112 P.3d 566 (2005)
STATE of Washington, Respondent,
v.
Jeremy D. REIER, Appellant.
No. 53193-0-I.
Court of Appeals of Washington, Division 1.
May 31, 2005.
Randall W. Yates, Snohomish Pros. Atty. Office, Charles Franklin Blackman, Everett, WA, for Respondent.
Suzanne Lee Elliott, Attorney at Law, Seattle, WA, for Appellant.
AGID, J.
¶ 1 Jeremy Reier appeals his vehicular homicide conviction. He argues that the trial court erred by admitting his blood alcohol test results because the toxicologist failed to comply with the Washington Administrative Code (WAC) when analyzing his blood sample. But the toxicologist conducted duplicate analyses which agreed within 0.01 grams of alcohol per 100 milliliters of blood, as required by the WAC. The toxicologist also properly calibrated the instruments to accurately test blood alcohol. And finally, because the WAC does not require the toxicologist to perform replicate analyses, there was no error in admitting the test results. We affirm.
FACTS
¶ 2 At 1:45 a.m. on December 1, 2001, Jeremy Reier was driving his pickup truck on a state road in Snohomish County when he lost control, left the roadway, and collided with construction barrels and a dumpster. Reier's passenger, Michael Lee, was thrown from the truck and later died from head injuries.
¶ 3 When police officers arrived at the scene, they noticed that Reier smelled of intoxicants, had slurred speech, and had poor coordination and balance. Reier's blood alcohol level, measured almost two hours after the accident, was 0.17 grams of ethanol per 100 milliliters of blood.[1]
¶ 4 The State charged Reier with one count of vehicular homicide. Before trial, he moved to suppress the blood alcohol test results. The trial court denied the motion, as well as a later motion to strike the test results. A jury convicted Reier, and the court sentenced him to 31 months' confinement. Reier appeals.
DISCUSSION
¶ 5 A person commits vehicular homicide if he or she operated a motor vehicle while under the influence of intoxicating liquor or drugs and his or her driving caused another's death.[2] A person is guilty of driving under the influence of intoxicating liquor or drugs if he or she drove a vehicle and had, within two hours of driving, "an alcohol concentration of 0.08 or higher as shown by analysis of the person's breath or blood made under RCW 46.61.506[.]"[3] Under RCW 46.61.506(3), a breath or blood analysis is valid only if it was performed in compliance with the methods approved by the state toxicologist. Those methods are codified in title 448 of the WAC.
¶ 6 Reier challenges the trial court's decision to admit his blood alcohol test results, arguing that the toxicologist failed to comply with the approved testing methods and relevant code requirements. Before a court may admit blood alcohol test results, "the State must present prima facie proof that the test chemicals and the blood sample are free from any adulteration which could conceivably introduce error to the test results."[4]
I. Duplicate Analyses
¶ 7 WAC 448-14-020 requires that certain procedures be used when analyzing a blood sample for alcohol content. Under that regulation, analytical procedures "should" include a control test, a blank test, and "[d]uplicate analyses that should agree to within 0.01% blood alcohol deviation from the mean."[5] Test results should be expressed as grams of alcohol per 100 milliliters of whole blood sample, and results should be reported to two significant figures.[6]
¶ 8 Reier argues that, in his case, the toxicologist neglected to specify whether the duplicate test results were within 0.01 percent blood alcohol deviation from the mean as required. The duplicate analyses of Reier's blood test were reported as 0.171 and 0.172 grams of alcohol per 100 milliliters of blood. According to Reier, these results were not within "0.01%" of the mean as required by the regulations, because the designation "0.01%" actually requires the duplicate analyses to agree within 0.0001 grams of alcohol per 100 milliliters of blood, rather than 0.01 grams.
¶ 9 In response to this argument, the trial court found that the term "0.01%" as it relates to blood alcohol is ambiguous. Nevertheless, the court denied Reier's motion on the ground that WAC 448-14-020 uses the word "should" instead of "shall." That is, the regulation states that the analytical procedure should include duplicate analyses that should agree to within 0.01 percent deviation. The court noted that "should" is a directive term with a different meaning from "shall," which is a mandatory term. The court therefore ruled that, while the duplicate analyses conducted in Reier's case may not have fully complied with the regulation, suppression is not required because the regulation itself is not mandatory.
¶ 10 This court has previously held that "WAC 448-14-020 does not bar admissibility of a blood alcohol analysis test when a second confirming test is not performed because the term `should' is directional and is not mandatory as is the term `shall' in this statute."[7] We agree with that holding and with the trial court. But even if this were not the case, Reier's blood alcohol test results were admissible because "0.01%" does not mean "0.0001" in this context. Rather, it means 0.01 grams of alcohol per 100 milliliters of blood.
