VÁSQUEZ, Presiding Judge.
¶ 1 In
Guthrie v. Jones,
202 Ariz. 273, ¶ 2, 43 P.3d 601, 602 (App.2002), this court held that evidence regarding breath-to-blood partition ratios
is inadmissible in a prosecution for di’iving or being in actual physical control of a vehicle with an alcohol concentration of .08 or more within two hours of driving, in violation of A.R.S. § 28-1381(A)(2). However, we also held that when the state uses breath-test results to establish a presumption of intoxication in a prosecution for driving under the influence of an intoxicant (DUI) while impaired to the slightest degree,
see
§ 28-1381(A)(1) and (G), the defendant may introduce partition-ratio evidence to rebut the presumption.
Id.
¶ 14.
¶ 2 In this ease involving a prosecution under both § 28-1381(A)(1) and (A)(2), the state filed a motion in limine to preclude Joseph Cooperman from introducing at his trial in the Tucson City Court partition-ratio evidence contesting the accuracy of the Intoxilyzer 8000 breath tests. The city court denied the state’s motion, and the state filed a special action challenging that ruling in the superior court. The state now appeals from the superior court’s denial of relief. Relying on
Guthrie,
the state contends the superior court abused its discretion by affirming the city court’s ruling. For the reasons stated below, we affirm.
Factual and Procedural Background
¶ 3 On June 20, 2010, a Tucson police officer cited and arrested Cooperman for DUI while impaired to the slightest degree, in violation of § 28 — 1381(A)(1), and, based on the results of duplicate Intoxilyzer breath tests, for driving or being in actual physical control of a vehicle with an alcohol concentration of .08 or more within two hours of driving, in violation of § 28-1381(A)(2).
¶ 4 Before trial, the state moved to preclude Cooperman from presenting evidence of partition ratios, breath and body temperatures, breathing patterns, and radio frequency interference (RFI)
to contest the accuracy of his breath-test results. At an evidentiary hearing on the motion, the city court heard conflicting testimony from the state’s expert, Michael Sloneker, and the defendant’s expert, Chester Flaxmayer, concerning the effect of using partition ratios and other physiological variables on the accuracy of breath tests. The court issued a thorough, five-page ruling denying the state’s motion, but granting an instruction
limiting the relevancy of the partition-ratio evidence to the (A)(1) charge.
¶ 5 The state then filed a petition for special action with the superior court. The respondent judge accepted jurisdiction but denied relief, affirming the city court’s ruling. The state filed this appeal. We have jurisdiction pursuant to A.R.S. §§ 12-120.21(A)(1), 12-2101(A)(1), and Rule 8(a), Ariz. R.P. Spec. Actions.
Discussion
¶ 6 Under Arizona’s statutory scheme, a defendant accused of drinking and driving or being in actual control of a vehicle can be charged with multiple DUI-related offenses.
See
A.R.S. §§ 28-1381 through 28-1383. Under § 28-1381(A)(1), the state must prove the defendant was driving or in actual physical control of a vehicle “[wjhile under the influence of intoxicating liquor ... if the person is impaired to the slightest degree.” And, for a violation of § 28-1381(A)(2), the state must prove the defendant had “an alcohol concentration of .08 or more within two hours of driving or being in actual physical control of the vehicle.” Cooperman was charged under both subsections.
¶ 7 Arizona’s implied consent law, A.R.S. § 28-1321(A), provides that any person operating a motor vehicle in this state and arrested for DUI “gives consent ... to a test or tests of [his] blood, breath, urine or other bodily substance for the purpose of determining alcohol concentration.” If the test shows the defendant’s alcohol concentration within two hours of the time of driving or being in actual physical control was .08 or more, “it may be presumed that the defendant was under the influence of intoxicating liquor” for purposes of the (A)(1) offense.
§ 28-1381(G)(3). The presumption, however, does not preclude the introduction of “any other competent evidence bearing on the question of whether or not the defendant was under the influence of intoxicating liquor.” § 28-1381(H).
