¶ 1 This case comes before us on an interlocutory appeal from the district court’s order granting the State of Utah’s motion to compel medication of the defendant, Wanda Barzee. This opinion contains the views of the majority of the court as to all parts of the analysis except Part III.A regarding the appropriate standard of review for the second Sell factor and Part III.D.2. Part III.D.2 of this opinion addresses whether involuntary administration of antipsychotic medication is substantially likely to render Ms. Barzee competent to stand trial. This opinion, as discussed in Part III.D.2, concludes that it is not, but this opinion is not the majority opinion on that issue. The majority opinion on that issue concludes otherwise, as set forth in Justice Durrant’s separate opinion, and thus affirms the district court’s order granting the State’s motion to compel medication. Justice Durrant’s opinion is joined by Associate Chief Justice Wilkins and Justice Parrish. The dissenting view concerning the standard of review for the second Sell factor in Part III.A and the entirety of Part III.D.2 of this opinion is that of myself; Justice Nehring concurs in my result.

BACKGROUND

I. PROCEDURAL HISTORY

¶2 In March 2003 Ms. Barzee and her husband, Brian David Mitchell, were arrested and charged with multiple felonies in connection with their alleged abduction of a minor. Ms. Barzee was charged with aggravated burglary, aggravated sexual assault, aggravated kidnaping, and attempted aggravated kidnaping, or in the alternative, conspiracy to commit aggravated kidnaping. The State filed a petition to inquire into Ms. Barzee’s competency. Two court-appointed evaluators determined that Ms. Barzee was suffering from mental illness of a psychotic nature and that her competency was “severely compromised.” Both evaluators found that, due to the nature of her psychosis, Ms. Barzee had “severe impairments” in her ability to engage in the reasoned choice of legal strategies and options, and thus concluding that Ms. Barzee was “severely impaired with respect to her present capacity to consult with her counsel and participate in the proceedings against her with the reasonable degree of rational understanding.” *51The district court concluded that Ms. Barzee was not competent to proceed. Ms. Barzee was then transferred to the Utah State Hospital, where she currently remains.

¶ 3 Since its initial ruling, the district court has conducted two hearings to review Ms. Barzee’s competence. After the first review healing in August 2004, the district court determined that while Ms. Barzee was still incompetent to stand trial, there was a “substantial probability that [she] may become competent in the foreseeable future.” One year later, after the second review hearing, the district court concluded that Ms. Barzee remained incompetent. Following that hearing, the Salt Lake District Attorney’s Office filed a motion to compel medication.1 At the Medication Hearing, the district court heard testimony from Drs. Kreg Jeppson, Paul Whitehead, Raphael Morris, and Xavier Amador; the court subsequently granted the State’s motion to compel medication. Ms. Barzee filed this interlocutory appeal; we have jurisdiction pursuant to Utah Code section 78-2-2(3)(h) (2002).

II. DIAGNOSES AND OPINIONS OF THE MENTAL HEALTHCARE PROFESSIONALS

¶ 4 In all, eight mental health care professionals were involved in this case. All eight agree that Ms. Barzee suffers from a psychotic disorder with the primary feature of nonbizarre grandiose delusions.2 However, Ms. Barzee’s precise diagnosis is in dispute, as is the question of whether antipsychotic medication is likely to render Ms. Barzee competent. Initially, two evaluators were appointed to determine if Ms. Barzee was competent to stand trial. Each offered an opinion on Ms. Barzee’s diagnosis.

¶ 5 First, Dr. Jeffrey A. Kovniek, a psychiatrist and court-appointed competency evaluator, diagnosed Ms. Barzee with shared psychotic disorder3 because of the development of her delusions during her relationship with Mr. Mitchell, who Dr. Kovniek believed was the dominant individual. He also opined that she qualified for a diagnosis of delusional disorder.4 According to Dr. Kovniek, Ms. Barzee suffers from nonbizarre delusions, delusions of reference, no verbal or communication symptoms apparent in schizophrenia, and no other psychotic symptoms; thus, shared psychotic disorder or delusional disorder, rather than schizophrenia, is the proper possible diagnosis. He found her incompetent to stand trial because of the impairment in her capacity to engage in reasoned choice of legal strategies and options. He suggested that Ms. Barzee’s treatment should include medication and that antipsychotic drugs would increase the likelihood that she would become competent.

¶ 6 Second, Dr. Nancy B. Cohn, a psychologist and court-appointed competency evaluator, diagnosed Ms. Barzee with schizophrenia,5 paranoid type, a conclusion not shared by any of the seven other experts who were asked to give opinions regarding Ms. Bar-*52zee’s mental health. Dr. Cohn reported symptoms including cognitive disorganization symptomatic of thought disorder, paranoid ideation, hallucinations, delusions, and referential thinking. Dr. Cohn also noted that there was no evidence to suggest that a head injury or substance abuse precipitated Ms. Barzee’s current condition. She stated that Ms. Barzee was incompetent to stand trial based on impairment in her ability to testify relevantly, in her capacity to communicate with her attorneys, and in her ability to make reasoned choices because her decisions are driven by her religiously based delusions. She noted that “[pjsychotropic medications have been minimally useful in diminishing delusional thinking in certain kinds of psychotic disorders, but it is not entirely clear that medication would be helpful in addressing Ms. Barzee’s deeply entrenched, delusional belief, as these are the symptoms that are most refractory to pharmacological intervention.”

¶7 After the initial evaluations, review hearings were held, and the district court heard from two additional experts, Dr. Gerald Berge, a psychologist, and Dr. Eric Nielsen, a social worker. Each evaluator expressed an opinion on Ms. Barzee’s progress toward competency and her diagnosis.

¶ 8 In a report dated July 2004, and in testimony before the court in August 2004, Dr. Berge stated that he agreed with Dr. Kovnick’s diagnosis of shared psychotic disorder. He disagreed with Dr. Cohn’s evaluation, stating that the symptoms were not as severe as Dr. Cohn suggested in her report. For example, Dr. Berge opined that Ms. Barzee does not suffer from hallucinations or pronounced disorganization in thinking. Like Dr. Kovnick, Dr. Berge’s diagnosis of shared psychotic disorder was connected to Ms. Barzee’s relationship with Mr. Mitchell. Dr. Berge stated, however, that he would shift the diagnosis to delusional disorder if tile delusions continued despite lengthy separation from Mr. Mitchell. In Dr. Berge’s opinion, Ms. Barzee remained incompetent to stand trial because nonbizarre delusions, the primary feature of her disorder, continued to impair her capacity to make reasoned choices regarding her legal options. Dr. Berge stated that the effect of medication on a patient like Ms. Barzee is “controversial,” but noted that if the delusions were eliminated, she would likely become competent.

¶ 9 One year later, at another review hearing, Dr. Nielsen diagnosed Ms. Barzee with psychotic disorder not otherwise specified (PDNOS).6 Dr. Nielsen stated that the primary feature of her illness is grandiose religious delusions and that she also suffers from anosognosia, or a lack of insight into her mental illness. He opined that Ms. Bar-zee’s disorder could not be shared psychotic disorder because despite separation from Mr. Mitchell and her decision, with God’s influence, to “leave him behind,” her delusions continued. Dr. Nielsen further opined that she does not suffer from schizophrenia because many of the symptoms associated with that diagnosis are not present in Ms. Bar-zee’s ease. He stated that the delusions affect her functioning generally and that, in his opinion, this was not typical of delusional disorder. He reported that, due to her psychosis, Ms. Barzee continued to be incompetent to stand trial.

¶ 10 Dr. Nielsen expressed the opinion that if Ms. Barzee has delusional disorder, the condition is refractory and rarely treatable with medication but that schizophrenia generally has some response to antipsychotic medication. He pointed out that symptoms of a delusional nature do not respond favorably to medication and that Ms. Barzee’s long duration of untreated psychosis suggests a poorer prognosis but that the schizophrenic symptoms of thought disorder — ram*53bling and vagueness — may suggest a more favorable response to the drugs.

¶ 11 Because Ms. Barzee’s incompetence was not improving without medication and because Ms. Barzee refused to be medicated, the State moved to medicate her forcibly. The Medication Hearing was held in February 2006. Four experts testified as to Ms. Barzee’s condition and the likely effects of medication.

¶ 12 First, Dr. Kreg Jeppson, a psychiatrist at the Utah State Hospital, Ms. Barzee’s treating physician, and the State’s primary witness, based his opinions on his experience at the state hospital where he treats many schizophrenic and bipolar patients, some patients with PDNOS, and very few patients suffering from delusional disorder. Dr. Jeppson stated that he was a “clinician,” not an “expert witness,” researcher, or peer-reviewer. As Ms. Barzee’s treating physician, he met with her “weekly to monthly” over a twenty-two-month period and testified that “a lot of times these aren’t 45-minute talks, I mean just a few minutes here and there.” Ms. Barzee met weekly with her social worker, but refused to meet with her “treatment team” for the ten months prior to the Medication Hearing.7

¶ 13 Dr. Jeppson testified that Ms. Bar-zee’s symptoms included grandiose delusions, some persecutory delusions, and anosognosia. He originally diagnosed her with delusional disorder, but in January 2005 he changed her diagnosis to PDNOS. He testified that this change was a move from a more specific diagnosis to a more general diagnosis because he had acquired “more facts.” Specifically, Ms. Barzee exhibited “referential thinking,”8 which he did not consider a part of delusional disorder. Even before Dr. Jeppson made his initial diagnosis of delusional disorder, however, he was aware that Drs. Kovniek, Cohn, and Berge had identified referential thinking as one of Ms. Barzee’s symptoms, and Dr. Jeppson knew that Ms. Barzee discussed receiving answers to prayers through movies “a long time ago.” Moreover, the one (and only) episode of referential thinking that Dr. Jeppson could point to when he testified happened months before he changed her diagnosis.

¶ 14 Dr. Jeppson concluded that restoration to competency was likely within eight to twelve months of initiating medication. He stated that “hopefully, [Ms. Barzee] would be restored to competency” and suggested that she had a seventy to eighty percent chance of restoration with medication.

¶ 15 Second, Dr. Paul D. Whitehead, also a psychiatrist at the state hospital, testified on behalf of the State. His experience included restoring over 100 patients to competency. However, with Ms. Barzee, Dr. Whitehead’s experience consisted of one meeting in November 2004. He was consulted to determine if she met the United States Supreme Court’s Harper standard for forcible medication. See Washington v. Harper, 494 U.S. 210, 110 S.Ct. 1028, 108 L.Ed.2d 178 (1990). He determined that Ms. Barzee could not be medicated under that standard because she was not dangerous or gravely disabled. Dr. Wdiitehead also discussed the Sell factors, which allow compelled medication for the sole purpose of rendering a defendant competent to stand trial.

¶ 16 Dr. Whitehead opined that delusional disorder was an appropriate diagnosis. He also testified that, given Ms. Barzee’s lack of cooperation and the inability to rule out other causes, “which are extremely unlikely,” such as a brain tumor and seizure disorder, PDNOS was also a reasonable diagnosis. Dr. Whitehead stated that while Ms. Barzee meets the definition of delusional disorder under the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), he thought she had more symptoms than a typi*54cal patient with that diagnosis. He indicated, however, that delusional disorder is “very-rare,” and that he had seen only four or five cases.

¶ 17 Dr. Whitehead cautioned that he did not like placing a number on the likelihood of restoration through medication, but that “a very conservative, even skeptical number [would be] fifty to sixty percent.” He continued, stating that the case is nowhere “close to a one hundred percent case” for restoration, but he thought “chances are better than not that she will be restored.” A gross estimate in his opinion was seventy percent.