¶ 11 We interpret agency regulations the same way we interpret statutes,[8] and statutory interpretation is a question of law that we review de novo.[9] When interpreting a statute or regulation, we must avoid any "absurd or strained consequences."[10] Reier's reading of the duplicate analyses provision is one that would lead to absurd results. Since the regulation was adopted in 1970, employees at the state toxicology laboratory have consistently interpreted the provision to mean that duplicate test results must agree within .01 grams of alcohol per milliliters of blood. And although Reier's expert witness testified that the percentage sign indicates that the duplicate analyses must agree within 0.0001 grams, he also acknowledged that during his 28 years working for the state toxicologist, he interpreted the regulation as requiring duplicate analyses to agree within 0.01 grams.
¶ 12 To read the regulations as requiring the duplicate analyses to agree within 0.0001 grams of alcohol would also contradict the regulation when read as a whole. First, WAC 48-14-020(2)(a) requires that results be expressed as grams of alcohol per 100 ml of blood. When a number is expressed per 100, that number becomes a percent. For example, 0.17 grams per 100 milliliters is 0.17 percent. Therefore, 0.01 percent blood alcohol is equivalent to 0.01 grams per 100 milliliters. And further, WAC 48-14-020(2)(b) requires analysts to report test results to two significant figures after rounding. Requiring duplicate analyses to agree within four significant figures is inconsistent with this regulation. This is especially true because the instruments used by the state toxicologist only report results to three significant figures, which are then rounded to two significant figures for reporting. We hold that the term "0.01%" in WAC 48-14-020(1)(a)(iii) means that duplicate analysis should agree within 0.01 grams of alcohol per 100 milliliters of whole blood.[11]
II. Instrument Calibration
¶ 13 During trial, Reier unsuccessfully moved to strike his blood test results on the ground that one of the instruments used to analyze his blood was not properly calibrated. We review a trial court's evidentiary ruling for an abuse of discretion.[12] A court abuses its discretion when it bases its decision on untenable grounds or exercises discretion in a way that is manifestly unreasonable.[13]
¶ 14 William Marshall, the state toxicology lab technician who analyzed Reier's blood sample, testified that the lab uses headspace gas chromatography to analyze the presence of alcohol in a blood sample. A chromatograph is an instrument that separates compounds, and by calibrating the instrument, one can determine the quantities of various compounds in a blood sample. Before testing a blood sample, an analyst must calibrate the chromatograph. This is done, in part, by testing different mixtures to verify that the instrument can differentiate between methanol, ethanol, isopropanol, acetone, and propanol. These mixtures are run on two separate instruments to test both instruments' abilities to separate those elements, as two different instruments must be used to obtain the duplicate analyses discussed above.
¶ 15 Marshall testified that he calibrated two chromatographs (instruments 1 and 2) when analyzing Reier's blood sample. While instrument 2 successfully recognized, identified, and quantified the various compounds, instrument 1 separated the compounds but did not quantify them. Marshall stated that for a blood alcohol test, the instrument must show only that it is capable of separating the compounds. That is, quantification of the compounds is unnecessary when testing blood for alcohol, as opposed to when one is testing blood for volatiles.
¶ 16 The trial court accepted Marshall's testimony that instrument 1 separated the compounds during calibration as required, and that nothing in the regulations or toxicologist's protocols requires quantification of the substances for the purpose of testing blood for alcohol. The court recognized that only the ethyl alcohol concentration in Reier's blood was at issue, and the State presented two blood tests obtained by two separate instruments that quantified Reier's blood alcohol to nearly identical values. The court ruled that this satisfied the WAC's admonition that the blood analysis procedures be precise, accurate, and specific.[14]
¶ 17 The trial court did not abuse its discretion here. The evidence presented shows that when determining a person's blood alcohol level, the instruments must be capable of separating compounds, but the quantity of those compounds is unimportant. Reier presents no evidence to the contrary. Here, both instruments showed adequate separation of compounds, and both instruments obtained almost identical results. The trial court correctly admitted the blood test results.[15]
III. Replicate Analyses
¶ 18 Finally, Reier argues that his blood test results are inadmissible because Marshall did not conduct a replicate analysis of the blood sample. According to Marshall, a replicate analysis involves analyzing the same sample on the same day with the same instrument, while duplicate analyses are run on different instruments. Marshall testified that the lab does not routinely perform replicate analyses because it is not required to do so. But Reier argues that Marshall was required to perform a replicate analysis under WAC 448-14-010, which states:
Any quantitative blood alcohol analysis method which meets the following criteria is approved by the state toxicologist and may be used in the state of Washington....
The blood analysis procedure should have the following capabilities:
(1) Precision and accuracy.
(a) The method shall be capable of replicate analyses by an analyst under identical test conditions so that consecutive test results on the same date agree with a difference which is not more than 3% of the mean value of the tests. This criterion is to be applied to blood alcohol levels of 0.08% and higher.
. . . .
