dissenting.

In Wetherhorn v. Alaska Psychiatric Institute,1 we considered the constitutionality of Alaska's statutory provisions that govern the cireumstances whereby the State can involuntarily commit a person with mental illness for 30 days in order to evaluate and treat that person. Wetherhorn provides the legal context in which to consider Joan's case.

One of the orders that Wetherhorn appealed was the superior court's order approving her involuntary commitment for 30 days. We began our discussion by observing that "Itlhe United States Supreme Court has characterized involuntary commitment for a mental disorder as a 'massive curtailment of liberty' that cannot be accomplished without due process of law."2 We emphasized the Supreme Court's "repeated admonition that, given the importance of the liberty right involved, a person may not be involuntarily committed if they 'are dangerous to no one and can live safely in freedom.'"3 We explained that the Supreme Court has determined "before a person can be involuntarily committed, the [trial] court must find in addition to mental illness either: (1) that the person presents a danger to self or others; or (2) that the person is 'helpless to avoid the *603hazards of freedom either through his own efforts or with the aid of willing family members or friends.'"4 We stated:

The two findings required in addition to a finding of mental illness are each aimed at different types of harm. The first finding, of "danger to self or others," is concerned with active forms of harm, where the respondent has demonstrated the affirmative ability or inclination to inflict harm to self or another person. The second finding is concerned with a more passive condition, whereby the respondent is so unable to function that he or she cannot exist safely outside an institutional framework due to an inability to respond to the essential demands of daily life.[5]

We then analyzed the Alaska statute that correlates with the requisite findings for both types of harm. We explained:

Alaska statutes address both types of harm. Alaska Statute 47.80.785(c) permits the court to "commit the respondent to a treatment facility for not more than thirty days if it finds, by clear and convincing evidence, that the respondent is mentally ill and as a result is likely to cause harm to the respondent or others or is gravely disabled."[6]

We emphasized the importance of the "clear and convincing" standard of proof. We noted that another Supreme Court case, Addington v. Texas,7 "was concerned with the standard of evidentiary proof required in civil commitment statutes."8 We explained that Addington held that the standard of proof:

must be greater than the preponderance of evidence standard but less than the beyond a reasonable doubt standard. As Adding-ton noted, "[i]ncreasing the burden of proof is one way to impress the factfinder with the importance of the decision and thereby perhaps to reduce the chances that inappropriate commitments will be ordered."[9]

As I will show, in Joan's case the superior court unfortunately failed to give meaning to the heightened "clear and convincing" standard of proof. Now, doubly unfortunate, this court also fails to meaningfully apply this heightened standard on review. In my view, the evidence adduced against Joan was conclusory and speculative, and may not have satisfied even the lower preponderance of the evidence standard. Because Joan was involuntarily committed to the Alaska Psychiatric Institute (APT) by evidence that was far less than clear and convincing, I respectfully dissent from the opinion of the court.

I.

In order to involuntarily commit Joan to API for 30 days, the superior court was required to find by "clear and convincing" evidence that she was "likely to cause harm to [herself] or others" due to her mental illness,10 meaning she posed "a substantial risk of bodily harm" to herself or others "as manifested by recent behavior causing, attempting, or threatening harm," or by "a current intent to carry out plans of serious harm."11 The superior court was also required to consider whether a less restrictive alternative was available.12 I urge the reader to carefully parse and weigh Dr. Bell's and Dr. Parker's testimony in light of these statutory standards as measured by the mandated standard of proof: clear and convincing evidence.

A.

Clear and convincing evidence is evidence that produces "a firm belief or conviction *604about the existence of a fact to be proved."13 Here, the State's experts Dr. Parker and Dr. Bell offered equivocal, speculative, and con-clusory opinions on the issues of whether Joan posed a substantial threat of harm to herself or others and whether a less restrictive alternative was available.

Dr. Parker testified that Joan was "potentially a risk to herself" due to her unstable emotional state (emphasis added), but admitted Joan had not directly expressed any intent to harm herself or anyone else during her present hospitalization and had not actually harmed herself or anyone else, other than by using drugs. When asked what danger, if any, Joan posed to herself based on the fact that she had used illegal substances, Dr. Parker responded, "amphetamines aren't good for the system, but ... I can't say there's some imminent danger from that. It's certainly self-destructive, but ... plenty of people engage in self-destructive behavior." (Emphasis added.)