¶ 18 Third, Dr. Raphael Morris, a psychiatrist, testified for the defense. Dr. Morris has experience in restoring patients to competency in his work at state hospitals and in private practice. He successfully restored approximately thirty patients to competency over a two-year period in the intensive management unit of his hospital. His experience with Ms. Barzee included one brief meeting and a review of her records.

II19 Dr. Morris did not offer his own opinion of Ms. Barzee’s diagnosis, but he did explain the diagnosis of PDNOS to the court. He testified that PDNOS is used when a patient suffers from a psychotic disorder and the practitioner is unable to rule out a medical condition or substance abuse as the cause of the disorder. Thus, PDNOS is a “differential diagnosis,” which means that the disorder is either substance-induced, resulted from another medical condition, or is a primary psychotic disorder. See DSM-IV 311. Because Ms. Barzee does not have a history of substance abuse, and there was no evidence that a medical condition was causing her symptoms, Dr. Morris concluded that she suffers from either schizophrenia or delusional disorder. Dr. Morris stated that he is a “stickler! ] at trying to pin down a diagnosis” and would accept PDNOS as the diagnosis for only the first two weeks after a patient’s admission.

¶20 Dr. Morris opined that Ms. Barzee had a twenty to thirty-five percent chance of being rendered competent with medication— the lower end representing the likelihood of restoration if she suffers from delusional disorder and the higher end representing the rate if she suffers from schizophrenia.

¶21 Fourth, Dr. Xavier Amador testified on behalf of Ms. Barzee. Dr. Amador is a clinical psychologist, a professor, and a peer-reviewer for psychiatric journals. He has authored books and articles on psychotic disorders and acted as the co-chair in revising portions of the current edition of the DSM-IV that define and explain psychotic disorders. In fact, Dr. Amador co-wrote the first draft definitions for all of the psychotic disorder terms in the DSM-IV, which is the authoritative tool for diagnosis of mental disorders. He stated that the difference between himself, a psychologist, and the other witnesses at the Medication Hearing, all psychiatrists, is the foundation of his training in research, along with the clinical aspects of treatment. Dr. Amador noted the importance of anchoring opinions in empirical evidence provided by research, rather than relying only on personal clinical experience to determine the efficacy of different treatments because all practitioners work with unique groups, and the experience from those groups can skew a practitioner’s view of treatment generally.

¶ 22 Dr. Amador’s experience is not with restoring patients to competency, but he does perform competency evaluations. His specific experience with Ms. Barzee entailed visiting with her multiple times over the past few years, spending over fifteen total hours with her, and reviewing her medical records.

¶23 Dr. Amador was perplexed by Dr. Jeppson’s diagnostic change from delusional disorder to PDNOS, because PDNOS is a diagnosis normally reserved for patients for whom there is a lack of information, and not, in Dr. Amador’s opinion, for a patient for whom there is a two-year period of observation. Dr. Amador agreed with the original diagnosis of delusional disorder with grandiose and persecutory delusions. He noted other important factors to consider in assessing the efficacy of medication for Ms. Barzee, such as the long duration of her untreated psychosis — at least ten to thirteen years— her grandiose delusions, and her lack of insight.

*55¶ 24 Dr. Amador suggested a “low likelihood” of restoring Ms. Barzee to competency through administration of antipsychotic medication. He suggested that the likelihood of restoration was twenty percent, increasing slightly to thirty or forty percent if a diagnosis other than delusional disorder were considered.

¶ 25 After hearing the foregoing testimony, the district court granted the State’s motion to forcibly medicate Ms. Barzee.

ANALYSIS

¶26 This case involves the State’s ability to involuntarily medicate a criminal defendant for the sole purpose of rendering her competent to stand trial. This is a case of first impression for Utah courts and thus presents our first opportunity to apply the analysis of the United States Supreme Court in Sell v. United States, 539 U.S. 166, 123 S.Ct. 2174, 156 L.Ed.2d 197 (2003). Before we begin our federal constitutional analysis under Sell, we pause to discuss two preliminary matters: Ms. Barzee’s state constitutional claim, and the applicability of the Harper standard to Ms. Barzee’s case. See Washington v. Harper, 494 U.S. 210, 110 S.Ct. 1028, 108 L.Ed.2d 178 (1990).

I.MS. BARZEE FAILED TO PRESERVE HER STATE CONSTITUTIONAL CLAIM

¶ 27 Ms. Barzee’s state constitutional claim was not preserved below. At no time prior to appellate briefing in this case did she request that the district court decide this issue as a matter of state law, see State v. Tiedemann, 2007 UT 49, ¶ 33, 162 P.3d 1106, nor did she elaborate on the separate state constitutional issue at oral argument. We decline to address whether the exceptional circumstances doctrine would warrant consideration of the unpreserved State claims, and wait for another day to determine whether Utah’s constitution forbids forcibly medicating a mentally ill defendant for the sole purpose of achieving competency to stand trial.

II.THE HARPER STANDARD IS INAPPLICABLE TO MS. BARZEE’S CASE

¶ 28 “[B]efore turning to the trial competence question,” we must “determine whether forced administration of drugs can be justified on ... alternative grounds.” Sell v. United States, 539 U.S. 166, 182, 123 S.Ct. 2174, 156 L.Ed.2d 197 (2003) (emphasis omitted). In Washington v. Harper, the United States Supreme Court determined that involuntarily medicating a mentally ill criminal defendant could be constitutionally permissible if the inmate is a danger to herself or others or if the inmate’s refusal to take drugs poses grave risks to her health. 494 U.S. 210, 225-26, 110 S.Ct. 1028, 108 L.Ed.2d 178 (1990). Harper does not apply to this case because the parties agree that Ms. Barzee is not gravely disabled and she is not a danger to herself or others while confined at the state hospital. Thus, we proceed with an analysis under Sell.

III.SELL ANALYSIS DETERMINING WHETHER THE STATE DEMONSTRATED THAT INTRUSION UPON MS. BARZEE’S LIBERTY INTEREST IN FREEDOM FROM UNWANTED ADMINISTRATION OF ANTIPSYCHOTIC MEDICATION IS PERMISSIBLE

¶29 Defendants “possess[] a significant liberty interest in avoiding the unwanted administration of antipsychotic drugs under the Due Process Clause of the Fourteenth Amendment” of the United States Constitution. Washington v. Harper, 494 U.S. 210, 221-22, 110 S.Ct. 1028, 108 L.Ed.2d 178 (1990); see also Riggins v. Nevada, 504 U.S. 127, 133-34, 112 S.Ct. 1810, 118 L.Ed.2d 479 (1992) (recognizing an individual’s liberty interest in avoiding forced administration of antipsychotic medication). In Sell v. United States, the United States Supreme Court outlined “limited circumstances” under which the government may “administer antipsy-chotic drugs involuntarily to a mentally ill criminal defendant[ ] in order to render that defendant competent to stand trial.” 539 U.S. 166, 169, 123 S.Ct. 2174, 156 L.Ed.2d *56197 (2003). The four-factor test created by the Court is as follows:

First, a court must find that important governmental interests are at stake.... Second, the court must conclude that involuntary medication will significantly further those [ ] state interests [by finding] that administration of the drugs is substantially likely to render the defendant competent to stand trial [and] ... that administration of the drugs is substantially unlikely to have side effects that will interfere significantly with the defendant’s ability to assist counsel in conducting a trial defense, thereby rendering the trial unfair. Third, the court must conclude that involuntary medication is necessary to further those interests.... Fourth, ... the court must conclude that administration of the drugs is medically appropriate.

Id. at 180-81, 123 S.Ct. 2174 (citations omitted); see also Utah Code Ann. § 77-15-6.5(4)(d)(i)-(iv) (Supp. 2006). The State is required to prove each of these factors by clear and convincing evidence. Id. § 77-15-6.5(6)(a); see also United States v. Bradley, 417 F.3d 1107, 1113-14 (10th Cir. 2005).

¶ 30 With respect to the second factor, the majority concludes that the administration of medication is substantially likely to render Ms. Barzee competent to stand trial. The separate opinion of Justice Durrant, joined by Associate Chief Justice Wilkins and Justice Parrish, contains the analysis supporting the majority conclusion. Section D.2 of this Part III contains my dissenting view, concluding that the evidence does not show that the administration of medication is substantially likely to render Ms. Barzee competent. Justice Nehring concurs in my dissenting view. We begin by addressing the standard of review applicable to the different Sell factors. Then we discuss the first, third, and fourth factors of the Sell analysis before moving on to examine the second Sell factor.

A. Standard of Review

¶ 31 While Sell established the test to determine whether antipsychotic medication can be involuntarily administered to a defendant solely to make her competent to stand trial, it did not outline the standards of review for appellate courts considering the matter. Other courts have considered the issue. For example, according to the Court of Appeals for the Second Circuit, “[w]hether the Government’s asserted interest is important is a legal question that is subject to de novo review.” United States v. Gomes, 387 F.3d 157, 160 (2d Cir. 2004). We agree that this is the appropriate standard for this issue. We also agree with the Second Circuit’s conclusion that the third and fourth factors— necessity and medical appropriateness — are “factual in nature and ... therefore subject to review for clear error.” Id. Where the second factor is concerned, however, there is disagreement among the members of this court. For the majority view, refer to the separate opinion of Justice Durrant.

¶ 32 My disagreement with the majority stems in part from the cursory review that they apply to the issue of whether medication will “significantly further” the State’s interest. Sell, 539 U.S. at 180-81, 123 S.Ct. 2174. In order for the State’s interest to be “significantly furthered,” medication must be “substantially likely” to render Ms. Barzee competent to stand trial, and it must be “substantially unlikely” to interfere with her fair trial rights. Id. at 181, 123 S.Ct. 2174. The focus of my opinion is on the first subis-sue, and I will limit my analysis accordingly.

¶ 33 I disagree with the majority that the second Sell factor is merely a question of fact. Instead, “[r]ecognizing the vital constitutional liberty interest at stake,” Bradley, 417 F.3d at 1114, I believe that heightened appellate scrutiny should be applied to the mixed question of law and fact encompassed in the inquiry of whether “involuntary medication will significantly further” the State’s interests because administration of the drugs is substantially likely to render the defendant competent to stand trial, Sell, 539 U.S. at 181, 123 S.Ct. 2174 (emphasis omitted). This second factor and its first subissue present a mixed question of law and fact because, while the inquiry is highly dependent upon diagnoses and factors influencing a particular patient’s likely reaction to medication, the relevant standard by which to weigh these facts is “competency.” Competency is necessarily a legal concept. In Utah, many factors *57must be considered when analyzing whether a defendant is competent to stand trial. In this ease, the most pertinent factor is whether Ms. Barzee is substantially likely to be rendered capable of engaging in reasoned choice regarding legal strategies and options. Expert testimony will enlighten the court regarding her diagnosis and the types of improvement that may result from administration of antipsychotic medication, but it is up to this court to determine whether those facts will place Ms. Barzee in a mental state adequate to render her competent to stand trial — a legal concept.

¶ 34 The appropriate standard of review for a mixed question of law and fact is assessed under the test set forth in State v. Levin, 2006 UT 50, 144 P.3d 1096. Applying this test, I conclude that “more rigorous appellate scrutiny” is mandated to protect defendants against constitutional deprivations of liberty. Jensen v. Sawyers, 2005 UT 81, ¶ 95, 130 P.3d 325. Levin requires us to consider three factors when determining the appropriate standard of review for mixed questions of law and fact:

(1) the degree of variety and complexity in the facts to which the legal rule is to be applied; (2) the degree to which a trial court’s application of the legal rule relies on “facts” observed by the trial judge, “such as a witness’s appearance and demeanor, relevant to the application of the law that cannot be adequately reflected in the record available to appellate courts;” and (3) other “policy reasons that weigh for or against granting discretion to the trial courts.”

