening nor closing argument mention it. Instead of relying on the presumption, the state provided Mr. Barnhill's expert testimony explaining to the jury the effect of the 0.16% blood alcohol level on driving. Accordingly, Rowell is not pertinent here.
When the state does not rely on the statutory presumption of intoxication, it may use all admissible evidence to prove that a defendant is guilty of vehicular homicide. We stated in McElroy, 553 So.2d at 458:
[W]ithout the benefit of the statutory presumption of intoxication, the state may nonetheless endeavor to prove that a defendant was guilty of driving while intoxicated, and, in the process, attempt to use all admissible evidence, including the hospital record, the testimony of the technologist performing the blood alcohol test, and expert testimony concerning the likely effect upon an individual of a given blood alcohol level.
In order for the results of a blood alcohol test to be admissible, the state must prove that the reliability of the test satisfies due process and fairness. We stated in McElroy, 553 So.2d at 458 n. 1:
Should there be no predominant evidence of "careful adherence to strict procedures in administering intoxication tests," State v. Jones, [316 So.2d 100, 105 (La. 1975)], or in the event that a defendant takes issue with the qualifications of a technician, the quality of the testing machine, or the maintenance of the equipment, and so forth, the court, in the interest of due process and fairness, may well be entitled to bar the evidence.
In the instant case, the technician who drew defendant's blood was trained, certified, and experienced. He drew the blood within three hours of the accident. The kit he used to draw the blood contained a preservative which inhibits any ethanol (alcohol) production in the tubes used to store the blood. He gave the blood to the chemist the day after the accident. The next day, the chemist performed the blood analysis. She was required to follow and did follow the crime lab's procedures, which exceeded the requirements of Department of Public Safety regulations. La. Admin.Code tit. 55:I, §§ 551-553 (1987); see also 14 La.Reg. 360 (1988) (to be codified at La.Admin.Code tit. 55:I, §§ 551-563). She was trained, took frequent proficiency tests, and was certified as a blood alcohol analyst by the Louisiana Department of Public Safety. She routinely inspected and maintained the gas chromatograph. Prior to analyzing defendant's blood sample, the chemist calibrated the machine by analyzing four samples of known blood alcohol content. The calibration procedure spot checks the chemicals used in the calibration because the results of the samples confirm each other. She analyzed defendant's blood sample in duplicate. The results of the gas chromatograph were shown to be reliable by proficiency tests regularly performed on samples sent from the University of Tennessee Toxicology Lab and crime labs in Monroe and Alexandria. Under the circumstances, we find that the reliability of the blood alcohol test satisfies due process and fairness. The results of the test were properly admitted into evidence. The court of appeal erred in holding otherwise. We must reverse. We will remand the case to the court of appeal for consideration of defendant's remaining assignments of error.
DECREE
For the reasons assigned, the judgment of the court of appeal is reversed. The case is remanded to the court of appeal for consideration of defendant's remaining assignments of error.
DENNIS, J., dissents and assigns reasons.
DENNIS, Justice, dissenting.
I respectfully dissent.
The majority finds that the holding of State v. Rowell, 517 So.2d 799 (La.1988), is inapplicable to this case because the state did not rely upon the presumption of intoxication found in La.R.S. 32:662. An examination of the rationale of Rowell and the circumstances of this case, however, compel me to disagree. In Rowell, the defendant was accused of driving while intoxicated in violation of La.R.S. 14:98(A)(1). One element of the offense was intoxication. We held that the conviction could not stand because the state used the potentially inaccurate test result to establish a presumption regarding an element of the offense. In the present case, the defendant is accused of vehicular homicide, one element of which is driving with a blood alcohol concentration of .10 or greater. See La.R.S. 14:32.1 A(2). Here, the potentially inaccurate test result is not used to establish a presumption regarding an element of the offense, but to establish an element of the offense itself. Since the test establishes an element of the crime rather than merely a presumption, this court should be more concerned, not less concerned, with the accuracy of the test.
The dicta from State v. McElroy, 553 So.2d 456, 458 n. 1 (La.1989), cited by the majority, does not compel a different result even if it is correct. In that case, the state argued that it could introduce the test results in a DWI case if it did not attempt to use the presumption of intoxication. In that case, the test result would be merely circumstantial evidence that would allow the factfinder to draw an inference of intoxication. Because that is a vastly different situation than the one present here, where the test result may establish an element of the offense without the necessity of any inference, the McElroy dicta should not be extended to cover this situation. Furthermore, McElroy itself may be incorrect, given the enormously prejudicial effect the admission of an erroneous test result would have. See La.Code Evid. art. 403.