Dr. Bell testified that Joan had the "potential to cause harm" to herself and others due to her unstable emotions and "excessive anger," and "speculate[d]" that her anger was one of the reasons she had a bruise around her eye.14 (Emphasis added.) When asked if he was aware of any recent threats or attempts to harm herself, Dr. Bell testified that Joan had not expressed any desire to harm herself and had done nothing to purposefully harm herself during her present hospitalization, other than the evidence of her drug use. He testified the harm from using illegal substances would be a "further destabilization" of Joan's mental and emotional state; he could not determine whether Joan had suffered any organic brain damage as a result of drug use. Dr. Bell testified, "I think if she leaves the hospital, [Joan] would go out and do drugs" in order to control the stress caused by her belief that she is responsible for keeping the universe centered. (Emphasis added.)

"Likely" is defined as "an equivalent to probably."15 "Probable" is defined as "likely to be or become true or real."16 "Possible" is defined as "being something that may or may not occur" and "being something that may or may not be true or actual."17 "Potentially" is the adverb form of "potential," which is defined as "existing in possibility" and "capable of development into actuality."18 Garner explains that "probable; likely; possible" "in order of decreasing strength-express gradations of the relative chance that something might happen."19 When an expert witness testifies that something is "possible" or "potential," or that something "might" be true or "might" occur, or that he "thinks" it might, this indicates a lesser chance that something is true or likely than if the expert testified that it was "probable"-'"probable" in this sense meaning more likely true than not true. Certainly testimony that something is "possible" or "potential" does not establish the truth or likelihood of the thing-here specifically a "substantial risk" of causing harm to self or others-and certainly not by clear and convincing evidence, that is, a "firm belief or conviction" that the substantial risk of harm existed.20

*605Based on this speculative and conclusory evidence, and on our decision in v. Alaska Psychiatric Institute,21 the superior court "found that Joan had bipolar disorder and that Joan's mental illness altered her perception of reality. This caused Joan to use drugs, and she was therefore likely to cause serious harm to herself through illegal drug use."22 But the court stressed that Joan was "not being detained because she [was] a drug addict."

This court also relies on E.P. in affirming the superior court's finding. But E.P. is manifestly distinguishable from Joan's case in several respects-there was clear evidence that E.P. was addicted to huffing gas, that his addiction had caused organic brain damage resulting in dementia and personality disorder, and that he intended to continue huffing gas if released.23 Unlike in E.P., there is no evidence that Joan was addicted to any drugs, that her drug use had caused or would likely cause substantial bodily harm such as organic brain damage, and she never expressed a clear intent to continue using drugs if released. Of greater significance are the concessions by both doctors that Joan had not directly expressed any intent to harm herself or anyone else during her present hospitalization and had not actually harmed herself or anyone else, other than by using drugs. It almost appears that the main reason underlying Joan's involuntary commitment was that she had abused drugs.

Though Joan's case is unlike E.P.'s case, E.P. reveals several relevant principles. We cautioned in K.P. that the statutory definition of "mental illness" does not include "drug addiction" in and of itself 24-and by necessary implication mental illness also does not include mere drug use and abuse, even if such abuse is harmful. We explained:

We distinguish [E.P.'s] case from one in which an addicted person with full mental capacity chooses to continue abusing harmful substances, no matter how unwise one might consider that choice. In such a case, the person's intent to harm himself by abusing substances results from drug addiction alone, which the legislature excluded from the definition of "mental illness." In [E.P.'s] case, E.P.'s decision to harm himself by abusing substances results from his brain damage, and therefore meets the statutory standards."[25]

It is undisputed that Joan has a diagnosis of bipolar disorder, thus meeting the statutory definition of mental illness. But there was no evidence that Joan was addicted to drugs; rather, it appears that she was simply an abuser of these harmful substances. There was also no evidence that Joan's drug use was caused by any organic brain damage, or even by her bipolar disorder. Joan's case thus appears to fall somewhere between the case of a person "with full mental capacity" (i.e., no mental illness) and E.P.'s case where he suffered organic brain damage from huffing gas and it was his brain damage that caused him to continue to huff gas. Dr. Parker's and Dr. Bell's testimony attempted to relate the risk of harm they argued Joan may cause to herself by her use of drugs to the potentially deleterious effect that these drugs could have on her mental condition. The superior court and this court rely on this testimony to conclude that Joan's case is like E.P.'s case and is not the kind of case that would be excluded because of the limiting definition of "mental illness" and the distin*606guishing example quoted above from E.P.26 I am unconvinced. Joan cannot legally be involuntarily committed merely because she suffers from bipolar disorder. Nor can she legally be involuntarily committed merely because she abuses drugs, or even because her abuse of drugs would be harmful in the same way that drug abuse would be harmful to "an addicted person with full mental capacity."27 Rather, in order to legally involuntarily commit Joan, there must be clear and convincing evidence that as a result of her mental ill-mess she is likely to cause harm to herself or others.28 For the reasons explained above, I believe the evidence is insufficient to support the superior court's finding by a clear and convincing standard of proof that Joan posed "a substantial risk of bodily harm" to herself or others.29