2006 UT 50, ¶ 25, 144 P.3d 1096 (citation omitted).

¶ 35 In determining whether involuntary medication is substantially likely to render a defendant competent, the pertinent evidence will most likely be in the form of expert opinion testimony by mental health professionals. This testimony will likely encompass mental health diagnoses, the results of research on the efficacy of medication, and factors specific to the defendant’s condition. Certainly, this type of information may be complex,9 and the credibility of witnesses may occasionally play a role in conclusions drawn by a judge as to the weight of testimony. However, much of the necessary information concerning the diagnoses and research “generally can be adequately reflected in a cold record” and will be supported by statistics and published research, unaffected by an individual’s appearance and demeanor while testifying. Id. ¶ 40. While determinations about credibility and complex evidence might weigh in favor of granting some level of deference to the trier of fact, the policy concerns that I address next clearly tip the balance in favor of indepth appellate review.

¶ 36 The issue of forcibly medicating a defendant for the sole purpose of making her competent to stand trial implicates constitutional liberty interests of the highest degree. Harper, 494 U.S. at 221, 110 S.Ct. 1028. Thus, we should be concerned with “promoting clarity and consistency in our state’s jurisprudence” and “offering predictable constitutional protections.” Levin, 2006 UT 50, ¶ 41, 144 P.3d 1096. When involuntary administration of antipsychotic medication is involved, this court must “define the boundaries of a substantial constitutional right.” Id. ¶ 44. With the weight of the constitutional policy considerations foremost among my concerns, I believe this court’s role must be to scrutinize the entire record upon which such decisions are made.10 As the Supreme *58Court has noted in the First Amendment context, “When the standard governing the decision of a particular case is provided by the Constitution,” the appellate court’s role is of “special importance” and “the stakes ... are too great to entrust them finally to the judgment of the trier of fact.” Bose Corp. v. Consumers Union of U.S., Inc., 466 U.S. 485, 499, 501 n. 17, 503, 104 S.Ct. 1949, 80 L.Ed.2d 502 (1984); cf. Turner v. Arkansas, 407 U.S. 366, 368, 92 S.Ct. 2096, 32 L.Ed.2d 798 (1972) (applying de novo standard of review to constitutional facts in collateral estoppel claims that implicate the “Fifth Amendment’s double jeopardy guarantee”). In my opinion, the “intermingling of law and fact” that is implicated in this case compels this court to review the district court’s decision de novo “in order to preserve the precious liberties established and ordained by the Constitution.” Bose, 466 U.S. at 509, 511, 104 S.Ct. 1949.11

¶ 37 Thus, the weight of the Levin test “dictate[s] that the application of the legal concept should be strictly controlled by the appellate courts.” Levin, 2006 UT 50, ¶ 23, 144 P.3d 1096. Appellate courts have “an obligation to make an independent examination of the whole record in order to make sure that the judgment does not constitute a forbidden intrusion on” an individual’s protected liberty interest in freedom from unwanted administration of antipsychotic medication. Bose, 466 U.S. at 499, 104 S.Ct. 1949 (internal quotation marks omitted). Thus, I believe that the question of whether a defendant is substantially likely to be rendered competent should be reviewed de novo.12

¶ 38 My conclusion is further bolstered by the Supreme Court’s conclusion that instances where forcible medication is administered solely to make a defendant competent to stand trial should be “rare” and occur only in “limited circumstances.” Sell, 539 U.S. at 169, 180, 123 S.Ct. 2174. After review of the testimony and record in this case and a review of all other cases applying the Sell test, I believe that it is clear that instances when the third and fourth factors of the Sell test — necessity and medical appropriateness — will not be met are themselves likely to be rare or even nonexistent.13 Because *59the Sell test should be used to invade an individual’s liberty interest on only rare occasions, if some factors of the test are in reality almost automatically present, courts must consider the remaining portions of the test with utmost care. In my opinion, this observation requires courts to closely examine the question of whether medication will significantly further the State’s interest in rendering the defendant competent to stand trial.

B. The First Sell Factor: Important Governmental Interest

¶ 39 The initial inquiry under the Sell analysis is whether “important governmental interests are at stake.” Sell v. United States, 539 U.S. 166, 180, 123 S.Ct. 2174, 156 L.Ed.2d 197 (2003). When involuntary medication is at issue, the governmental interest will be in bringing the accused to trial. Id. (citing Riggins, 504 U.S. at 135-36, 112 S.Ct. 1810 (1992) (“Power to bring an accused to trial is fundamental to a scheme of ‘ordered liberty’ and prerequisite to social justice and peace.” (citations and internal quotation marks omitted))). In order to do so, the defendant must be charged with a “serious crime” and “special circumstances,” such as lengthy civil commitment, must not diminish the government’s interest in bringing the defendant to trial. Id. This standard recognizes the State’s interest in protecting “the basic human need for security” undeterred by lost evidence or faded memories that can inhibit effective prosecution when a defendant regains competence after years of commitment. Id.

¶ 40 In Ms. Barzee’s case, she is charged with six first degree felony offenses, each potentially punishable by life sentences, and one second degree felony, punishable by up to fifteen years in prison. In United States v. Gomes, the Court of Appeals for the Second Circuit recognized that “ ‘evident from the substantial sentence’ ” a defendant faces if convicted is the “ ‘seriousness of the crime and [the] perceived dangerousness to society.’” 387 F.3d 157, 160 (2d Cir. 2004) (quoting United States v. Gomes, 289 F.3d 71, 86 (2d Cir. 2002)). Other jurisdictions have concluded that a maximum statutory term of imprisonment of ten years is sufficiently serious to create an important government interest in bringing an accused to trial. See, e.g., United States v. Evans, 404 F.3d 227, 238 (4th Cir. 2005); United States v. Archuleta, 2006 WL 2476070, at *2, 2006 U.S. Dist. LEXIS 63526, at *6 (D.Utah 2006), aff’d 218 F. App’x 754 (10th Cir. 2007). We are convinced that Ms. Barzee is charged with crimes sufficiently serious to evidence an important state interest in prosecution.

¶ 41 Furthermore, we conclude that special circumstances do not diminish this interest. It is unclear whether civil commitment is an option for Ms. Barzee. As long as she is confined in the state hospital, the parties agree that she is not a danger to herself or others. We presume the State would argue that, outside of that environment, she does pose a danger, but there is no certainty that she will be subjected to lengthy confinement through a civil proceeding. We cannot conclude that potential and speculative civil commitment undermines the State’s interest in bringing Ms. Barzee to trial for such serious offenses. Nor can we conclude that the limited time Ms. Barzee has already spent confined at the state hospital in any way undermines the State’s interest. Even if the time already spent in confinement were credited to a future sentence, it would constitute only a fraction of the potential sentence she faces if convicted. See, e.g., United States v. Bradley, 417 F.3d 1107, 1117 (10th Cir. 2005) (recognizing that nine months of confinement “pales in comparison to the fifty years [of] imprisonment [defendant] faces”); United States v. Rivera-Guerrero, 426 F.3d 1130, 1143 (9th Cir. 2005) (stating that a three-year credit toward a possible two-year sentence undermined the government’s interest in prosecution). Thus, Ms. Barzee is charged with serious crimes creating an important State interest in timely prosecution that is not undermined by her past or potential future confinement in the state hospital.

*60C. The Third Sell Factor: Whether Anti-psychotic Medication Is Necessary to Further the State’s Interest; and the Fourth Sell Factor: Whether Administration of Antipsychotic Medication Is Medically Appropriate

¶ 42 Sell requires that the administration of antipsychotic medication be “in the patient’s best medical interest in light of [her] medical condition.” Sell v. United States, 539 U.S. 166, 181, 123 S.Ct. 2174, 156 L.Ed.2d 197 (2003). Furthermore, involuntary medication must be necessary to further the state’s interest in making a defendant competent. In other words, a “court must find that any alternative, less intrusive treatments are unlikely to achieve substantially the same results.” Id. In Ms. Barzee’s case, all of the experts at the Medication Hearing agreed that if she were their patient, they would want to treat her with medication. Dr. Morris stated that he would “absolutely” encourage Ms. Barzee to take medication. Dr. Amador testified that he “would do everything in [his] power to convince a patient to take medication if they had never” before been medicated even if there was only a five-percent chance of a successful outcome. Dr. Jeppson believed medication was in Ms. Barzee’s best medical interest because it might allow her to “have a more full life, a more functional life, and be able to proceed with her case.” Additionally, Dr. Jeppson testified that no less intrusive means of treatment were likely to restore her to competency, and that any progress toward restoration would require medication. Dr. Whitehead testified that he believed medication would increase her ability to function in general and would “help her better relate to her family.” He reiterated Dr. Jeppson’s testimony that less intrusive means of restoration had not and would not be successful at restoring Ms. Barzee to competency.

¶ 43 In Sell, the Court stated that the “specific kinds of drugs at issue” should be considered. Id. The other witnesses did not testify that they had any concerns with the drug regimen suggested by Dr. Jeppson, although Dr. Whitehead preferred using a different antipsychotic medication, and Dr. Morris would possibly have used different dosages. However, Dr. Amador testified that he would use “motivational enhancement therapy” to encourage Ms. Barzee to take medication rather than simply forcing it upon her through a “doctor knows best” approach. While treatment with antipsychotic medication would be medically appropriate for Ms. Barzee “in a vacuum,” the trauma associated with “treating] people against their will” could result in harmful side effects including depression, stress reaction, and post-traumatic stress disorder. Even in light of these cautionary statements by Dr. Amador, we conclude that the district court did not clearly err in its conclusions that treatment with antipsychotic medication is medically appropriate in Ms. Barzee’s case and that less intrusive means of treatment are unlikely to accomplish restoration to competency.

D. The Second Sell Factor: Whether Involuntary Medication Will Significantly Further the State’s Interest

¶ 44 The inquiry into whether the administration of involuntary medication will significantly further the State’s interests in rendering Ms. Barzee competent to stand trial requires us to consider two issues: (1) whether medication is “substantially unlikely to have side effects that will interfere significantly with the defendant’s ability to assist counsel in conducting a trial defense,” and (2) whether medication is “substantially likely to render the defendant competent.” Sell v. United States, 539 U.S. at 181, 123 S.Ct. 2174 (emphasis added). Before we address each of these issues, we pause to discuss the meaning of the substantially likely or substantially unlikely standards in this context.

1145 We read “substantially likely” within the context of the greater question that it is designed to address: whether the State’s interest in a competent defendant will be significantly furthered through involuntary medication. This leads us to the conclusion that “substantially likely” requires the likelihood of restoration to be significant, rather than requiring merely “some” likelihood of restoration. This conclusion is in keeping with that reached by other courts that have considered the issue. In United *61States v. Gomes, 387 F.3d 157, 161-62 (2d Cir. 2004), a seventy-percent chance at restoration to competence was considered substantially likely; in United States v. Ghane, 392 F.3d 317, 320 (8th Cir. 2004), a ten percent chance of restoration, described as a “glimmer of hope,” was held to be inadequate to meet this standard. Other courts have determined that “a chance of success that is simply more than a 50% chance of success does not suffice to meet this standard.” United States v. Rivera-Morales, 365 F.Supp.2d 1139, 1141 (S.D.Cal. 2005); see also People v. McDuffie, 144 Cal.App.4th 880, 50 Cal.Rptr.3d 794, 798 (2006) (holding that a fifty- to sixty-percent chance of “improving” did not meet the substantially likely standard). We agree; the substantially likely standard requires that the chance for restoration to competency be great. To the extent that such a likelihood can be quantified, it should reflect a probability of more than seventy percent. Likewise, in order for side effects to be considered substantially unlikely to interfere with a defendant’s right to a fair trial, any side effect that would impede a defendant’s ability to assist in her defense must have a very low rate of occurrence.