¶ 8 “Alcohol in the breath does not cause impairment; impairment results when aleo-hol enters the body, is absorbed into the bloodstream, and is transported to the central nervous system and the brain.”
Guthrie,
202 Ariz. 273, ¶ 5, 43 P.3d at 602. “[B]reath alcohol readings nonetheless indicate blood alcohol levels, and ... the percentage of alcohol in 100 milliliters of blood could be equated to the percentage of alcohol in 210 liters of breath.”
Id.
¶¶ 5-6. Accordingly, our legislature adopted a 2,100:1 ratio of breath-to-blood alcohol.
Id.
¶ 6. The 2,100:1 partition ratio is “an estimation” and varies among individuals based on several factors, including: body temperature, breathing patterns, blood consistency, and environmental factors.
Id.
¶¶ 7-8. Evidence of such variability, nevertheless, is admissible only in certain circumstances.
¶ 9 For example, in
Guthrie,
this court held partition-ratio evidence is inadmissible in a prosecution under § 28-1381(A)(2). 202 Ariz. 273, ¶ 10, 43 P.3d at 603. Section 28-1381(A)(2) prohibits a person from driving or physically controlling a vehicle if the person has an “alcohol concentration” of .08 or more. And A.R.S. § 28-101(2) defines alcohol concentration, when expressed as a percentage, as either “[t]he number of grams of alcohol per one hundred milliliters of blood,” or “[t]he number of grams of alcohol per two hundred ten liters of breath.” “The statutes thus permit[ ] either a breath alcohol reading or a blood alcohol reading to establish the element of alcohol concentration without regard to the question how the former might be converted to the latter.”
Guthrie,
202 Ariz. 273, ¶ 10, 43 P.3d at 603. Accordingly, in
Guthrie,
we reasoned that the accuracy of the 2,100:1 partition ratio was irrelevant in a prosecution under (A)(2) because “it [i]s illegal to drive or physically control a vehicle if, according to a test within two hours, the number of grams of alcohol per 210 liters of breath exceeded [.08].”
Id.
However, we also held that, in a prosecution under (A)(1), where the state uses the defendant’s breath-test results to establish a presumption of intoxication pursuant to § 28-1381(G), the defendant may introduce evidence challeng
ing the accuracy of the partition-ratio calculation to rebut that presumption.
Id.
¶ 14.
¶ 10 In its petition for special action to the superior court, the state contended the city court had abused its discretion by ruling that (1) the § 28-1381(G) presumption arises by operation of law when test results are admitted in evidence, even when the state “never sought to take advantage of the ... statutory presumption”; (2) once the presumption arises, evidence of the variability of the partition ratio in the general population is relevant and admissible in an (A)(1) ease; and (3) evidence of hematocrit, breathing patterns, and breath and body temperature — and their effect on breath tests — are relevant and admissible in both types of DUI eases. On appeal, the state essentially reurges these arguments, asserting the superior court abused its discretion by affirming the city court’s ruling, which it contends is “contrary to [the] easelaw.” We address each argument in turn.
Appellate Review
¶ 11 When a special action initiated in the superior court is appealed to this court, we must conduct a bifurcated review.
Bazzanella v. Tucson City Court,
195 Ariz. 372, ¶ 3, 988 P.2d 157, 159 (App.1999). First, we consider the superior court’s decision to accept or refuse jurisdiction; second, we consider the merits of the superior court’s decision.
Id.
Here, the court accepted jurisdiction, finding there was “no equally plain, speedy, and adequate remedy available by appeal.” Neither party argues the court erred in so finding, and we agree the state had no avenue to appeal the city court’s order and special-action jurisdiction therefore was appropriate.
See State ex rel. Thomas v. Duncan,
216 Ariz. 260, ¶ 4, 165 P.3d 238, 240 (App.2007) (state can appeal order granting motion to preclude evidence, but not order denying preclusion).