B.

Both doctors also dismissed the possibility of a less restrictive alternative for Joan, even though they had not explored alternative options. Dr. Parker and Dr. Bell testified that they never communicated with any of Joan's family members to see if they could provide a wraparound plan for her. And even though Joan had previously been treated by Dr. Baker, a psychiatrist, neither doctor contacted Dr. Baker to discuss Joan's condition or to see if Dr. Baker could further treat Joan as an alternative to commitment to a psychiatric institution. Dr. Bell testified he did not know whether Joan would be an appropriate candidate for voluntary treatment, but he did not believe a less restrictive alternative would meet her needs because she was "liable to erupt with labile [i.e., changeable] emotions" and it was "possible she could injure another person in that state of mind." (Emphasis added.) Dr. Parker testified that he believed less restrictive options would not work for Joan because of her unstable temperament and because she had not acknowledged that she had a mental illness. Notwithstanding this testimony, the superior court found there were no less restrictive alternatives for Joan.

I start with the proposition that a mentally ill person's belief that she is not mentally ill cannot be the measure by which a court finds that there are no less restrictive alternatives; nor can a doctor's testimony that a person might display changeable emotions or "possibly" cause harm. Of greater significance, it is illogical and insufficient for a doctor to opine that there are no less restrictive alternatives when the doctor has done nothing to evaluate any less restrictive alternative. The doctors were aware, or should have been aware, that Joan was brought to the hospital by her mother. They were aware that Dr. Baker had previously provided psychiatric treatment to Joan. Their failure to contact family and Dr. Baker or to explore any other possible alternative should have caused the superior court to conclude that the State, which bears the burden of proof, failed to prove that there were no less restrictive alternatives.

In Wetherhorn we said:

[We] agree[d] with the Supreme Court of Washington that "[it is not enough to show that care and treatment of an individual's mental illness would be preferred or beneficial or even in his best interests. Indeed, AS 47.80.7830 does require more than a best interests determination. For example, it requires that the petition for commitment "allege that the evaluation staff has considered but has not found that there are any less restrictive alternatives available" .... As further protection, the statute directs the court to make these findings by "clear and convincing" evidence.[30]

*607Under these circumstances, where the testifying doctors utterly failed to make any effort to contact Joan's prior treating physician or her family to explore less restrictive alternatives, I believe the doctors' conclusory opinions are insufficient under any standard of proof to support the superior court's finding that no less restrictive option was available for Joan.

Because there was no clear and convincing evidence that Joan presented a substantial risk of harm to herself or others, and that she was "helpless to avoid the hazards of freedom ... with the aid of willing family members or friends"31 or her prior treating psychiatrist-in other words, that there were no less restrictive alternatives to involuntary commitment-I would reverse the superior court's involuntary commitment order.

II.

I also disagree with this court's resolution of the mootness question. The court today recognizes for the first time that the collateral consequences doctrine will permit an appeal of an otherwise moot order of involuntary commitment, provided that no previous commitments have been ordered. The court hedges on whether it will recognize in later cases the collateral consequences exception to the mootness doctrine for persons who have been involuntarily committed more than onee. While I agree that collateral consequences justify not applying the mootness doctrine in involuntary commitment cases, I would go farther: it is my view that the supreme court must accept and decide on the merits every appeal of an order of involuntary commitment.