1. Although Side Effects Exist, Administration of Antipsychotic Medication Is Substantially Unlikely to Produce Side Effects that Will Interfere with Ms. Bar-zee’s Right to a Fair Trial

¶ 46 Many side effects can result from the administration of antipsychotic medications, but those side effects that can be quite severe and could impede Ms. Bar-zee’s right to a fair trial are rare. Thus, we hold that the district court did not clearly err in concluding that antipsychotic medication is substantially unlikely to interfere with Ms. Barzee’s right to a fair trial.

¶ 47 The side effects considered likely by Drs. Jeppson and Whitehead include fatigue; dry mouth; blurry vision; constipation; ort-hostatic hypotension; and metabolic syndrome, which would require monitoring for weight gain, lipid profile, cholesterol, and diabetes. According to Drs. Jeppson and Whitehead, those side effects would not interfere with Ms. Barzee’s ability to assist in her defense, nor would they interfere with her abilities to consult with her attorneys, engage in reasoned choice of legal options, recall memories, or testify relevantly. Dr. Morris generally agreed with these likely side effects but added cardiac problems to the list and stated that weight gain is likely and could lead to other health problems.

¶ 48 The physicians also recognized other unlikely side effects that can occur and that could have profound effects on Ms. Barzee’s ability to assist with her defense. Tardive dyskinesia creates tics, can be irreversible, and would impair Ms. Barzee’s ability to assist defense counsel. According to Dr. Jeppson, tardive dyskinesia occurs in only one patient out of 200, and according to Dr. Whitehead, it occurs in two to five percent of patients with continuous exposure to antipsy-chotic medications, more often in women than men. Another unlikely side effect is increased risk of stroke, but this generally occurs only in older patients, not patients within Ms. Barzee’s age group. Even though these side effects could dramatically impair Ms. Barzee’s ability to assist with her defense, because of the low probabilities associated with them, the district court’s conclusion that the side effects are substantially unlikely was not clear error.

¶ 49 Drs. Jeppson and Amador also recognized additional side effects that are of concern. Due to the nature of Ms. Barzee’s psychotic disorder and the connection her delusions have to her identity, a medication that could possibly alter these fixed beliefs and insult her identity may make Ms. Barzee vulnerable to depression, suicidal ideation, stress reaction, or posttraumatic stress disorder. These effects could significantly interfere with Ms. Barzee’s ability to assist counsel. Because, as I will discuss below, I am persuaded that the drugs are unlikely to alter her fixed delusions, I consider these side effects associated with a crisis in her identity substantially unlikely. The majority does- not agree with my conclusion that the drugs are unlikely to alter Ms. Barzee’s fixed delusions, but similarly concludes that these side effects identified by Drs. Jeppson and Amador are substantially unlikely.

*622. Administration of Antipsychotic Medication Is Not Substantially Likely to Render Ms. Barzee Competent to Stand Trial

¶ 50 For clarity’s sake, I remind the reader that this portion of my opinion does not represent the view of the majority of the court. For the majority’s conclusion on whether administration of antipsychotic medication is substantially likely to render Ms. Barzee competent to stand trial, please refer to Justice Durrant’s separate opinion.

¶ 51 In order for the State to medicate Ms. Barzee, it is required to prove that the medication is substantially likely to make her competent based upon clear and convincing evidence. Utah Code Ann. § 77-15-6.5(6)(a) (Supp. 2006); see also United States v. Bradley, 417 F.3d 1107, 1113-14 (10th Cir. 2005). Clear and convincing evidence requires that the evidence “place in the ultimate factfinder an abiding conviction that the truth of its factual contentions are highly probable. This would be true ... only if the material it offered instantly tilted the evidentiary scales in the affirmative when weighed against the evidence ... offered in opposition.” Colorado v. New Mexico, 467 U.S. 310, 316, 104 S.Ct. 2433, 81 L.Ed.2d 247 (1984) (internal quotation marks and citations omitted). I do not believe the State’s evidence in this case rises to this standard. Further, I believe that the district court did not properly weigh all of the evidence presented within the framework of this high burden. Instead, the court ignored the testimony of the defense experts and relied exclusively on the testimony of Drs. Jeppson and Whitehead. It did not consider the evidence presented by the defense, but found the State witnesses to be in the “best position” to determine the likelihood of Ms. Barzee’s restoration to competency and afforded complete deference to their opinions in abrogation of the court’s duty to actually consider and weigh all the evidenee presented.14 Proper balancing does not simply require a court to pick one expert or one side to defer to, but instead requires that the evidence presented by each side is thoughtfully considered and weighed. See United States v. Valenzuela-Puentes, 479 F.3d 1220, 1228-29 (10th Cir. 2007) (refusing to affirm the district court in a Sell hearing because competing testimony rebutted its conclusion that the defendant was substantially likely to be rendered competent with medication). The district court deferred to the opinions of the State witnesses based upon the purported familiarity that Drs. Jeppson and Whitehead had with Ms. Bar-zee. This reasoning is not supported by the evidence. First, Dr. Whitehead had only a single one-and-a-half-hour meeting with Ms. Barzee. The fact that he works in the same hospital where Ms. Barzee is committed does not establish that he is particularly familiar with her case. Second, while Dr. Jeppson is the psychiatrist assigned to Ms. Barzee’s case, the record does not show that he in fact spent significantly more time with her than Dr. Amador. The district court stated that he had “weekly” meetings with Ms. Barzee; however, Dr. Jeppson testified that his meetings with her were on a “weekly to monthly” basis over a twenty-two-month period and that “a lot of times these aren’t 45-minute talks, I mean just a few minutes here and there.” In addition, Ms. Barzee had not met with her treatment team for the ten months preceding the Medication Hearing. Dr. Amador visited with Ms. Barzee multiple times over the past few years, spending over fifteen total hours with her. Thus, the amount of time Dr. Amador spent with Ms. Barzee rivals the time she spent with Dr. Jeppson. My brief review of the record reveals that Drs. Jeppson and Whitehead clearly did not have a “quantity or quality of information” so superior to that of Drs. Morris and Amador such that their opinions should not even have been considered on the *63issue of Ms. Barzee’s likelihood of restoration. However, that is what the district court did. I reject the notion that a treating physician’s opinion can be looked to for the sole source of information when competing testimony, based on proper foundation, challenges the treating physician’s conclusions. In this ease, we should closely examine all of the witnesses’ testimony and look to the foundation for that testimony in weighing its impact. In examining that testimony, I conclude that the defense experts, not the state hospital physicians, most carefully considered and relied on Ms. Barzee’s particular condition and history. I am persuaded that the State witnesses lacked adequate foundation for their opinions in many respects, including a rebanee on general statistics rather than statistics particular to the defendant, failure to rely on the DSM-IV, failure to recognize the effect of a particular diagnosis on restoration to competency, failure to consider the presentation of Ms. Barzee’s particular symptoms, and the inability of medication to alter Ms. Barzee’s fixed delusions. I will address each of these issues below.

a. The State Experts’ Reliance on General Statistics Failed to Account for Ms. Barzee’s Particular Condition

¶52 Drs. Jeppson and Whitehead, the State experts, relied on their clinical experience at the state hospital in assessing the efficacy of medication on the restoration of Ms. Barzee’s competence. Dr. Jeppson cited seventy- to eighty-percent restoration rates at the state hospital and concluded that there would be a seventy-five-percent chance of restoration “with anybody.” He stated that “hopefully, [Ms. Barzee] would be restored to competency.” Dr. Whitehead noted that the state hospital has a restoration rate of eighty percent. He stated that Ms. Barzee’s case was nowhere “close to a one hundred percent case” for restoration, offering a gross estimate of seventy percent. He stated that a seventy-five-percent restoration rate exists for all psychotic disorders and opined that psychotic disorder not otherwise specified (PDNOS) may possibly have a higher rate. Dr. Amador’s testimony, by contrast, cautioned against looking only to personal clinical experience in drawing conclusions about the efficacy of treatment because the types of patients in any particular treatment population may skew a practitioner’s views on treatment generally.

¶ 53 I am not persuaded that the rates of restoration for the general population at the state hospital would have any bearing on Ms. Barzee’s particular case without any showing that the population resembled Ms. Barzee. As the Court of Appeals for the Fourth Circuit has recognized, the general population tells us nothing of the response of a particular patient. United States v. Evans, 404 F.3d 227, 240-41 (4th Cir. 2005) (refusing to rely on the state’s evidence when there was no indication that the federal hospital staff had considered the defendant’s particular medical condition in reaching conclusions). While Dr. Whitehead did opine that PDNOS patients may have higher rates of restoration than those with other psychotic disorders, he also stated that a particular diagnosis has “little, if any, ramifications” for treatment and that diagnosis does not have “much bearing” on restoration.15 The district court looked at the success rate of the general population in the state hospital and compared it with the statistical data presented by defense experts, who identified a much lower rate of restoration for patients with symptoms similar to Ms. Barzee’s — twenty percent for patients with delusional disorder, increasing to thirty or forty percent if other conditions, including schizophrenia, were considered. The court attributed the discrepancy, however, not to the individual symptoms and diagnoses of those similar patients, but to what the district court considered to be the “expertise” of the state hospital physicians. I find no basis for the court’s conclusion that Utah’s state hospital physicians are remarkably better than other psychiatrists at medicating and restoring incompetent patients. The fact that the rates for the general hospital population at the state hospital are identical to the rates reported by the *64Federal Bureau of Prisons’ hospital system clearly indicates no special expertise on the part of Drs. Jeppson and Whitehead. Rather, the discrepancy between the conclusions of the State witnesses and those of the defense witnesses is explained by the fact that the State’s witnesses based their opinions on the rates for general populations and saw little need to look at the particular sympto-mology of the defendant, while the defense experts refined their opinions to consider Ms. Barzee’s particular characteristics. Thus, in my opinion, to the extent that the testimony of the State witnesses relied on the restoration rates for the general population at the state hospital, it should be given little, if any, weight.

¶ 54 Likewise, I reject the State witnesses’ reliance on similar general statistics from the federal hospital system, but I am even more troubled by the State witnesses’ use of those particular statistics. Citing a report issued by the Federal Bureau of Prisons,16 Drs. Jeppson and Whitehead testified that the restoration rate for the general population at federal hospitals was seventy to eighty percent. As noted above, I have serious concerns with relying on statistics of the general population of a hospital. I also have concerns with the basic reliability of the federal statistics. Dr. Amador explained that the federal study was not peer reviewed; it was an unpublished, internal hospital report, and the patients were not broken down by diagnoses or symptoms. In fact, the physician who conducted the study did not even know the diagnoses of the patients involved in the study, nor did he know if any of the patients suffered from delusions similar to those suffered by Ms. Barzee. Those facts led Dr. Amador to conclude that the federal hospital data was “completely useless” in predicting whether medication would be successful in restoring Ms. Barzee to competency. Dr. Morris also testified that Ms. Barzee’s poor prognostic factors must be taken into account because each case is individual and statistics for general populations are not predictive of Ms. Barzee’s response. The rates identified for general populations were inconsistent with the rates found in studies cited by the defense experts, which were more focused and examine restoration of patients with symptoms mirroring those of Ms. Barzee. Thus, in my opinion, the statistics relied on by the State witnesses are entitled to little weight because the statistics do not account for the individual history and symptomology of Ms. Barzee. See United States v. Cruz-Martinez, 436 F.Supp.2d 1157, 1162 (S.D.Cal. 2006) (noting that the court had “serious doubts about the predictive value and applicability of the government’s statistic regarding the likelihood of success” when “[i]t [was] not even clear that the statistic applie[d] to individuals in defendant’s condition”).