¶ 12 Accordingly, we turn to the merits of the superior court’s decision. “The denial of special action relief is a discretionary decision for the superior court,”
State ex rel. Dean v. City Court of City of Tucson,
123 Ariz. 189, 192, 598 P.2d 1008, 1011 (App.1979), but we review questions of law, such as the interpretation of a statute, de novo,
Files v. Bernal,
200 Ariz. 64, 66, 22 P.3d 57, 59 (App.2001). “Generally, a court abuses its discretion where the record fails to provide substantial support for its decision or the court commits an error of law in reaching the decision.”
Id.
at 65, 22 P.3d at 58. Here, the court denied relief and confirmed the city court’s order, finding it “thorough, well considered and consistent with the evidentiary record and cited case law.”
We conclude the record and the case law provide ample support for the superior court’s decision.
Statutory Presumption
¶ 13 First, we address the state’s argument that evidence regarding the partition-ratio calculation is inadmissible where “the [s]tate does not [rely on] the statutory presumption of impairment.”
Guthrie
permits partition-ratio rebuttal evidence in a prosecution under (A)(1) when the state introduces breath-test results and thereby raises the § 28-1381(G) presumption. But the state apparently suggests that if it does not raise the presumption expressly, then partition-ratio evidence is neither relevant nor admissible. And, it asserts it does not intend to use Cooperman’s breath-test results in order to take advantage here of the statutory presumption on the (A)(1) charge.
¶ 14 Cooperman counters that the state has misinterpreted
Guthrie.
He contends
Guthrie
stands for the proposition that in an (A)(1) prosecution, the statutory presumption effectively is raised whenever the state introduces evidence of alcohol concentration, and a defendant is entitled to rebut that pre
sumption by challenging the breath-to-blood partition ratio. Cooperman points to § 28-1381(H) as evidence of the legislature’s intent to ensure that defendants prosecuted for DUI are afforded their due process right to present a complete defense, which includes introducing evidence to rebut the § 28-1381(G) presumption.
¶ 15 The city court found there is no language in
Guthrie,
or § 28-1381(G), limiting use of the presumption in the manner suggested by the state. Rather, the court concluded the presumption “will always apply” in an (A)(1) prosecution once the state introduces evidence of the Intoxilyzer results to prove the defendant was driving under the influence of intoxicating liquor. In other words, the presumption “does not offer an opt out clause” for the state.
¶ 16 The interpretation of a statute is a question of law we review de novo.
State v. Bolding,
227 Ariz. 82, ¶ 5, 253 P.3d 279, 282 (App.2011). Our primary goal in interpreting a statute is to effectuate the intent of the legislature.
State v. Ross,
214 Ariz. 280, ¶ 22, 151 P.3d 1261, 1264 (App.2007). We look first to the plain language of the statute as the best indicator of that intent and give that language effect when it is clear and unambiguous.
Fragoso v. Fell,
210 Ariz. 427, ¶ 7, 111 P.3d 1027, 1030 (App.2005). We “will not read into a statute something which is not within the manifest intent of the legislature as reflected by the statute itself.”
State v. Ritch,
160 Ariz. 495, 497, 774 P.2d 234, 236 (App.1989).
¶ 17 We find the language of § 28-1381(G) clear and unambiguous. According to that statute, the defendant’s alcohol concentration — as determined by an analysis of his blood, breath, or other bodily substance taken within two hours of driving or being in actual physical control of a vehicle — “gives rise” to certain presumptions. And, contrary to the state’s position, the statutory presumption of intoxication is raised in a prosecution for an (A)(1) offense whenever the state introduces evidence that a defendant had an alcohol concentration of .08 or more. We find nothing in the statute’s language to support the state’s argument that the presumption applies only when expressly invoked by the state.
Our interpretation of the statute is consistent with
Guthrie
which, as the city court points out, “carefully never implies that the [s]tate could choose not to use the presumption” despite introducing alcohol-concentration evidence in an (A)(1) prosecution.
¶ 18 Trial courts have a duty to instruct the jury on the general principles of law that pertain to a criminal offense.