The mootness doctrine is a judicially constructed doctrine to give the courts a means to avoid addressing cases that no longer present "live controversies.32 But the Alaska Legislature has codified by statute the citizen's right to appeal a superior court order for involuntary commitment, and rightly so. Alaska Statute 47.30.765 unambiguously provides that a "respondent has the right to an appeal from an order of involuntary commitment." The statute's language could not be any plainer. I believe that the statute supplants the judicially created mootness doctrine in the limited cireumstance of involuntary commitment orders. This is similar to the situation where the legislature supplants the common law with statutory law: the courts are bound to apply statutory law that supersedes common law.33

The right to appeal is no right at all if it is merely the right to pay the filing fee and file an appellate brief, only to be told that your appeal is moot and the court will not reach the merits. Even though the respondent will in every case have already completed her 80-day commitment by the time her appeal is ripe, and therefore the supreme court could not undo the commitment if the respondent's commitment order were wrongly issued, I contend that any order for involuntary commitment that is erroneously issued remains a "live controversy" for the respondent for the remainder of the respondent's life. Of first importance, the citizen's liberty has been alleged to have been wrongfully taken by court process; the court should afford the citizen the opportunity to prove the error and, if proven, obtain judicial acknowledgment that the order was erroneously issued. Giving the citizen this opportunity will assure the citizen that she will be heard, and that if a lower court has erred, that error will not go unnoticed or unremedied, at least to the extent that the erroneous order will be reversed and vacated. Public confidence in the *608judicial branch demands that we hold ourselves accountable.

Second, in this age of prevalent information mining, collection, and storage into increasingly large, interconnected, and search-able data banks, the fact that a citizen has been involuntarily committed to a mental institution will follow that individual for all of her life. She should be given the means to effectively challenge that order through appeal regardless of the fact that by the time her appeal is ripe for decision, the 30 days will have long since expired and she will have been released from State custody. The injury inflicted by an erroneously issued order of involuntary commitment "lives" until the wrong is righted. I am at a loss to understand how a citizen can be ordered to be involuntarily committed for 230 days and be precluded from appealing this order merely because it is practically impossible to perfect an appeal of an order that by its terms will expire in 30 days.

Ending where I began, we-the Alaska Supreme Court-along with our legislature and the United States Supreme Court, have recognized that "involuntary commitment for a mental disorder [is] a 'massive curtailment of liberty' that cannot be accomplished without due process of law."34 I believe that this court should accept every appeal of an order of involuntary commitment as a matter of being faithful to the citizen's right to due process of law. I also believe that, apart from considerations of due process, where the legislature has codified the right to appeal, we have a duty to honor and give real meaning to the law and to the right that it bestows.

OPINION

I. INTRODUCTION

An adult woman diagnosed with a mental illness appeals her already completed 30-day *595involuntary commitment to Alaska Psychiatric Institute (APT), arguing the evidence did not support the superior court's findings that: (1) she was likely to cause harm to herself or others due to her mental illness; and (2) API was the least restrictive alternative placement for her. Because our existing case law provides that an evidentiary-based "weight of the evidence" challenge to a completed involuntary commitment is moot absent accompanying legal issues appropriate for decision under the mootness doctrine's public interest exception, we asked the parties to submit supplemental briefing on mootness. As a result, we now confront a question not directly raised in our earlier cases: should our application of the mootness doctrine in this context accommodate the importance of collateral consequences arising from an involuntary commitment? We answer that question "yes" and therefore reach the merits of this appeal. On the merits, we affirm the superior court's involuntary commitment order.

II. FACTS AND PROCEEDINGS

In February 2010 Joan K. disappeared from her mother's house.1 Three weeks later Joan's mother received a telephone call from an unknown woman saying Joan was "confused or impaired" and should be picked up. Joan's mother found Joan and brought her to Fairbanks Memorial Hospital (FMH). Emergency room staff examined Joan and found her "very confused"; she also tested positive for amphetamines and cocaine. Joan was admitted to the psychiatric ward, where she had been voluntarily hospitalized twice in November 2009.

The next day a FMH staff physician applied for an ex parte order authorizing Joan's involuntary hospitalization for a mental health evaluation.2 The superior court granted the application. Dr. Victor Bell, a psychiatrist, observed Joan over the course of five days. Dr. Stephen Parker, a psychologist, also observed Joan on two occasions. Neither Dr. Bell nor Dr. Parker contacted Joan's family or the psychiatrist who had previously treated Joan for a short time.

The superior court held a 30-day commitment hearing on March 1, 2010.3 The court found that Joan had bipolar disorder and that this mental illness altered her perception of reality, causing Joan to use drugs. The superior court found she was therefore likely to cause serious harm to herself through illegal drug use. The court stressed that Joan was "not being detained because she [was] a drug addict."