¶ 55 Today, the majority permits forcible medication of patients at the state hospital based primarily on the statistic that seventy to eighty percent of the general population at the state and federal hospitals were restored without regard to individual diagnosis and prognosis. Thus, every patient committed to the state hospital is substantially likely to be restored. Allowing this analytical charade renders the second part of the Sell test meaningless; such flawed logic does not, in my view, comport with common sense. The decision today allows courts to order forced administration of antipsychotic medication without regard to the individual symptoms and history of a particular patient. It allows courts to do so based on general statistics even when a multitude of evidence suggests that a particular patient with unique characteristics is not likely to be restored. The majority is not the first court to rely on general statistics, but I believe it is the first court to do so when the defense has presented evidence controverting the reliability of *65those statistics. See United States v. Dallas, 461 F.Supp.2d 1093, 1095, 1099-1100 (D.Neb. 2006) (refusing to rely on general statistics cited by government witnesses who did not take into account the symptoms and history of the particular defendant); United States v. Cruz-Martinez, 436 F.Supp.2d at 1161-62 (same); see also United States v. Gomes, 387 F.3d at 159, 161-62 (upholding the district court’s conclusion that defendant was substantially likely to be restored based on the unchallenged testimony of government doctors who cited the Bureau of Prisons’ seventy percent success rate); United States v. Milliken, 2006 WL 2945957, at *9-10, 2006 U.S. Dist. LEXIS 82413, at *29-31 (M.D.Fla. 2006) (relying on unchallenged testimony citing the Bureau of Prisons’ seventy-six-percent success rate in restoring individuals to competence); United States v. Leveck-Amirmokri, 2005 WL 1009791, at *4, 2005 U.S. Dist. LEXIS 7610, at *14 (W.D.Tex. 2005) (relying on physician’s conclusion that medication fails to work only one time out of twenty when defendant had provided “no reason to doubt” the government witnesses). I am troubled by the district court’s decision in light of the evidence, but I am even more troubled that my colleagues are willing to undertake only highly deferential review of that decision — one that I believe should be recognized as clearly erroneous even under a deferential standard of review. In my opinion, when the State relies on statistics from the general population and competing testimony establishes that those statistics are inapplicable to a defendant with a particular history and particular symptoms, the general statistics are inapposite. When faced with competing evidence, general statistics cannot rise to the level of clear and convincing evidence that a patient is substantially likely to be restored. I believe this is the only reasonable conclusion that can be reached if the Sell test is to have continued validity in protecting the constitutional liberty interest in freedom from unwanted antipsychotic medication.

b. The State Experts’ Failed to Rely on the DSM-IV

¶ 56 The DSM-IV is the authoritative tool for diagnosis in the field of mental health; however, Drs. Jeppson and Whitehead stated that the manual was of little or no use to their analysis of Ms. Barzee’s ease. Dr. Jeppson stated, “I am not tied to DSM-IV.” When asked about the characteristics of PDNOS, he responded that he had not “reviewed that recently. I don’t pack [the DSM-IV] around.” When Dr. Whitehead was asked about specific diagnoses and their implications for treatment decisions, he stated that diagnosis has “little, if any, ramifications” for treatment, despite extensive research cited by the defense experts suggesting that diagnosis and symptoms have significant ramifications for a patient’s response to medication. Indeed, Dr. Ama-dor testified that the difference in diagnosis at “this particular hearing” is relevant to “predicting response” to antipsychotic medication. The state hospital physicians’ dismissal of the standard diagnostic system used in the field of psychiatry is perplexing in light of the important questions they were asked to address by the State in this case. While diagnosis may not be critical to the question of whether medication is medically appropriate — because for any form of psychosis, drugs will almost undoubtedly be appropriate — diagnosis is of 'utmost importance for predicting the effect of medication on a particular patient. In this case, the court was not asked to consider the likelihood of any psychotic patient being restored to competency through medication; instead, the court was to decide whether forcible medication was likely to restore Ms. Barzee to competence. I am at a loss to comprehend the State witnesses’ disregard of the DSM-IV, an important and integral tool for diagnosis in the mental health arena.

¶ 57 Not only did the State witnesses disregard a basic tool of their trade, but Dr. Jeppson, without the aid of the DSM-IV, made a puzzling change in Ms. Barzee’s diagnosis from delusional disorder to PDNOS. Three of the four witnesses at the Medication Hearing agreed that delusional disorder was a reasonable diagnosis for Ms. Bar-zee. Dr. Jeppson knew that Ms. Barzee had experienced referential thinking prior to his *66initial diagnosis of delusional disorder,17 yet he pointed to no other factor explaining his subsequent change in diagnosis except her continued reports of referential thinking.18 Although Dr. Jeppson testified that referential thinking was “one of many changes” influencing his decision, he failed to cite any other factor, and the factor he did identify was not even a “change.” He stated that he did not believe that referential thinking was part of the symptomology of delusional disorder pursuant to the DSM-IV — which he had not read recently — but “if it is, it is certainly a small part.” While Dr. Jeppson eventually softened his assertion that referential thinking removes one from the diagnosis of delusional disorder, this assertion was flatly rebutted by other witnesses. Dr. Morris testified that delusional ideas of reference are simply one delusion of the delusional disorder and that the symptom does not remove one from a diagnosis of delusional disorder according to the DSM-W. Dr. Amador, the co-chair for revising the DSM-IV section on psychotic disorders, stated that referential thinking as exhibited by Ms. Barzee — receiving messages from movies— did not remove her from the diagnosis of delusional disorder in the DSM-IV. Not surprisingly, the DSM-IV section on delusional disorder states, “Ideas of reference (e.g., that random events are of special significance) are common in individuals with [delusional] disorder. Their interpretation of these events is usually consistent with the content of their delusional beliefs.” DSM-IV 325-26. Thus, I am skeptical of Dr. Jeppson’s change in diagnosis based on factors that apparently do not withstand the scrutiny of other mental health professionals and published professional standards. I am further troubled because the DSM-IV, the standard tool for diagnosing mental illness, appears to have been given little, if any, weight by Drs. Jeppson and Whitehead. Furthermore, in view of the DSM-IV’s flat rejection of the notion that ideas of reference are not a symptom of delusional disorder, I find Dr. Jeppson’s testimony unpersuasive.19

¶ 58 Equally perplexing is the change after two years of treatment from the more specific diagnosis of delusional disorder to the more general one of PDNOS.20 Dr. Morris pointed out that there is no evidence that Ms. Barzee’s condition is the product of substance abuse or a medical condition, and Dr. Whitehead noted that the chance that a medical condition was causing her symptoms was “extremely unlikely.” Thus, the remaining diagnoses beneath the umbrella definition of PDNOS, which are available to this particular patient, are schizophrenia and delusional disorder.21 Only one initial evaluator, Dr. Cohn, suggested that schizophrenia was the appropriate diagnosis; all the other practitioners doubted the existence of symptoms as severe as Dr. Cohn’s report suggested. Neither of the State witnesses at the Medication Hearing suggested that Ms. Barzee met the *67criteria for schizophrenia or even opined that it might have been an appropriate diagnosis. Thus, while the boundaries between psychotic disorders may at times be “fuzzy,” according to Dr. Whitehead, the opinions of all the mental health professionals involved in this case persuade me that it is extremely likely that Ms. Barzee suffers from delusional disorder.

c. The Diagnosis of Delusional Disorder Suggests a Poor Prognosis for Response to Medication

¶59 If in fact Ms. Barzee suffers from delusional disorder, the prognosis for restoration is poor. Dr. Nielsen, an initial evaluator, opined that if Ms. Barzee suffered from delusional disorder, her condition was refractory and rarely treatable with medication. Dr. Jeppson explained that he had treated very few patients with delusional disorder. He admitted that delusional disorder is more refractory to treatment than schizophrenia, but was unsure if it is the most refractory of psychotic disorders. Dr. Amador considered the specific diagnosis of delusional disorder when he opined that Ms. Barzee had a twenty-percent chance of restoration, noting that delusional disorder is harder to treat than other psychotic disorders. He based this opinion on research and his clinical experience. Dr. Whitehead discounted the literature and research upon which Dr. Amador relied, calling it “traditional clinical lore” and claiming that it “would be an error to say that a delusional disorder is refractory to medication” because “the jury is out on that.” He stated that very little research has been conducted on delusional disordered patients because the diagnosis is “so rare.” He admitted to seeing only four or five cases of delusional disorder over the course of his career. Dr. Whitehead cited one study that suggested an eighty percent response rate to medications for people with delusional disorder. However, it is unclear how the presence of a response to medication equates to restoration to competency, especially when the experts agree that Ms. Barzee’s delusions would persist even if medication was administered. Moreover, Dr. Whitehead relied on the information reported by the federal hospital study; he testified that eighty percent of the delusional patients in that study were restored and that five to ten delusional patients were involved in the study. That information, however, was unequivocally rebutted by Dr. Amador, who explained to the court that he spoke directly with the physician who conducted the study and that that physician did not know the diagnoses of the patients involved. Thus, Dr. Whitehead’s testimony about the presence of delusional patients in the federal study appears to be without foundation. Despite Dr. Whitehead’s opposition to looking to “dogmatic conclusions based on research on delusional disorders” because it is rare, I am persuaded by Dr. Amador’s review of the relevant scientific literature and his conclusions based on numerous studies that find delusional disorder refractory to medication even though “some therapeutic effect” may occur with medication for any psychotic disorder.22 The State had the burden of proving by clear and convincing evidence that Ms. Barzee was substantially likely to become competent through the administration of antipsychotic medication. In my opinion, the high threshold of this burden has not been met. If delusional disorder is the appropriate diagnosis, the evidence presented fails to establish that medication is substantially likely to render Ms. Barzee competent.

d. The Presentation of Ms. Barzee’s Particular Symptoms and Her History Suggest a Poor Prognosis for Response to Medication

1Í60 If Ms. Barzee suffers from PDNOS or another variety of psychotic disorder, her specific symptoms also lead to the conclusion that restoration to competency is unlikely. The Court of Appeals for the Fourth Circuit recognized that in order to determine whether a patient is substantially likely to be restored to competence, a mental *68health professional must consider the “particular mental and physical condition” of the patient. Evans, 404 F.3d at 240-41; see also United States v. Baldovinos, 434 F.3d 233, 240 n. 5 (4th Cir. 2006) (recognizing that Sell requires “an exacting focus on the personal characteristics of the individual defendant”); United States v. Valenzuela-Puentes, 479 F.3d 1220, 1229 (10th Cir. 2007) (remanding the case to the trial court with instructions to “specifically consider” the particular symptoms and characteristics of the defendant). Thus, in analyzing the opinions of the experts in this ease, I believe we must look closely at the factors each expert considered in reaching his conclusion with regard to Ms. Barzee’s likely response to antipsychotic medication. In doing so, it becomes clear that the defense experts were the only witnesses in this case who gave due consideration to Ms. Barzee’s history and symptoms. Although Dr. Jeppson referred to Ms. Barzee’s gender as a positive factor for restoration (without citing authority for his assertion), he and Dr. Whitehead essentially dismissed all of the specific facts apparent in Ms. Barzee’s illness.23 The most notable of the features particular to Ms. Barzee is the duration of her untreated psychosis, a period in excess of ten years and possibly as much as thirteen years. She also exhibits grandiose delusions and some persecutory delusions as well as delusional ideas of reference or referential thinking. Further, Ms. Barzee does not believe that she is mentally ill. In addition, no positive prognostic factors, such as family history or past successful treatment with medication, exist in Ms. Barzee’s case. See United States v. Archuleta, 218 F. App’x 754, 756 (10th Cir. 2007) (relying on previous restoration with antipsychotic drugs and “well-documented history of resolution of his psychotic symptoms” as a result of medication in concluding that the defendant was substantially likely to be restored (internal quotation marks omitted)); United States v. Morris, 2005 WL 348306, at *3-4, 2005 U.S. Dist. LEXIS 38785, at *13-14, 20 (D.Del. 2005) (holding that defendant was substantially likely to be rendered competent based on past history of positive response to anti-psychotic medication).