State v. McAl-vain,
104 Ariz. 445, 448, 454 P.2d 987, 990 (1969);
see also State v. Clark,
112 Ariz. 493, 495, 543 P.2d 1122, 1124 (1975) (failure to instruct on matter vital to rights of defendant constitutes fundamental error). This duty exists regardless of who requests a particular instruction.
McAlvain,
104 Ariz. at 448, 454 P.2d at 990. Although the presumption contained in § 28-1381(G) is permissive and “nothing more than an inference,”
State v. Seyrafi,
201 Ariz. 147, ¶ 9, 32 P.3d 430, 433 (2001), the trial judge still has a duty to instruct the jury on this general principle of law pertaining to a prosecution under (A)(1) once evidence is introduced of the defendant’s alcohol concentration.
General Partition-Ratio Evidence
¶ 19 Next, we consider the state’s argument that “any defendant who wishes to challenge the ‘standard’ or ‘generally accepted’ [pai’tition] ratio on which the Intoxilyzer (5000 or 8000) is based, must present evidence of his own ratio at the time of the test.” The state maintains that “the logic [and language] of
Guthrie
” compel that result and “[t]he probative value of any other evidence (such as a ‘hypothetical’ person) is substantially outweighed by the danger of unfair prejudice.” In response, Cooperman argues that
Guthrie
“d[oes] not preclude a defendant from introducing studies or other evidence about variations in the population.”
¶ 20 For purposes of our review, we treat the state’s motion in limine as a motion to suppress,
State v. Rodriguez,
126 Ariz. 28, 30, 612 P.2d 484, 486 (1980), and consider only the evidence presented at the suppression hearing, viewing that evidence in the light most favorable to the city court’s ruling,
State v. Gay,
214 Ariz. 214, ¶ 4, 150 P.3d 787, 790 (App.2007). We are mindful that a trial court has broad discretion to determine the admissibility of evidence and is best suited to conduct any balancing of probative value and prejudicial effect pursuant to Rule 403, Ariz. R. Evid.
State v. Harrison,
195 Ariz. 28, ¶ 21, 985 P.2d 513, 518 (App.1998). On review, we also consider “the evidence in the ‘light most favorable to its proponent, maximizing its probative value and minimizing its prejudicial effect.’”
Id., quoting State v. Castro,
163 Ariz. 465, 473, 788 P.2d 1216, 1224 (App.1989).
¶ 21 We believe the state interprets
Guthrie
too broadly. In that case, we said partition-ratio evidence is relevant and admissible in an (A)(1) but not an (A)(2) prosecution. Although dictum in
Guthrie
may suggest otherwise, we did not address specifically whether a defendant is limited to evidence of his own partition ratio, or whether he may show generally that partition ratios vary from person to person, and even from moment to moment for the same person.
Noting that
Guthrie
does not provide definitive guidance on this issue, the city court relied on cases from other jurisdictions, namely,
State v. Hanks,
172 Vt. 93, 772 A.2d 1087 (2001), and
People v. McNeal,
46 Cal.4th 1183, 96 Cal. Rptr.3d 261, 210 P.3d 420 (2009). And, based on those eases, the court concluded that if “the [sjtate has relied upon or will rely upon an [Ijntoxilyzer result in any way to argue impairment, evidence of the variability of the partition ratio in the general population is relevant and admissible” to challenge the presumption of intoxication in an (A)(1) case. It noted that, because an individual’s partition ratio constantly varies, a calculation of his ratio at some later time would not be relevant to discredit the earlier Intoxilyzer test or the partition ratio it employed.
The court concluded that the general variability of partition ratios makes the evidence relevant and that, although “questions can be raised in individual cases about whether th[e] assum[ed ratio] tends to underestimate most people’s result, the defense has a right to raise the issue, and east doubt upon the theory” that the Intoxilyzer results accurately demonstrate the defendant’s impairment. We agree with the court’s reasoning and conclusion. To be relevant, evidence need only have “any tendency to make a fact more or less probable than it would be without the evidence.” Ariz. R. Evid. 401. And our supreme court has observed that “[t]his standard ... is not particularly high.”