The superior court also found that as a result of her mental illness Joan was likely to cause harm to others, based on an incident at FMH and evidence of her unstable emotions. The court said Joan might "present aggressively out in the public in front of lay people who may not know of [Joan's] mental disability and who may react violently or who may be hurt by her, may not know how to talk her down and certainly are not going to have drugs available [such as Valium] to ameliorate her mood as was true here."

The superior court found no less restrictive facility than API would adequately protect Joan and the public. Finding Joan had refused voluntary treatment, the court ordered her committed to API for a period not to exceed 30 days. Although the record does not indicate when API actually released Joan, 30 days from her commitment date was April 1, 2010.

Joan appeals the superior court's 30-day commitment order.

III. STANDARD OF REVIEW

Mootness is a matter of judicial policy and its application is a question of *596law.4 We adopt the rule of law that is "most persuasive in light of precedent, reason, and policy."5 We review fact findings in involuntary commitment proceedings for clear error, reversing only if we are left with a "definite and firm conviction that a mistake has been made."6 We review related questions of law de novo, "including whether the fact findings meet the statutory standards for involuntary commitment."7

IV. DISCUSSION

A. Mootness And Collateral Consequences

1. Framing the issue

In Wetherhorn v. Alaska Psychiatric Institute we established that commitment-order appeals based on assertions of insufficient evidence are moot if the commitment period has passed, subject to the public interest exception.8 Because Joan's post-release appeal from the superior court's commitment order is based on an assertion of insufficient evidence and neither Joan nor the State discussed mootness in their original briefs, we ordered supplemental briefing on that issue.

In her supplemental briefing, Joan suggests we should overrule Wetherkhorn because it mistakenly focused on release from commitment, rather than vacating the commitment order, as the relief sought in a commitment-order appeal. She also argues the public interest exception to mootness applies because "(unless this court reviews commitment orders for sufficiency of the evidence, the masters and trial court judges hearing these cases will have no standards by which to measure the cases before them." Finally, she argues that we should adopt the collateral consequences exception to mootness in commitment-order appeals. The State responds that Wetherkorn mandates dismissal of Joan's appeal as moot because: (1) the public interest exception to mootness does not apply; and (2) Joan has not established any actual collateral consequences resulting from her commitment order.

We ordered oral argument on the mootness question, directing that the parties be prepared to discuss the authority and appropriateness of issuing a vacatur order to remedy possible collateral consequences arising from an otherwise-moot commitment order.9

2. We decline to consider overturning Wetherhorn, but we adopt the collateral consequences exception to mootness in this context.

a. Issues not considered

In response to our supplemental briefing order, Joan asserts that in Wetherhorn we "erred in focusing on the period of the commitment rather than the relief that Wetherhorn sought" when we held that an evidentiary-based challenge to a completed commitment is generally moot. Joan asserts the true relief Wetherhorn sought was vacating the "gravely disabled" finding; although not expressly stating it, she suggests the true relief she seeks is vacating the finding that she is a danger to herself or others.

Our order for supplemental briefing did not anticipate questioning Wethkerkorn's fun*597damental holding, nor did Joan address the standards we impose for overturning our precedent.10 We therefore do not consider overturning Wetherhorn's holding that an ev-identiary-based challenge to a completed 30-day commitment generally is moot.

Joan also argues her evidentiary-based appeal of her 30-day commitment should be considered under the public interest exception to mootness. She contends that given her bipolar disorder, she is likely to face future commitment proceedings and the issues of this case are likely to be repeated. She also contends we "must provide guidance to the trial courts" on the evidence necessary to meet the legal standards for: (1) finding someone a danger to self or others; and (2) least restrictive treatment alternatives. In connection with this latter argument, Joan cursorily asserts in her supplemental opening brief that her statutory right to appeal her commitment order is "meaningless" if we choose not to review the order.11 Although Joan's interpretation of the statute as overriding the judicial policy of not deciding moot cases appears overbroad, we do not need to address this argument, or her overall public interest exception argument, because we agree with Joan that we should adopt the collateral consequences exception to mootness in this context and consider the merits of her appeal.12

b. Collateral consequences exception to mootness

Joan notes that several other courts have applied the collateral consequences exception to mootness in the involuntary commitment context. She points to social stigma,13 adverse employment restrictions,14 application in future legal proceedings,15 and restrictions on the right to possess firearms16 as recognized consequences from involuntary commitment orders. She argues we should adopt the collateral consequences exception to mootness and urges us to do so without adopting case-specific requirements; she contends a commitment proceeding will not focus on future collateral consequences and the record available for appellate review will be inadequate. Joan also argues that when we decline to review the merits of involuntary commitment orders, we should vacate them rather than leaving them in place.