¶ 61 First, Drs. Morris and Amador opined that a duration of untreated psychosis in excess of one year significantly decreased the likelihood that a patient will respond to medication.24 In Ms. Barzee’s case, her psychosis has impacted her functioning and behavior for at least ten years and possibly as much as thirteen years. Dr. Morris stated that duration of untreated psychosis as a predictor of response to medication was not a novel idea in psychiatry. The longer a patient is ill— even more than six months to a year — the chances of improvement with medication significantly decrease over time. Dr. Morris cited his experience and the current literature, referring to numerous studies, for the “very well established” proposition that duration of untreated psychosis is an “important factor” in predicting response to medication. Dr. Amador similarly identified Ms. Barzee’s lengthy duration of untreated psychosis as a poor prognostic factor. Dr. Amador based this conclusion on his experience and on the research reported in the relevant scientific literature; in fact, he had personally participated in peer review of approximately fifteen articles on this specific subject. He stated that after one year of psychosis without treatment, there is typically no response for the negative or positive symptoms of a psychotic disorder,25 nor is there a response for social functioning. Dr. Amador opined that Ms. Barzee’s lengthy duration of untreated psychosis was a “significant factor ... [that] reduce[d] her chances of responding to anti-*69psychotic medication” and also reduced the chances that her delusions will cease. Contrary to the testimony of the defense experts, Dr. Whitehead stated that the duration of untreated psychosis was an “important” variable, but that it should not be “overstated.” He later said it was only a “small to moderate” factor in predicting response to treatment. He did not address the numerous studies cited by the defense, but pointed to studies from the 1950s and one study from 2005 where the duration of untreated psychosis impacted treatment of only negative symptoms — not Ms. Barzee’s most prominent positive symptom of grandiose delusions. Similarly, Dr. Jeppson, while admitting that duration of untreated psychosis was “definitely a factor” in predicting response, did not think that the duration of Ms. Barzee’s disorder had implications for her response to medication. He did admit, however, that patients experiencing their first episode of psychosis would “probably respond better,” but that he would “have to read” on whether duration of untreated psychosis impacts the prognosis for response to medication. He opined that just because a newly psychotic patient may respond more favorably to medication, that does not mean that “someone who has [suffered from psychosis] for ten or eleven years is not going to respond.” Dr. Whitehead made a similar statement in his testimony, suggesting that illness untreated for one day may not respond to medication while illness untreated for decades may respond. He also stated that the state hospital had had success in treating patients with previously untreated illness in excess of one year. However, he did not state that the hospital had ever successfully treated anyone with untreated psychosis to the extent of Ms. Barzee’s, and I find no support in the record for the district court’s conclusion that the state hospital physicians somehow have expertise in restoring delusional patients with a “significant delay from the onset of psychosis to initial treatment.” While I recognize that some patients with a lengthy duration of untreated psychosis may in fact respond to medication,26 the overwhelming evidence from the empirical research presented at the Medication Hearing identified duration of untreated psychosis as a factor negatively impacting the response to administration of an-tipsychotic medication. The speculation that an individual patient with years of untreated psychosis may respond favorably does not meet the clear and convincing standard of the State’s burden to prove that this particular patient is substantially likely to be restored to competency by forced medication.27 The evidence presented at the hearing establishes that it is not substantially likely that someone with ten to thirteen years of untreated psychosis will respond to medication at all, much less be rendered competent.

¶ 62 Second, the presence of grandiose delusions, Ms. Barzee’s “primary competency impairing symptom,” was also identified as a poor prognostic factor for psychotic patients’ response to medication. In his initial review, Dr. Nielsen pointed out that symptoms of a delusional nature do not respond favorably to medication. Dr. Morris stated that the prominent delusion affecting Ms. Bar-zee’s competency is her “leave it to the Lord” grandiose delusion. He stated that grandiose delusions are more difficult to treat than other delusions, such as persecu-tory delusions, which Ms. Barzee also exhibited to a lesser extent. Dr. Amador explained that grandiose delusions permeate an individual’s personality and self-esteem and impede treatment of nonmood-related *70psychosis generally when compared to other types of delusions. His conclusion that grandiose delusions are refractory to treatment was informed by his clinical experience and by the relevant scientific research. Dr. Whitehead opined differently; he stated that grandiose delusions were not particularly refractory and cautioned not to look to the “dogmatic conclusions” of research with delusional disordered patients. He stated that Ms. Barzee exhibited positive symptoms,28 and he then relied on research suggesting an eighty-percent response rate to medication for psychotic patients generally who suffer from positive symptoms. However, Dr. Whitehead did not identify whether this research referred to hallucinations or delusions or if it focused on grandiose delusions specifically. Similarly, Dr. Jeppson (without citing any authority) opined that positive symptoms and lack of negative symptoms suggested a better prognosis. As noted above, the research highlighted by defense experts concluded that the impact of medication was less favorable for grandiose delusions than for other types of delusions, hallucinations, and disorganized thinking. Dr. Jeppson again attempted to rebut the testimony of defense witnesses by focusing on the idea that an individual patient with grandiose delusions may respond to medication, stating that he had seen people who think they are Jesus respond to medication. Again, I am unpersuaded by a single, particular example that would appear to contradict the findings of a substantial amount of research. I recognize, and all the testimony supports, the fact that some individuals who suffer from grandiose delusions will respond to medication, but this does not inform the court as to Ms. Barzee. Even if only one percent of patients responded to medication, examples would exist, but the court should not extrapolate from the experience of one patient, or a handful of patients, the notion that Ms. Barzee will respond to medication.

¶ 63 Third, Drs. Amador and Jeppson identified Ms. Barzee’s lack of insight into her illness as a significant symptom of her psychosis. Dr. Amador has specific experience with psychotic patients who do not believe they are mentally ill. He has conducted research on treatment response with these particular individuals, and he has coauthored approximately fifty peer reviewed articles on the topic of involuntary medication of individuals -who do not recognize that they are mentally ill. Based on his experience and research, he identified this symptom as a factor that would have a negative impact on Ms. Barzee’s response to medication, stating that it “will take many years [for a patient] to ever understand there is an illness.”

¶ 64 Taking into account the poor prognostic factors and the absence of positive factors, Dr. Morris concluded that medication is unlikely to resolve Ms. Barzee’s delusions to the point of restoration of competency. He stated, “I am of the opinion that it is unlikely that she is going to make any kind of significant response to medication. Based on that, it is even more unlikely that she is going to be restored to competency.” Dr. Amador’s conclusion was identical. Their reasoning is persuasive because of the specificity with which they analyzed Ms. Barzee’s condition and because of the foundations upon which they based their opinions — not statistics derived from a general psychiatric population, but rather from research tied to the specific characteristics exhibited by Ms. Barzee. Accordingly, even setting aside the witnesses’ disputes about diagnosis, I am still persuaded that medication is unlikely to restore Ms. Barzee to competence.

e. Even With Medication, Ms. Barzee Will Remain Incompetent to Stand Trial Because Medication Is Not Likely to Eliminate Her Fixed Delusions

¶ 65 It is important to understand why Ms. Barzee is incompetent to stand trial in order to determine whether restoration is substantially likely. Prior to the Medication Hearing, Drs. Kovnick, Cohn, Berge, and Nielsen evaluated Ms. Barzee for the sole purpose of determining whether she was competent. Those professionals concluded that Ms. Bar-zee’s primary area of incompetency was her inability to engage in reasoned choice of legal *71strategies and options. Drs. Kovnick, Cohn, Berge, and Nielsen all agreed that Ms. Bar-zee showed severe impairment in this capacity because her religious delusions, and not her best interests, drove her decision-making process. Ms. Barzee told Dr. Kovnick that to “fight for her life during this court process would be not allowing God to put her through the suffering that he feels is required of her.” All of the professionals agree that Ms. Bar-zee’s delusions currently make her incompetent to stand trial, and the district court concluded that she was incompetent on three separate occasions prior to the Medication Hearing.29 Within the framework of Ms. Barzee’s specific impairment, I cannot conclude that medication is likely to render her competent to stand trial.

¶ 66 Because they did not take into account the changes that must occur for Ms. Barzee to be rendered competent, the State witnesses’ conclusions that medication will actually restore her to competency do not persuade me. The State focused on, and the district court was persuaded by, the witnesses’ conclusion that, with the aid of medication, Ms. Barzee would be less preoccupied with her delusions and talk about them less. That conclusion ignores the actual cause of Ms. Barzee’s incompetency: in order for Ms. Barzee to be rendered competent, her delusions must be eliminated. As long as she has the delusions, regardless of whether they dominate her conversation, they will impact her ability to engage in the reasoned choice of legal strategies and options. If Ms. Bar-zee talks less often about God directing her life, but continues to experience the delusion that she is subject to God’s direction and that her participation in the legal proceedings would go against God, she will lack the capacity to engage in reasoned choice of legal strategies and options. This will be true even if it is masked by a medication-induced ability to talk less about the delusions. The delusions will nonetheless continue to drive her decision-making process.

¶ 67 My conclusion is derived directly from the generally uncontroverted testimony presented at the Medication Hearing. Dr. Morris stated that Ms. Barzee’s delusions influence her competency because they inhibit her ability to weigh legal strategies. He stressed that the “grandiose delusions become predominant” when it comes to Ms. Barzee’s strategy in legal proceedings. Dr. Jeppson stated that Ms. Barzee will continue to have delusions even if medicated, but that with medication, she will focus on them or talk about them less, and the medication “[will] hopefully diminish the strength or maybe the intensity of the delusion.” Dr. Morris agreed with Dr. Jeppson’s assessment, testifying that there is a “poor prognosis for ending her delusion” with medication and that she “might” talk less about the delusion and not focus on it as much. Dr. Amador also agreed with Dr. Jeppson on this point, stating that while medication may make Ms. Barzee feel less pressure to discuss her delusions, there is a “high likelihood that the fundamental delusional beliefs aren’t going to change”; she may talk about her case, but that will not change the fact that she believes God is directing everything, and thus, the delusions will continue to influence her decisions and behavior. According to Dr. Amador, in order to render Ms. Barzee competent, medication would have to reduce the certainty of Ms. Barzee’s beliefs, not merely her discussion of those beliefs, and it was Dr. Amador’s opinion that medication would not impact Ms. Barzee’s certainty.30 Even Dr. Cohn, who believed Ms. Barzee suffered from schizophrenia, which generally responds more favorably to medication than delusional disorder, opined that it was unclear whether Ms. Barzee’s “deeply entrenched, delusional belief’ would be impacted by medication because those symptoms *72“are most refractory to pharmacological intervention.”