¶ 19 While Reier contends that section (1)(a) requires lab technicians to perform replicate analyses, the trial court concluded that neither the regulations nor toxicology protocols requires anything more than a methodology that is capable of replicate analyses. And because head space gas chromatography is capable of replicate analyses, the court found no error here. We again apply de novo review to this interpretation of agency regulations.[16]
¶ 20 We conclude that WAC 448-14-010(1)(a) does not require lab technicians to perform replicate analyses on blood samples. The regulation plainly states that blood analysis methods must be capable of replicate analyses. And in contrast to WAC 448-14-020, which encourages lab technicians to perform duplicate analyses,[17] WAC 448-14-010 neither encourages nor requires the technicians to perform replicate analyses with each blood sample. Because the State introduced evidence that head gas chromatography is capable of replicate analyses, which Reier did not contradict, there was no error here.[18]
¶ 21 We affirm.
WE CONCUR: KENNEDY and GROSSE, JJ.
NOTES
[1] The legal level in Washington is 0.08 grams per 100 milliliters.
[2] RCW 46.61.520(1)(a); State v. Morgan, 123 Wash.App. 810, 815, 99 P.3d 411 (2004) (citing State v. Rivas, 126 Wash.2d 443, 451, 896 P.2d 57 (1995)).
[3] RCW 46.61.502(1)(a).
[4] State v. Clark, 62 Wash.App. 263, 270, 814 P.2d 222 (1991) (citing State v. Weston, 54 Wash.App. 105, 108, 772 P.2d 1036 (1989); State v. Barefield, 47 Wash.App. 444, 458, 735 P.2d 1339 (1987), aff'd, 110 Wash.2d 728, 756 P.2d 731 (1988)).
[5] WAC 448-14-020(1)(a)(iii).
[6] WAC 448-14-020(2)(a), (b).
[7] Tennant v. Roys, 44 Wash.App. 305, 313, 722 P.2d 848 (1986).
[8] Cobra Roofing Serv., Inc. v. Dep't of Labor & Indus., 122 Wash.App. 402, 409, 97 P.3d 17 (2004) (citing Children's Hosp. & Med. Ctr. v. Dep't of Health, 95 Wash.App. 858, 864, 975 P.2d 567 (1999), review denied, 139 Wash.2d 1021, 994 P.2d 847 (2000)).
[9] Agrilink Foods, Inc. v. State Dep't of Revenue, 153 Wash.2d 392, 396, 103 P.3d 1226 (2005) (citing W. Telepage, Inc. v. City of Tacoma Dep't of Fin., 140 Wash.2d 599, 607, 998 P.2d 884 (2000)).
[10] State v. Stannard, 109 Wash.2d 29, 36, 742 P.2d 1244 (1987) (citing State v. Richardson, 81 Wash.2d 111, 499 P.2d 1264 (1972)).
[11] Accord, State v. Babiker, 126 Wash.2d 664, 110 P.3d 770, (2005) (interpreting WAC 48-14-020 to mean that duplicate analyses should agree within 0.01 grams of alcohol per 100 milliliters of blood).
Reier argues that neither our interpretation nor the holding in Babiker comports with the Supreme Court's recent holding in City of Seattle v. Clark-Munoz, 152 Wash.2d 39, 93 P.3d 141 (2004). There, the court held that blood alcohol test results are admissible only if they were analyzed in strict compliance with the relevant statute and regulations. Id. at 48-50, 93 P.3d 141. But our holdings here and in Babiker in no way depart from Clark-Munoz. We do not hold that blood alcohol analyses need not strictly comply with the statute and regulations, but rather we are interpreting what exactly those regulations require.
[12] State v. Markle, 118 Wash.2d 424, 438, 823 P.2d 1101 (1992) (citing State v. Mak, 105 Wash.2d 692, 718 P.2d 407 (1986), cert. denied, 479 U.S. 995, 107 S.Ct. 599, 93 L.Ed.2d 599 (1986)).
[13] State v. Zunker, 112 Wash.App. 130, 140, 48 P.3d 344 (2002) (citing State v. Valdobinos, 122 Wash.2d 270, 279, 858 P.2d 199 (1993)), review denied, 148 Wash.2d 1012, 62 P.3d 890 (2003).
[14] WAC 448-14-010.
[15] For the first time on appeal, Reier argues that the State failed to show that the two instruments agreed within .01 of the mean during the control tests, and that instrument l's inability to quantify the separate compounds meant that the State could not show that the testing was free from interferences native to the sample as required by the regulations. Because Reier raises these issues for the first time on appeal, we decline to discuss them. RAP 2.5(a).
[16] Cobra Roofing Service, 122 Wash.App. at 409, 97 P.3d 17 (citing Children's Hosp., 95 Wash. App. at 864, 975 P.2d 567); Agrilink Foods, 153 Wash.2d at 396, 103 P.3d 1226 (citing W. Telepage, 140 Wash.2d at 607, 998 P.2d 884).
[17] See WAC 448-14-020(1)(a)(iii).
[18] At the trial court level, the State repeatedly noted that Reier's blood sample was still available for retesting, and the court noted that fact when deciding Reier's calibration and replicate analysis arguments. The record does not indicate whether Reier ever retested the blood.