The State acknowledges there are collateral consequences from an involuntary commitment order, but argues that: (1) an exception from the general rule of mootness requires a case-specific analysis; and (2) Joan has not established any actual collateral consequences arising from her involuntary commitment order. At oral argument the State also argued that certain collateral consequences from an involuntary commitment order, such as restrictions on the right to possess a firearm, are important, and we therefore should consider an appeal's merits rather than simply vacate the underlying commitment order.

We have previously recognized that the collateral consequences doctrine "allows *598courts to decide otherwise-moot cases when a judgment may carry indirect consequences in addition to its direct force, either as a matter of legal rules or as a matter of practical effect."17 Both Joan and the State have articulated sound reasons to adopt the doctrine, at least to some extent, in the involuntary commitment order context.

We conclude that there are sufficient general collateral consequences, without the need for a particularized showing, to apply the doctrine in an otherwise-moot appeal from a person's first involuntary commitment order. But we do note that some number of prior involuntary commitment orders would likely eliminate the possibility of additional collateral consequences, precluding the doctrine's application.18

3. Joan's commitment order is reviewable under the collateral consequences exception to mootness.

Based on our adoption of the collateral consequences exception to mootness in the involuntary commitment order context, we agree that Joan's commitment order, her first, is reviewable.

B. Merits of Joan's Appeal

1. Legal framework

To involuntarily commit someone to a treatment facility for up to 80 days, a court must first find, by clear and convincing evidence, that the person "is mentally ill and as a result is likely to cause harm to [self] or others or is gravely disabled."19 Although the statute does not define "harm," we have found AS 47.30.915(10) relevant,20 defining "likely to cause serious harm" as when a person:

(A) poses a substantial risk of bodily harm to that person's self, as manifested by recent behavior causing, attempting, or threatening that harm;

(B) poses a substantial risk of harm to others as manifested by recent behavior causing, attempting, or threatening harm, and is likely in the near future to cause physical injury, physical abuse, or substantial property damage to another person; or

(C) manifests a current intent to carry out plans of serious harm to that person's self or another[.]

The court must also consider whether a less restrictive alternative would provide adequate treatment.21 Alaska Statute *59947.80.915(9) defines "least restrictive alternative" as treatment conditions that:

(A) are no more harsh, hazardous, or intrusive than necessary to achieve the treatment objectives of the patient; and

(B) involve no restrictions on physical movement nor supervised residence or inpatient care except as reasonably necessary for the administration of treatment or the protection of the patient or others from physical injury[.]

As discussed below, we affirm the superior court's findings by clear and convincing evidence that Joan suffered from a mental illness and that as a result she posed a substantial risk of bodily harm to herself under AS 47.30.915(10)(A)"22 We therefore do not need to discuss the evidence or the superior court's finding that Joan posed a substantial risk of harm to others.

2. Expert witness testimony

Dr. Bell testified Joan suffered from bipolar disorder, a serious mental illness which caused Joan to suffer mixed emotional states with manic elements, depressed elements, and psychotic manifestations. Joan told Dr. Bell that she could sense other people's feelings, that Fairbanks was the center of the galaxy and the separation point between heaven and hell, and that she was responsible for maintaining the balance between heaven and hell. Joan also told him he was preserib-ing the wrong medicines because she needed amphetamines and OxyContin to deal with the stress of sensing others' thoughts. Dr. Bell was unable to determine whether Joan had organic brain damage because her manic bipolar disorder symptoms prevented a complete examination. Dr. Bell thought Joan had been abusing drugs for the three weeks she disappeared because her admission drug sereens were positive for amphetamines and cocaine and because she appeared to be in a state of "lethargy or torpor," which typically follows prolonged stimulant abuse.

Dr. Bell was primarily concerned Joan would harm herself by further illegal drug use. -He thought Joan would use drugs if she left the hospital because she believed she needed amphetamines and OxyContin to manage the stress associated with her delusions. Dr. Bell testified that using illegal substances in Joan's mental state would make her thought process "so completely disorganized" that she would not "know what she was doing." Using drugs would also cause "further de-stabilization of [Joan's] mental and mood state, which would aggravate her ability to conform ... to reasonably safe behaviors."