¶ 68 At the Medication Hearing, the district court judge stated that she was not concerned with eliminating Ms. Barzee’s delusions; rather, she was concerned with restoring Ms. Barzee to competency. I conclude that, in Ms. Barzee’s case, the continued presence of delusions would constitute continued incompetence because her inability to make reasoned choices about her legal predicament, for example, her incompetence, is driven by fixed delusions that, even Dr. Jeppson admits, will continue despite the administration of antipsy-chotic medication.

f. Medication Is Not Substantially Likely to Render Ms. Barzee Competent to Stand Trial

¶ 69 The testimony of the defense experts, which is founded on empirical research documented in the relevant scientific literature, peer-reviewed studies, DSM-IV diagnoses, and clinical experience, is highly persuasive. Unlike the physicians from the state hospital, Drs. Amador and Morris focused on Ms. Barzee’s particular symptoms when rendering their opinions. I conclude that the district court’s deference to the State witnesses was not justified by the evidence. The State witnesses’ clinical experience and general statistical data do not outweigh the focused analyses of the defense witnesses. The district court was in error in its statement that “the statistical data ... is neither as authoritative or weighty as the testimony of [the] actual treating physician.” The statistical data that was particular to Ms. Barzee’s symptoms and history was the most authoritative and weighty information that was provided to the district court. Anecdotal evidence based on limited clinical experience cannot be used to extrapolate information about Ms. Barzee. Nor can statistics derived from psychotic patients generally be applied to Ms. Barzee without a showing of similarities in disorder, symptoms, and history. When numerous studies rebut a particular practitioner’s limited and possibly inapposite experience, the information grounded in empirical research is the most authoritative and weighty evidence; it should not have been disregarded here.

¶ 70 I believe that Ms. Barzee is not substantially likely to be rendered competent to stand trial through the forced administration of antipsychotic medication. This conclusion is based on the weight of the evidence presented at trial, which persuades me that someone with Ms. Barzee’s particular symptoms and history is unlikely to respond to medication, much less be rendered competent. Thus, in my view, the State’s interest in bringing Ms. Barzee to. trial will not be significantly furthered by involuntarily medicating her.

¶ 71 In conclusion, I do not doubt that medication is the appropriate medical treatment for Ms. Barzee. As Dr. Whitehead pointed out, antipsychotic medication is the “cornerstone” of treatment for all psychotic disorders and “has been for the past fifty years.” I empathize with Dr. Morris’ statement that the inability to treat psychotic patients is “very frustrating” when treatment providers know there is “no way” patients are “going to get better” without medication. I also recognize the concerns expressed by Dr. Jeppson, that Ms. Barzee is unlikely to make any progress without medication and that the state hospital is simply “warehousing” her without treatment or care, which is not in her best medical interest.

¶ 72 However, this case involves the State’s ability to involuntarily medicate a criminal defendant for the sole purpose of rendering her competent to stand trial. This court is not charged with determining what is best for Ms. Barzee from a. mental health standpoint — if that were the case, there would be little need for the judicial branch to exercise any authority in these matters. This court’s duty is to determine whether Ms. Barzee’s constitutional liberty interest in freedom from unwanted medication can be overcome by the State’s interest in rendering her competent to stand trial. Unless the State’s interest is highly likely to be furthered by the intrusion upon Ms. Barzee’s liberty, this court cannot allow the State to forcibly medicate her with antipsychotic drugs.

¶ 73 I agree with Dr. Whitehead’s testimony that “predicting the treatment effect is often difficult.” I feel that when this court is faced with such a difficult task that impli*73cates the constitutional guarantee of liberty, we must carefully scrutinize the evidence to ensure that the State has met its high evi-dentiary burden. In this ease, I cannot conclude that the State proved by clear and convincing evidence that Ms. Barzee, given her specific disorder, history, and symptoms, is substantially likely to be rendered competent through the involuntary administration of antipsychotic medication. The district court overlooked a multitude of compelling evidence in favor of the State witnesses’ presumed expertise and familiarity. This purported expertise and familiarity does not hold up to a review of the evidence presented. Reliance upon the evidence of the State witnesses in this case — general statistics without regard to the individual characteristics of a defendant’s disorder — undermines the constitutional right to freedom from unwanted antipsychotic medication, a significant right that the Supreme Court has recognized on numerous occasions.

¶ 74 I cannot conclude that the State may constitutionally involuntarily medicate Ms. Barzee solely in an attempt to restore her to competence for trial when it has failed to prove that medication is substantially likely to accomplish that end. Therefore, I conclude that the State has failed to meet its burden in overcoming Ms. Barzee’s liberty interest in freedom from unwanted antipsy-chotic medication. I would reverse the order of the district court.

CONCLUSION

¶ 75 Notwithstanding my analysis with respect to the second Sell factor and my conclusion that it is impermissible for the State to intrude upon Ms. Barzee’s federal constitutional liberty interest in freedom from unwanted medication, the majority concludes otherwise. As set forth in Justice Durrant’s separate opinion, the majority affirms the district court’s order providing that the State may medicate Ms. Barzee.

¶ 76 Justice NEHRING concurs in the entire result of Chief Justice DURHAM’S opinion. Associate Chief Justice WILKINS, Justice DURRANT and Justice PARRISH concur in all parts of this opinion, save Part III.A and Part III.D.2.

writing for the majority:

¶ 77 We concur in Chief Justice Durham’s opinion except her analysis with respect to the second Sell factor. As to this factor, we agree with Chief Justice Durham’s determination that “the district court did not clearly err in concluding that antipsychotic medication is substantially unlikely to interfere with Ms. Barzee’s right to a fair trial,”1 but we respectfully disagree with the remainder of the analysis.

¶ 78 We disagree with Chief Justice Durham’s analysis, set forth in Part III.A, as to the appropriate standard of review for the second Sell factor. Chief Justice Durham would apply a de novo review to the question of whether the administration of medication to Ms. Barzee is substantially likely to render her competent to stand trial — the first inquiry under the second Sell factor. Because this inquiry is essentially a factual one, we apply a clear error standard of review.

¶79 We also disagree with Chief Justice Durham’s substantive analysis, set forth in Part III.D.2, as to the first inquiry under the second Sell factor. Chief Justice Durham concludes that the State’s experts wholly disregarded the particular characteristics of Ms. Barzee’s psychosis and relied exclusively on reports and statistics regarding general populations. Our review of the record indicates that the State’s expert witnesses did, in fact, consider the particular characteristics of Ms. Barzee’s condition and based their conclusions, in part, on their assessment of those characteristics.

I. THE DISTRICT COURT’S FINDING THAT THE ADMINISTRATION OF ANTIPSYCHOTIC MEDICATION TO MS. BARZEE IS SUBSTANTIALLY LIKELY TO RENDER HER COMPETENT TO STAND TRIAL IS PRIMARILY A FACTUAL FINDING THAT SHOULD BE REVIEWED FOR CLEAR ERROR

¶ 80 As to the standard of review question, while the United States Supreme *74Court in Sell v. United States2 set forth four criteria to be considered in determining whether a defendant may be forcibly medicated to restore competency for the purpose of standing trial,3 it did not speak to the standard an appellate court should apply in reviewing a lower court’s decision as to these criteria. A number of federal circuit courts have, however, considered the question of the appropriate standard of review to be applied to a lower court’s determination regarding competency under the first part of the second Sell factor. The Second and Fourth Circuit Courts of Appeals have concluded that the second Sell factor addresses factual questions that should be reviewed for clear error.4 These courts have analyzed the second Sell factor in a way consistent with our traditional approach to review of lower court decisions. As to purely factual questions, we review the fact finder’s determinations for clear error.5 As to purely legal questions, we review the lower court’s determination for correctness (de novo).6 Only mixed questions of fact and law present a difficult challenge in determining the appropriate standard of review.7

¶ 81 Chief Justice Durham concludes that the question of competency under the second Sell factor is a mixed question and that because of the important constitutional right involved — the right to be free of forced medication — a de novo standard should be applied. But we conclude that this issue presents a predominantly factual question. Certainly, in a sense, even traditionally factual inquiries can be viewed as mixed questions. Issues such as reasonableness, knowledge, and proximate cause are fact questions to be resolved by the fact finder, but the definition of the terms themselves is a legal matter. For example, in a jury trial a judge will instruct a jury as to the legal definition of “reasonableness,” but the jury makes the factual determination as to whether a defendant was reasonable. So, too, in this case, the definition of “competency” requires a legal analysis, but whether a particular defendant is competent or is likely to be rendered so is a question of fact for the fact finder to answer.

¶ 82 In Utah, the Legislature has set forth various criteria to be considered by a court in determining whether a defendant is competent to stand trial;

(a) the defendant’s present capacity to:

(i) comprehend and appreciate the charges or allegations against him;

(ii) disclose to counsel pertinent facts, events, and states of mind;

(iii) comprehend and appreciate the range and nature of possible penalties, if applicable, that may be imposed in the proceedings against him;

(iv) engage in reasoned choice of legal strategies and options;

(v) understand the adversary nature of the proceedings against him;

(vi) manifest appropriate courtroom behavior; and

(vii) testify relevantly, if applicable;

(b) the impact of the mental disorder, or mental retardation, if any, on the nature and quality of the defendant’s relationship with counsel;

(c) if psychoactive medication is currently being administered:

(i) whether the medication is necessary to maintain the defendant’s competency; and

(ii) the effect of the medication, if any, on the defendant’s demeanor and affect *75and ability to participate in the proceedings.8

Each of these criteria presents the kind of question that we have traditionally considered factual in nature.9 Essentially, in assessing competency, a judge is required to consider whether the defendant understands her circumstances well enough to understand the crime with which she is charged, to assist her counsel in defending against those charges, and to understand the potential punishments.10 Thus, it is for the courts or the Legislature to define “competency,” but whether a particular defendant is competent as so defined is a factual determination that we review for clear error.

¶ 83 Although the question of competency involves predominantly empirical inquiries and requires a clear error standard of review on appeal, defining competency is, as we have noted, a legal matter. Even assuming that, therefore, the question of whether Ms. Barzee would likely be rendered competent by medication should be reviewed as a mixed question of fact and law, as Chief Justice Durham views it, our application of the three-part test under State v. Levin11 leads us to a different conclusion than that reached by Chief Justice Durham. Under the Levin test, to the extent the question before us is a mixed one, we would accord the trial court’s finding substantial deference.

¶ 84 As to the first Levin factor — “the degree of variety and complexity in the facts to which the legal rule is to be applied”12— we agree with Chief Justice Durham that the issue before us is complex. Indeed, this complexity is illustrated by the fact that Chief Justice Durham devotes a substantial number of pages to the issue of whether medicating Ms. Barzee is substantially likely to render her competent to stand trial. This factor, therefore, favors deferential review.13 We disagree with Chief Justice Durham, however, as to her assessment of the second and third Levin factors.

