concurring.
I concur with the result of this opinion. However, I disagree with the majority that the trial court committed error, albeit harmless error, when it failed to require that the thousandth-place digit of Defendant’s intoxi-lyzer test result, which revealed that Defendant had a .113 percent (.113%) blood alcohol concentration (BAC) following his arrest for driving under the influence of intoxicating liquor (DUI), be redacted before the test result was admitted into evidence.
According to the majority, the thousandth-place digit, i.e., the “3,” was “irrelevant” because the intoxilyzer manufacturer’s recognized margin of error for the accuracy of the test result was one one-hundredth percent (.01%).
While I agree that the thousandth-place digit was irrelevant for purposes of determining Defendant’s guilt under Hawai’i Revised Statutes (HRS) § 291-4(a)(2) (1985 & Supp. 1990), I believe that it was relevant for purposes of determining Defendant’s guilt under HRS § 291-4(a)(l) (1992),1 which Defendant was alternatively charged with violating.
At the time Defendant was arrested, the standards for admitting into evidence Defendant’s BAC test result were set forth in HRS § 291-5 (1985):
(a) In any criminal prosecution for a violation of section 291-4, ten-hundredths per cent or more by weight of alcohol in the defendant’s blood within three hours after the time of the alleged violation as shown by chemical analysis or other approved analytical techniques of the defendant’s blood or breath shall be competent evidence that the defendant was under the influence of intoxicating liquor at the time of the alleged violation.
(b) In any criminal prosecution for a violation of section 291-4, the amount of *148alcohol found in the defendant’s blood within three hours after the time of the alleged violation as shown by chemical analysis or other approved analytical techniques of the defendant’s blood or breath shall be competent evidence that the defendant was under the influence of intoxicating liquor at the time of the alleged violation and shall give rise to the following presumptions:
(1) If there were five-hundredths per cent or less by weight of alcohol in the defendant’s blood, it shall be presumed that the defendant was not under the influence of intoxicating liquor at the time of the alleged violation.
(2) If there were in excess of five-hundredths per cent but less than ten-hundredths per cent by weight of alcohol in the defendant’s blood, such fact may be considered with other competent evidence in determining whether or not the defendant was at the time of the alleged violation under the influence of intoxicating liquor but shall not of itself give rise to any presumption.
(c) Subsection (b) shall not be construed as limiting the introduction of any other competent evidence bearing upon the question of whether or not the defendant was under the influence of intoxicating liquor at the time of the alleged violation.
Pursuant to HRS § 291-5, Defendant would be presumed to have driven under the influence of intoxicating liquor in violation of HRS § 291-4(a)(2) if, following his arrest, the result of an intoxilyzer test, adjusted for the manufacturer’s .01 percent margin of error, revealed that Defendant had a .10 percent BAC. Evidence of the thousandth-place digit of the intoxilyzer test result would thus not be relevant for determining Defendant’s BAC, which by statute is based on the less precise hundredth-place digit.
Since the testimony at trial indicated, however, that the intoxilyzer was capable of accurately reading an individual’s BAC to the thousandth-place digit, as long as the machine’s margin of error was factored into the BAC test result, I believe the test result was “competent evidence bearing upon the question of whether or not Defendant was under the influence of intoxicating liquor at the time of the alleged violation.”
For example, when asked why the BAC test result was displayed digitally on the intoxilyzer machine as a three-digit, rather than two-digit, figure, i.e., to the thousandth, rather than hundredth, place, Honolulu Police Department criminalist Joanne Furuya (Ms. Furuya) explained as follows:
Q. Now is it possible for the machine to accurately measure to the thousandth of a percentage if it has a margin of error in the hundredths of a percentage?
A. Yes.
Q. It is. And how is that possible?
A. When you take the thousandths of a place it is a more precise number than the hundredth of a place, it will be hard to have the margin of error to the thousandths of a place when you are only referring to the second of a place. Do you understand what I am saying?
If your margin of error is plus, minus point 0 one, what we do is if there’s a reading of point 0 nine five, that five is not automatically brought up to the point one 0, it’s totally intricate.
So we are giving the benefit of the doubt by that third digit.