Dr. Bell testified there was not a less restrictive facility than API that could meet Joan's needs. He testified Joan needed a "very secure mental health unit" that would "very closely control[ ]" her and provide "a lot of emotional support, with careful control of her sleep pattern and regular appetite." Dr. Bell thought Joan would "need a long period of treatment, though it could be concluded within 30 days were she to accept medication reliably and begin to understand how drugs and alcohol impact on the exacerbation of her mental illness."

Dr. Parker testified he had not personally diagnosed Joan, but her records indicated she had bipolar disorder and a history of attention deficit hyperactivity disorder, poly-substance abuse, and alcohol abuse. In Dr. Parker's opinion, Joan was at risk of causing herself bodily harm through drug use if she were not committed. He said Joan's mental stability "can change very rapidly" due to her bipolar disorder. He also noted she had disappeared for three weeks and used drugs prior to her hospital admission. Dr. Parker testified Joan locked herself in a bathroom and threatened to blow herself up in November 2009. When the superior court asked what danger, if any, illegal drug use posed for Joan, Dr. Parker replied it would not be "good for the system" and was "self-destructive," but he could not say it posed "immi*600nent danger." Dr. Parker acknowledged Joan had not expressed suicidal ideations after her February hospital admittance, nor did she directly indicate any intent to harm herself or others.

Like Dr. Bell, Dr. Parker testified there was not a less restrictive facility than API that could meet Joan's needs. Dr. Parker thought Joan needed to come out of her current manic episode and return to her baseline before release, otherwise she would continue her "uncontrolled manic behavior" and substance abuse. Dr. Parker noted that the day before the hearing nurses had to medicate Joan after an incident at the nurses' station. He also thought outpatient treatment was not a viable option for Joan because outpatient psychiatry or psychology requires patients to "have some kind of insight on their behavior and some ... sort of consistency of behavior." Joan had denied to him that she had any mental illness or needed treatment; due to her lack of insight, Dr. Parker thought it "very unlikely" Joan would follow through with outpatient treatment even if she said she would. Dr. Parker also testified that for a family wrap-around plan or 24-hour surveillance by a family member to work Joan would have to agree, and Joan changed her mind too frequently for such plans.

3. Substantial risk of harm to self

Joan does not contest the superior court's finding of a mental illness. Joan instead contends the superior court erred by finding that due to her mental illness she was likely to cause harm to herself, arguing: (1) there was no evidence illegal drug use physically harmed her and she did not make affirmative statements that she would use illegal drugs if released; and (2) she did not manifest a current intent to carry out a plan to seriously harm herself.23

We decided a similar issue in L.P. v. Alaska Psychiatric Institute.24 E.P. had a history of inhaling gasoline fumes and other substances to get high (huffing),25 which caused organic brain damage.26 He was involuntarily committed to API several times and "maintained that, if discharged from API, he [would] likely go back to huffing."27 We held E.P.'s organic brain damage was "a condition apart from, and more than, his drug addiction" and met the statutory definition of "mental illness" under AS 47.30.915(12).28 We also held the evidence supported the multiple masters' reports and superior court orders, which found huffing gas damaged E.P.'s brain and E.P. would continue huffing gas if released.29 Thus "E.P.'s continued intent to huff gas, as a result of his impaired judgment and understanding, [met] the standards of AS 47.30.915(10)(A) and (C)."30 "E.P.'s intent to huff gas constitute[d] intent to cause himself bodily harm, and ... resultled] from his mental illness.31 Here, the superior court expressly stated it was not finding Joan was likely to harm herself based on drug addiction alone. The superior court noted and, as we did in E.P., distinguished Joan's case "from one in which an addicted person with full mental capacity chooses to continue abusing harmful substances, no matter how unwise one might consider that choice."32 The superior court stated:

I find that [Joan] is suffering from a mental illness. She has bipolar disorder, which according to Dr. Bell's testimony, renders her unable to perceive reality. Her bipolar disorder is manifesting itself in psychosis right now. She is feeling that she is experiencing the feelings of other people, both close and far away, that Fairbanks is at the center point between heaven and hell and that she is responsible for *601maintaining balance between heaven and hell and keeping the universe centered.

She has advised Dr. Bell that she believes that she must take [amphetamines] and oxycontin to manage the stresses that are caused by her current situation; that is, by the fact that she is responsible for balancing the equal point between heaven and hell....