In addition to my disagreement with the court's refusal to apply the Rowell standards in this case, I am not convinced that even those standards adequately protect defendants' constitutional rights. The crucial time for the blood alcohol concentration is the time of operation of the motor vehicle. See, e.g., La.R.S. 14:32.1; La.R.S. 32:662. In most cases decided under Rowell, however, the factfinder must extrapolate to determine the blood alcohol concentration at the time of operation from a test that shows the blood alcohol concentration at the time of the test-taking. While there was expert testimony on the point in the present case, in many cases it is impossible to determine whether the defendant was in the absorption phase (i.e., his blood alcohol concentration was rising) or the elimination phase (blood alcohol concentration was decreasing) at the time of the test. Given this initial uncertainty, combined with the fact that alcohol absorption and elimination rates vary among individuals, I do not believe that a single test some time later than the vehicle operation is reliable enough to form a basis for a criminal conviction. The regulations should require at least two readings, so that the factfinder can at least determine whether the blood alcohol concentration is on the way up or on the way down. The administration of two blood alcohol concentration tests is the most effective and accurate way to determine where on the absorption curve an individual falls at any given moment, and it thus allows the factfinder to more accurately extrapolate the defendant's blood alcohol concentration while driving. Note, In Vino Veritas: The Truth about Blood Alcohol Presumptions in State Drunk Driving Law, 64 N.Y.U.L.Rev. 141, 178 (1989); see also S. Brent & S. Stiller, Handling Drunk Driving Cases § 3:11, at 46 (1985); Cohen, Understanding and Using Blood Alcohol Concentration in Liquor Liability Cases, in Practitioner's Guide to Liquor Liability Litigation 63-64 (R. Beitman ed. 1987); Fitzgerald & Hume, The Single Chemical Test for Intoxication: A Challenge to Admissibility, 66 Mass.L.Rev. 23, 33, 36 (1981). The use of two tests would also have the salutary effect of affording the benefit of evidence regarding absorption and elimination rates to those defendants who are unable to afford to present expert testimony on this subject. Minnesota and Nevada by statute require two consecutive tests to confirm the results of the first test. See Minn.Stat.Annot. § 169.123(2)(b); Nev.R.S. § 484.386(1). Several other states, including Michigan, New Jersey, California, Rhode Island, Maine and North Carolina, routinely give two tests though not required by statute. Note, supra, at 179. The practice of these states indicates that the burden of performing a second test within a short period of time is not prohibitive.
Finally, I note that the trial judge in this case gave careful instructions to the jury that it was to base a conviction upon a finding that the blood alcohol concentration exceeded .10 at the time of operation. But if the court carelessly gives the jury the impression that it may presume that the defendant's blood alcohol concentration was the same at the time of testing and the time of operation, this would be unconstitutional. The effect of such a presumption would be to shift the burden of persuasion to the defendant, which the Supreme Court held unconstitutional in Sandstrom v. Montana, 442 U.S. 510, 99 S.Ct. 2450, 61 L.Ed.2d 39 (1979). In order to avoid this problem, this court should consider adopting the rule established by the Vermont Supreme Court in State v. Dumont, 146 Vt. 252, 499 A.2d 787 (1985), which requires the prosecution to always introduce expert testimony extrapolating the defendant's blood alcohol concentration as shown by the test result back to the time of driving. (However, relation-back evidence, when supplementing a single BAC test, may be just as prejudicial to the defendant as no extrapolation at all.) Cf. Note, supra; Comment, Driving with 0.10% Blood Alcohol: Can the State Prove It?, 16 U.S.F.L. Rev. 817 (1982).
NOTES
[*] Judge Melvin A. Shortess of the First Circuit Court of Appeal participated in this decision as Associate Justice Pro Tempore.
[1] 545 So.2d 698 (La.App. 2d Cir. 1989).
[2] 556 So.2d 46 (La.1990).
[3] La.R.S. 14:32.1 A(2) provides:
A. Vehicular homicide is the killing of a human being caused proximately or caused directly by an offender engaged in the operation of, or in actual physical control of, any motor vehicle, aircraft, vessel, or other means of conveyance whether or not the offender had the intent to cause death or great bodily harm whenever any of the following conditions exist:
(2) The offender's blood alcohol concentration is 0.10 percent or more by weight based upon grams of alcohol per one hundred cubic centimeters of blood.
[4] According to the chemist, a gas chromatograph is an instrument which is used to separate compounds and to identify and measure them. It has a heated oven which contains a column of glass tubing packed with a solid material. A constantly flowing stream of gas is passed through the column. When a sample is injected it is vaporized and swept onto the column. Different compounds in the sample will travel through the column at different rates and thus be separated. When they emerge from the end of the column their presence is registered by a detector and recorded on a piece of graph paper as a peak. The area under the peak is proportional to the amount of the compound that is present. When the chemist runs a blood alcohol analysis, the alcohol in the blood sample is ethanol. She mixes one millileter of the blood sample with one millileter of an internal standard. The internal standard used by the crime lab is n-propanol. It serves as a measuring device to determine how much ethanol is present. After the mixture passes through the gas chromatograph, she gets a graph with two peaks on it. The first is the ethanol peak and the second is the internal standard peak. Every mixture will have the same amount of internal standard, so she can compare the sizes of the ethanol peak and the internal standard peak to determine how much ethanol is in the blood sample. A computer program in the integrator of the gas chromatograph also calculates and prints out the blood alcohol level based on the size of the peaks. The method used by the crime lab for running the gas chromatograph is referred to as direct injection with internal standard.
[5] The court of appeal stated that even if the crime lab procedures were adequate to satisfy the deficiencies of the procedures found in Rowell, they could not supplement the defective regulations because they were not adopted and promulgated by the Department of Public Safety as required by La.R.S. 32:663. The court further noted that the Department of Public Safety has promulgated new regulations concerning the analysis of blood which became effective June 20, 1988. 14 La.Reg. 360 (1988) (to be codified at La.Admin.Code tit. 55:I, §§ 551-563). These regulations were not in effect at the time of this incident.
[6] La.R.S. 32:662 A(1)(c) provides:
A. The chemical test or tests as provided for by this Part shall be subject to the following rules and shall be administered as provided for hereafter:
(1) Upon the trial of any criminal action or proceeding arising out of acts alleged to have been committed by any person while driving or in actual physical control of a vehicle while under the influence of alcoholic beverages the amount of alcohol in the person's blood at the time alleged as shown by chemical analysis of the person's blood, urine, breath or other bodily substance shall give rise to the following presumptions:
(c) If there was at that time 0.10 per cent or more by weight of alcohol in the person's blood, it shall be presumed that the person was under the influence of alcoholic beverages.