State v. Oliver,
158 Ariz. 22, 28, 760 P.2d 1071, 1077 (1988).
¶ 22 The state nonetheless maintains that evidence regarding variances in partition ratios generally should be precluded because it “could only serve to confuse the issue andWor mislead the jury.” Rule 403 provides that relevant evidence can be excluded if “its probative value is substantially outweighed by a danger of ... unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence.”
¶ 23 As did the city court, we find
Hanks
and
McNeal
instructive. In
Hanks,
the Vermont Supreme Court concluded general partition-ratio evidence was admissible in a prosecution for DWI, similar to our (A)(1) DUI offense, because “any evidence raising a doubt as to defendant’s condition ... is rele
vant.” 772 A.2d at 1092. Although the court acknowledged the general evidence the defendant sought to admit was “not highly probative given the unlikelihood that defendant’s partition ratio ... fell outside the normal curve,” the court specifically rejected the notion that allowing the evidence would confuse the jury.
Id.
The court reasoned:
Evidence on the variability of partition ratios would simply reveal to the jury that the breath-test result is based on a statutorily accepted conversion rate that tends to favor defendants, but that the result is not unassailable with respect to demonstrating impairment. We fail to see how such evidence would be confusing to the jury or unduly prejudicial to the State; to the contrary, not allowing defendants to reveal these scientifically recognized facts would make it difficult, if not impossible, for a defendant to challenge a test result that is admissible in generic [DUI] prosecutions only as a permissive inference on the ultimate question of impairment.
Id.
at 1093. Similarly, in
McNeal,
the California Supreme Court approved the use of general partition-ratio evidence, explaining that all evidence tending to rebut the presumption of intoxication is admissible. 46 Cal.4th at 1200, 96 Cal.Rptr.3d 261, 210 P.3d 420.
¶ 24 We agree with the court in
Hanks
that although the probative value of general partition-ratio evidence usually will be low, 772 A.2d at 1092, we cannot say it is substantially outweighed by the danger of confusing the issues or misleading the jury. Because the ultimate issue in an (A)(1) ease is whether the defendant was driving while “under the influence of intoxicating liquor,” § 28-1381(A)(1), evidence showing the Intoxilyzer may have overstated a defendant’s alcohol concentration, and thus his impairment, could be an important part of the defense, especially if the defendant tested at or near the .08 threshold.
Cf. Fuenning v. Superior Court,
139 Ariz. 590, 598, 680 P.2d 121, 129 (1983) (defendant may offer expert testimony to show “for one reason or another” test results did not establish DUI). Although the Intoxilyzer produces accurate, even understated, results in most cases,
State v. Velasco,
165 Ariz. 480, 485, 799 P.2d 821, 826 (1990), a defendant charged under § 28-1381(A)(l) nonetheless is entitled to introduce reliable evidence challenging the state’s alcohol-concentration evidence in his particular case.
¶ 25 Accordingly, we hold that when a defendant is charged with DUI under § 28-1381(A)(1) and the state introduces evidence of his breath-alcohol concentration at trial, he may offer evidence explaining how partition ratios vary within an individual and among the general population and how that variability may result in breath-test results
that overstate a defendant’s actual level of intoxication. It is then for the jury to decide the weight to be given such evidence in the particular ease.
See State v. Williams,
209 Ariz. 228, ¶ 6, 99 P.3d 43, 46 (App.2004) (“Although the record contains some conflicting evidence, it was for the jury to weigh the evidence and determine the credibility of the witnesses.”). We also reaffirm
Guthrie’s
holding that partition-ratio evidence is inadmissible in an (A)(2) case and agree with the city court that the state is entitled to a limiting instruction to that effect.
We find no abuse of discretion in the superior court’s denial of relief.