¶ 85 With respect to the second Levin factor — “the degree to which a trial court’s application of the legal rule relies on ‘facts’ observed by the trial judge, ‘such as a witness’s appearance and demeanor, relevant to the application of the law that cannot be adequately reflected in the record available to appellate courts’ ”14 — we disagree with Chief Justice Durham’s conclusion that “much of the necessary information concerning the diagnoses and research ‘generally can be adequately reflected in a cold record’ and will be supported by statistics and published research, unaffected by an individual’s appearance and demeanor while testifying.”15 Rather, we conclude that the district court’s assessment of the witnesses did involve some credibility determinations. The issue of whether medication is substantially likely to render Ms. Barzee competent to stand trial ultimately involves a disagreement among experts. “Because a trial court is in a better position to judge credibility and resolve evi-dentiary conflicts,” we review a trial court’s findings of fact for clear error.16

¶ 86 The district court carefully listened to evidence presented by competing experts for several days. Each of the experts was quali-*76ñed and took defensible positions as to Ms. Barzee’s potential competency. Thus, the district court ultimately had to make credibility determinations and resolve evidentiary conflicts. Although credibility assessment is not as critical here as it would be in many eases, we should give weight to the fact that the district court viewed the witnesses — and Ms. Barzee — first hand. It may well be possible for us to make an assessment as to the likelihood that Ms. Barzee will be rendered competent based on the record alone, but we are not as well positioned as the district court to make that assessment. Thus, our analysis under the second Levin factor also favors deferential review.17

¶ 87 Chief Justice Durham concludes that the third Levin factor — “other ‘policy reasons that weigh for or against granting discretion to trial courts’ ”18 — tips the balance in favor of de novo appellate review.19 Specifically, she explains that “constitutional liberty interests of the highest degree” are implicated when the state seeks to forcibly medicate a defendant for the purpose of rendering the defendant competent to stand trial.20 Although we have made a determination to grant less deference to factual findings in certain areas of the law that implicate constitutional interests, in our view we should do so rarely. For example, we apply nondeferential review “to the application of law to the underlying factual findings in search and seizure cases.”21 We do so because of our “interest in having uniform legal rules regarding consent to search, given the substantial Fourth Amendment interests lost as a result of such consents.”22 Furthermore, we have held that, although there may be varying fact patterns that are relevant to the permissibility of searches and seizures, they are “not so unmanageable in their variety as to outweigh the interest in having uniform legal rules.”23 For example, in reviewing the reasonableness of a traffic stop and protective search, we often need look only to a few objective facts, shared in many cases. As a result, we are well positioned to make the ultimate determination of reasonableness.24

¶ 88 But competency cases inherently involve case-by-case evaluations, and the facts presented by such cases are less manageable in their variety and less susceptible to categorization into recurring objective fact patterns than is the case in the Fourth Amendment context. Due to the different individual histories, diagnoses, drugs, treatments, and expert testimony that will likely be involved in every competency case, we are in no better position than the lower court to arrive at the ultimate determination of competency. Indeed, Chief Justice Durham’s opinion shows how complex and fact sensitive a competency evaluation can be with respect to a particular individual such as Ms. Barzee.

¶ 89 In the context of criminal trials, due process concerns are always present, so unless there is an overriding policy that would favor de novo review by an appellate court of a mixed question, as is the case in the Fourth Amendment context, we should defer to the fact finder. Such an overriding policy is not present here. The mere fact that the factual inquiry into whether Ms. Barzee is substantially likely to be rendered competent to stand trial if forcibly medicated implicates an important constitutional right does not warrant reviewing this inquiry with less deference than we review other factual issues. That is, the mere fact that this factual inquiry concerns a constitutional right *77does not alter the nature of that inquiry or the standard of review. Every criminal ease involves important constitutional rights, yet as to purely factual questions, we review for clear error. Moreover, the question of competency is, in other contexts, widely regarded as a factual issue.25 Thus, even viewed as a mixed question subject to the Levin test, the district court’s finding that it is substantially likely Ms. Barzee will be rendered competent by medication should be given substantial deference.

II. THE DISTRICT COURT’S FINDING THAT THE ADMINISTRATION OF ANTIPSYCHOTIC MEDICATION TO MS. BARZEE IS SUBSTANTIALLY LIKELY TO RENDER HER COMPETENT TO STAND TRIAL IS NOT CLEARLY ERRONEOUS

¶ 90 Under the second Sell factor, the district court found that the administration of antipsychotic medication is “substantially likely” to render Ms. Barzee competent to stand trial. Reviewing this finding de novo, Chief Justice Durham concludes that administration of antipsychotic medication is not “substantially likely” to render Ms. Bar-zee competent to stand trial. And Chief Justice Durham further concludes that, even under a deferential standard of review, the district court’s factual findings were clearly erroneous. We respectfully disagree.

¶ 91 Chief Justice Durham states that the district court “ignored the testimony of the defense experts and relied exclusively on the testimony of Drs. Jeppson and Whitehead,” thereby failing to properly weigh all of the evidence.26 Furthermore, she asserts that “the defense experts were the only witnesses in this case who gave due consideration to Ms. Barzee’s history and symptoms.”27 She states that, “[u]nlike the physicians from the state hospital, Drs. Amador and Morris focused on Ms. Barzee’s particular symptoms when rendering their opinions.”28 Chief Justice Durham further concludes that the State’s experts relied almost exclusively on federal and state reports from general hospital populations “without a showing of similarities in disorder, symptoms, and history” to Ms. Barzee.29 But our review of the record and the district court’s decision convinces us that the district court did not “ignore” the testimony of the defense experts and rely exclusively on the State’s experts. The district court’s opinion discusses the qualifications and conclusions of the key expert witnesses: Drs. Amador and Morris for the defense, and Drs. Jeppson and Whitehead for the State. The district court carefully noted the crucial differences in their testimony and discussed the relevant merit in each of their opinions.

¶ 92 Furthermore, we conclude that the State’s experts did not disregard Ms. Bar-zee’s particular ease, although it is true that they relied on federal and state statistics taken from general hospital populations. Dr. Whitehead was the clinical director of the forensic unit during much of the time Ms. Barzee was a patient, and he had frequent discussions with Dr. Jeppson concerning her diagnosis and symptoms. Dr. Jeppson was Ms. Barzee’s treating physician for over two years and testified that he met with her on a *78“weekly to monthly” basis. He was familiar with her particular psychosis and her particular symptoms, and his testimony reflected this experience. Moreover, Dr. Jeppson’s and Dr. Whitehead’s opinions were based on their own clinical experience, not only with patients suffering from psychosis generally, but with those patients whose symptoms were similar to those of Ms. Barzee. Indeed, Dr. Jeppson and Dr. Whitehead each had treated and restored to competency well over one hundred patients. And Dr. Whitehead testified that fifteen patients at the state hospital who had the same diagnosis as Ms. Barzee were all restored to competency through the administration of antipsychotic medication.30

¶ 93 The State’s experts did rely on reports regarding general populations in the federal hospital system and at the state hospital, in addition to other statistical information. These reports indicated that the success rate in restoring patients suffering from some form of psychosis to competency through the administration of antipsychotic medication was seventy to eighty percent. Chief Justice Durham maintains that the State’s experts’ reliance on these general statistics “should be given little, if any, weight.”31 She also argues that these restoration rates do not reflect Ms. Barzee’s particular symptoms and that the State’s experts almost exclusively relied on these restoration rates.32 But while the State’s experts used these restoration rates to support their conclusions, it is clear from the record that their opinions were formed by their own clinical experience and, in the case of Dr. Jeppson, his first-hand knowledge of Ms. Barzee as her treating physician.

¶ 94 Further, even though all of the patients in the reports relied on by the State’s experts did not share Ms. Barzee’s precise diagnosis or symptoms, the restoration rates indicated in the reports still have relevance. At the very least, the reports show success in treating patients with various forms of psychosis with antipsychotic medication. And the restoration rates are further relevant in light of Dr. Jeppson’s own experience with Ms. Barzee and restoring patients to competency in general, as well as those with symptoms similar to those of Ms. Barzee.

¶ 95 Ultimately, the district court was faced with complicated and conflicting expert testimony. In finding “the testimony of the State’s witnesses ... more persuasive,” the district court did not “ignore” the defense experts’ testimony, but carefully weighed and reviewed all of the evidence. Moreover, the district court’s decision is well reasoned and supported by the record. Therefore, we cannot say that the district court’s finding that the administration of antipsychotic medication was “substantially likely” to render Ms. Barzee competent to stand trial was “clearly erroneous.” We therefore affirm the district court’s order providing that the State may medicate Ms. Barzee.

concurring in the majority:

¶ 97 I concur in all respects with the majority opinion authored by Justice Durrant.

¶ 98 In March 2003, Defendant and her husband, Brian David Mitchell, were arrested in the company of a child who had been kidnapped and held for months. Both were charged with multiple felonies in connection with their alleged abduction of the child. Defendant was charged with aggravated burglary, aggravated sexual assault, aggravated *79kidnaping and attempted aggravated kidnap-ing, or in the alternative, conspiracy to commit aggravated kidnaping. The State filed a petition to inquire into Defendant’s competency. Two court-appointed evaluators determined that Defendant was suffering from mental illness of a psychotic nature and that her competency was “severely compromised.” Both evaluators found that due to the nature of her psychosis, Defendant had “severe impairments” in her ability to engage in the reasoned choice of legal strategies and options, and thus concluded that Defendant was “severely impaired with respect to her present capacity to consult with her counsel and participate in the proceedings against her with the reasonable degree of rational understanding.” The district court concluded that Defendant was not competent to proceed. Defendant was then transferred to the Utah State Hospital, where she currently remains.

¶ 99 Since its initial ruling, the district court has conducted two hearings to review Defendant’s competence. After the first review hearing, in August 2004, the district court determined that while Defendant was still incompetent to stand trial, there was a “substantial probability that [she] may become competent in the foreseeable future.” One year later, after the second review hearing, the district court concluded that Defendant remained incompetent. The State filed a motion to compel medication. At the Medication Hearing, the district court heard testimony from Drs. Kreg Jeppson, Paul Whitehead, Raphael Morris, and Xavier Amador. Concluding that “the testimony of the State’s witnesses is more persuasive,” the district court granted the State’s motion to compel medication. In that process, the district court made, and entered, a number of factual findings based upon the persuasiveness, content, and import of the testimony.

¶ 100 It is sufficient for my purpose here to simply note that the mental health professionals presented by the State and by the defendant did not agree on the need, danger, or outcome of the proposed forced medication of Ms. Barzee. The district court, to whom all of this evidence was presented, thoughtfully considered the differences and, as all judges are required to do, made a decision.

¶ 101 The Chief Justice is of the opinion that the factual findings of the district court, since they obviously implicate legal conclusions, deserve more careful examination on appellate review. This is accomplished, in this case, by reviewing the expert testimony presented to the trial court, considering it in the abstract, and, essentially, reweighing it to reach a different result.

¶ 102 I find no reason to reject the work of the district court. I am distressed by the trend in our eases of late to substitute our opinion of how the facts should be valued for that of the trial courts. I do not see that as our role. Apparently, some of my colleagues do.

¶ 103 As is always the case, the record on appeal and the argument presented by counsel in briefs and orally are woefully inadequate substitutes for the proceedings before the trial court. Our role, in my opinion, is to give helpful, clear, and usable direction on how the trial courts are to approach these tasks, not to accept the invitation by the unsuccessful party at trial to rejudge the matter in total. I believe this to be true in all cases, not just cases deemed to be somehow less important. I certainly agree that deprivation of essential and fundamental rights protected by the federal and state constitutions are important, and demand of us our best work. Such is the case of Ms. Barzee, and of the victim of the crimes with which Ms. Barzee is charged. However, to the parties, and the victims, all matters that come before us are just as important. All matters deserve the same care by us.

¶ 104 Exhaustive review of evidence in the record, and additional commentary on mental health generally, do not justify substitution of our judgment for that of the trial judge. Evaluation of witnesses has traditionally been left in the able hands of the trial courts. Assuming that role at the appellate level furthers a new trend, one with which I do not agree. Appellate courts should review, not redo. To do otherwise cripples the adversary process and invites not only additional *80appeals, but delay, inefficiency, and in my view, injustice.