Q. Okay. I guess my question is, if you have a machine that has a margin of error that is it’s assuming that this machine can only read or that the final reading will be adjusted by one one hundredth of a percentage, how is it possible that the machine can then be assumed to be accurately reading to a police [sic] ten times more refined than that?
A. Okay. We are saying that the margin of error is point 0 one, right?
Q. Yes.
A. If the machine had a reading of point 0 [sic]2 seven five, the error or the margin between the point one six five and the point one eight five, and the fives could *149be truncated so it could be—it can go between point one six and point one eight if you truncate the digit. Are we getting things across?
Transcript 2/2/93, at 74-75 (footnote supplied).
In other words, given Ms. Furuya’s .175 intoxilyzer result and accounting for the .01 margin of error, a one one-thousandth of a percentage point reading creates a BAC range from .165 to .185. A one one-hundredth of a percentage reading, achieved by truncating the “5” from Ms. Furuya’s original .175 result, creates a BAC range from .16 to .18. By reading the result at the one one-hundredth of a percentage point, a BAC range is created that is .005 lower than what the actual BAC may be. Therefore, the one one-thousandth of a percentage point reading may be more accurate than the one one-hundredth of a percentage point reading.
Ms. Furuya also testified that an intoxilyzer operator was not allowed to even begin a BAC test unless the reading on the machine’s light display is a maximum of “point zero zero three” (.003). Ms. Furuya explained the significance and accuracy of this thousandth-place reading:
A. It’s just a reading of the instrument and it can—now that’s in the thousands [sic] range, okay, it can fluctuate—the machine can fluctuate but the range that is allowable before one is able to use a test is a nonminus point zero zero one, point zero zero two, point zero zero three value.
Q. But then the machine is not accurate to the thousandth’s digit completely because point zero zero three is not point zero zero zero?
A. Point zero zero three is not point zero zero zero, that’s right.
Q. So the machine is not reading completely accurately that there is no alcohol in that air blank sample?
A. That’s why the margin of error is at the hundredth place. That’s why it’s not at the thousandth place.
Q. And so when we get the number in the thousands [sic] column, there is a question as to just how accurate the machine is reading at that point?
A No. Because as I mentioned before if one gets a point one seven three reading it’s plus or minus point zero one which a hundredth place it can vary between one six three and one eight three so that third digit is irrelevant.
Q. The third digit is irrelevant?
A. For the reading as proposed here.
Q. It only becomes relevant as applied to the defendant?
A. Yes, it’s—you are mistaken for what I am saying. When you apply the hundredths place margin of error, your third digit becomes irrelevant because you are moving the range in the hundredths place which is a less accurate value. Now you understand what I am saying?
Id. at 78-79.
Criminalist Claire Chun similarly testified that although a two-digit readout of a defendant’s BAC suffices for conviction purposes since HRS § 291-4(a)(2), the statute defining the unacceptable BAC level, is stated in two, rather than three, digits, the intoxilyzer machine can nonetheless read an individual’s BAC to a thousandth of a percent. Id. at 110-11.
Since the intoxilyzer was capable of reading Defendant’s BAC to a more precise thousandth-place digit, I believe that the entire test result, as adjusted for the machine’s margin of error, was competent to determine Defendant’s guilt under HRS § 291-4(a)(l), and in accordance with HRS § 291-5, should not be precluded as evidence at trial.
. At the time Defendant was charged with driving under the influence of intoxicating liquor in violation of HRS § 291-4(a)(l) or (2) (Supp. 1992), the statute provided as follows:
(a) A person commits the offense of driving under the influence of intoxicating liquor if:
(1) The person operates or assumes actual physical control of the operation of any vehicle while under the influence of intoxicating liquor, meaning that the person concerned is under the influence of intoxicating liquor in an amount sufficient to impair the person’s normal mental faculties or ability to care for oneself and guard against casualty; or
(2) The person operates or assumes actual physical control of the operation of any vehicle with .10 per cent or more, by weight of alcohol in the person's blood.
. Based on our reading of Ms. Furuya's testimony, we believe that Ms. Furuya intended to base her example on a reading of "point one seven five."