[Joan] is not being detained because she is a drug addict. She is taking drugs because of her perception of reality caused by her bipolar disorder and ... she's being detained because her mental illness is causing her to take the drugs.

The superior court's finding is amply supported by Dr. Bell's testimony.

The superior court also heard evidence that fllegal drug use would "pose a substantial risk of bodily harm" to Joan by exacerbating her mental illness.33 Dr. Bell testified if Joan used illegal drugs in her current mental state, her thought process would get "so completely disorganized" that she would not "know what she was doing." IMegal drug use would also cause "further de-stabilization of [Joan's]! mental and mood state, which would aggravate her ability to conform ... to reasonably safe behaviors." In short, Dr. Bell's testimony supported the conclusion that Joan's continued illegal drug use would exacerbate her mental illness and cause a self-destructive downward spiral of her mental and physical health.

Finally, the superior court heard evidence regarding Joan's "recent behavior causing, attempting, or threatening" harm to herself by illegal drug use.34 Although Dr. Bell and Dr. Parker both acknowledged Joan neither articulated a desire to harm herself nor did so beyond using illegal drugs, both thought she would continue using illegal drugs if she were not committed. Joan's emergency room toxicology report showed traces of amphetamines and cocaine in her system. Dr. Bell testified Joan showed symptoms of "lethargy or torpor" that follow stimulant abuse. Dr. Bell also testified Joan showed symptoms of opioid withdrawal. The plain text of AS 47.80.915(10)(A) directs courts to consider "recent behavior" and does not, as Joan argues, require affirmative statements regarding future drug use.35 Even if affirmative statements were required, Joan told Dr. Bell he was prescribing the wrong medicines because she needed amphetamines and OxyContin.36 This is sufficient to support the superior court's finding that Joan was likely to continue using illegal drugs if released.

Based on these findings, the superior court did not err by finding clear and convincing evidence that, under AS 47.30.785 and AS 47.30.915(10)(A), Joan was likely to cause harm to herself due to her mental illness.

4. Least restrictive alternative placement

An important principle of civil commitment in Alaska is to treat persons "in the least restrictive alternative environment consistent with their treatment needs."37

Joan argues the superior court erred in finding commitment to API would be the least restrictive alternative placement. Joan also contends no testimony supported a finding that she refused outpatient treatment or a home placement, particularly in light of Dr. Bell's and Dr. Parker's decisions not to contact her family or prior psychiatrist to ask about Joan's potential success in such alternative settings. Joan's second argument reflects a misunderstanding of the superior court's findings-the court found outpatient treatment was not a viable option, and there*602fore the lack of evidence that Joan refused voluntary outpatient treatment is irrelevant.

The superior court found there was "[nlo less restrictive facility [that] would adequately protect [Joan] and the public." The court explained:

API is an appropriate treatment facility, that there is no less restrictive facility that would adequately protect the respondent and the public at this time. The reason for that finding is Dr. Parker's testimony that in order for a family wraparound to work or 24-hour surveillance by a family to work, she would have to agree to it and he has witnessed her changing her mind rapidly about what she will do and what she will not do. So, I can't trust that committing her to her family's care would be a less restrictive alternative that would likely work for her.

The record supports the superior court's finding.

First, Dr. Bell and Dr. Parker both testified there was no less restrictive facility than API that could meet Joan's needs. Dr. Bell testified Joan needed a "very secure mental health unit" that would "very closely control[ ]" her and provide "a lot of emotional support, with careful control of her sleep pattern and regular appetite." Dr. Bell thought Joan "need[ed] a long period of treatment, though it could be concluded within 30 days were she to accept medication reliably and begin to understand how drugs and alcohol impact on the exacerbation of her mental illness." Dr. Parker testified Joan needed to come out of her current manic episode and return to her baseline before being released, otherwise she would continue her "uncontrolled manic behavior" and substance abuse.

Second, Dr. Parker testified outpatient psychiatry or psychology require a patient stable enough to have insight into one's behavior and some "sort of consistency of behavior." Joan was not stable because she had changeable emotions and could change her mind "from one minute to the next." Joan also lacked perspective regarding her bipolar disorder, denying she had any mental illness or needed treatment. Because of Joan's lack of insight, Dr. Parker thought it "very unlikely" she would follow through with outpatient treatment even if she said she would.

The superior court did not err by finding API was the least restrictive placement.

V. CONCLUSION

We AFFIRM the superior court's involuntary commitment order on its merits.