Hematocrit, Breathing Patterns, and Breath and Body Temperature
¶ 26 Finally, the state argues that evidence of the possible effect on breath tests of he-matocrit, breathing patterns, and breath and body temperature should be excluded in both (A)(1) and (A)(2) cases unless the defendant can offer evidence of his own physiology at the time of the test. The state maintains the possibility these factors could affect the breath test is not relevant because it “has no basis in fact — as applied to this defendant.” To support its argument, the state asserts the experts agreed that “blood hematocrit (blood concentration) was irrelevant to the issue of what might affect the differences between contemporaneous breath and blood tests”; a properly administered test — one in which the subject is instructed to take a deep breath and blow into the machine as long as he can — would “negate” any potential effect breathing patterns could have on the test results; and one study had found “no direct correlation shown between body temperature, breath temperature and the effect it would have on simultaneous breath and blood tests for alcohol concentration.”
¶ 27 But quoting
State ex rel. McDougall v. Superior Court,
178 Ariz. 544, 546, 875 P.2d 203, 205 (App.1994), Cooperman contends a “defendant may attack the accuracy of a breathalyzer on any relevant ground, including the inherent margin of error.” He claims “evidence of temperature and breathing patterns are independent of partition ratio ... [and] are always relevant and admissible in challenging the § 28-1381(A)(2) charge,” and “requiring the defendant to provide specific evidence of independent physiological variables is improper burden shifting.”
¶ 28 As noted above, evidence is relevant if it has “any tendency to make a fact more or less probable.” Ariz. R. Evid. 401. In reaching its conclusion that evidence of these physiological variables is relevant and admissible in both types of eases — even without defendant — specific evidence — the city court necessarily rejected the opinion of the state’s expert and, instead, relied upon the testimony of defense expert Flaxmayer. Although Sloneker testified that hematocrit “doesn’t make any difference in the simultaneous breath/blood comparisons” based on the results of one study, Flaxmayer opined that “hematocrit can change [either breath or blood alcohol concentration] by about plus or minus five percent” based on the results of a different study. Where the experts’ testimony differed, it was within the court’s discretion to reject Sloneker’s opinion.
State v. Ellison,
213 Ariz. 116, ¶ 32, 140 P.3d 899, 911 (2006).
¶ 29 Flaxmayer also testified that “how you breathe and how you blow into the instrument changes the reading that you obtain on the instrument. It changes your breath alcohol concentration.” As to temperature, he stated that “[t]he amount of alcohol that leaves the blood is a function of [body] temperature”; that studies have shown a statistically significant difference in the
breath temperature among participants; that breath machines were calibrated to thirty-four degrees centigrade and that one degree above or below that level could result in a six-to-eight-percent change in the breath-test result. He stated:
[T]emperature affects the amount of alcohol in your breath. Your breathing affects the amount of alcohol in your breath, regardless [of] if you then take the additional step to use [the] partition ratio to make the calculation. It changed what was in your breath.
¶ 30 The city court noted that Sloneker “did not address this exact issue” of how breathing patterns and temperature may affect breath alcohol readings, other than to state that these factors were related to the partition ratio. And although Flaxmayer agreed that if an individual takes a deep breath and blows into the machine as instructed it should cancel out the potential physiological effects, he also testified “you can’t always tell if somebody’s breathing exactly as instructed.” In sum, Cooperman presented competent expert testimony that these physiological factors, apart from partition-ratio evidence, can impact the ability of the machine to accurately register a defendant’s breath alcohol concentration, which is relevant to both charges. The court thus did not abuse its discretion in concluding the evidence had some tendency to make a fact in issue more or less probable.
Cf. State v. Storholm,
210 Ariz. 199, ¶¶ 11-12, 109 P.3d 94, 96 (App.2005) (defendant has right to obtain blood test to “east doubt on the validity of the breath alcohol concentration results”);
Moss v. Superior Court,
175 Ariz. 348, 352, 857 P.2d 400, 404 (App.1993) (accuracy of breath readings means due process does not require state to preserve breath sample for independent testing; “focus inherently shifts from the breath sample to the machine itself and its proper operation”).
Disposition
¶ 31 For the reasons stated above, the superior court’s ruling is affirmed.
CONCURRING: VIRGINIA C. KELLY, and PETER J. ECKERSTROM, Judges.