Abstract

A discussion of the use of advance directives in health care deals very often with end of life issues and focuses on matters concerning the continuance or discontinuance of life sustaining treatments and/or life supporting technology.  As the use of advance directives for psychiatric/mental health care becomes more widespread, the respect for the wishes of those who have completed these psychiatric advance directives (PADs) as well as protocols for overriding their directives will need to be discussed and explored.  On the positive side, PADs serve to promote patient involvement in treatment, to expedite psychiatric care and to reduce coercive crisis interventions.  On the other hand, as with other health care directives, unless they are carefully constructed with competent involvement of the author of the directive, they can invite a host of legal and ethical challenges in their implementation.  A study of the limited but growing use of PADs will help to inform this evolving contribution to self-determination and autonomy in the treatment of mental illness.

Psychiatric Advance Directives:

Safeguarding Patient Autonomy and Promoting Recovery

Introduction

Psychiatric advance care directives are an emerging method of treatment planning designed to improve treatment outcomes and enhance the autonomy of mental health patients (Srebnik, 2006).  They are analogous to but somewhat different from general health care directives in several significant ways.  The two types of directives are both based on the same concept – care planning for a time when an individual may lack the capacity to make treatment decisions for him/herself.  Their content, however, is quite different.  General health care directives attempt to guarantee, for those who execute them, a good death, while psychiatric health care directives (PADs) endeavor to secure, for a more specific population, a good life (Backlar, 1997).

            A general health care directive deals with terminal/end of life conditions and most individuals will have the occasion to access a general health care advance directive only once.  In the case of PADs, however, it is often the individual’s experience of one mental health crisis treatment that prompts him/her to communicate preferences and directives for anticipated future crises (Backlar, 1997).

            Beginning with California in 1976, all states have subsequently enacted advance directive statutes of some sort including living wills, durable powers of attorney or both (Swanson, 2006).  In 1990, the federal Patient Self-Determination Act (PDSA) was enacted to promote the use of these written advance directives.  Under the provisions of this law, health care facilities receiving federal funds are required to inform patients of their rights to prepare an advance directive, to inquire about and document whether patients have executed a directive, to ensure compliance with state laws and to educate health care providers regarding these legal instruments.   This reflects a commitment to value a patient’s autonomy as a part of the right to personal liberty (Szasz, 1998). 

            Psychiatric advance directives evolved from the more general advance care directives (Harvard, 2007) and twenty-five states have now enacted PAD statutes (Swartz, 2007). Proponents of PADs envision these documents encouraging collaboration between the patient and the mental health care provider and facilitating discussions regarding plans for crisis treatment.  They also believe that an individual’s present access to providers of choice for a projected time of crisis may enhance the subject’s treatment adherence and his therapeutic confidence (Kim, 2007).

            A concern regarding advance directives, especially PADs, is that the instructions of the PAD are not consistently followed[i] and that they may be overridden (Srebnik, 2007).  Ethical issues arise when the existing instructions are ignored or overridden and when patient autonomy is not held in highest regard.  This is illustrated in the case which follows.

Case Study

Nancy Hargrave is a resident of Vermont who suffers from paranoid schizophrenia, for which she has been hospitalized multiple times since 1995 in the Vermont State Hospital.  While hospitalized in 1997, she was twice the subject of proceedings for involuntary medication; the earlier proceeding yielded a finding that Hargrave was competent to refuse medication, but the latter resulted in a finding that she was incompetent to do so.  Upon the second finding, she was administered psychiatric medication over her objection, in a non-emergency situation. 

On April 14, 1999, Nancy Hargrave executed a Durable Power of Attorney (DPOA) designating a guardian in the event of future incapacity and refusing the administration of “any and all anti-psychotic, neuroleptic, psychotropic or psychoactive medications,” and electroconvulsive therapy.[ii]

Under Vermont law, an adult may execute a DPOA which allows him to appoint a guardian in the event of his incapacity and/or to articulate preferences for or limitations to treatment.  Prior to 1998, Vermont statutes provided two mechanisms for overriding DPOA’s:  an individual could explicitly revoke his own previously executed DPOA, or a third party could petition the probate court to suspend an individual’s DPOA in conjunction with that court’s appointment of a guardian.  This override was contingent upon a ruling from the Vermont Human Services Board.   In 1998, Vermont legislature passed Act 114 which established a third procedure for overriding DPOAs of certain patients who are imprisoned or committed.  Under Act 114, health care providers may petition to involuntarily medicate imprisoned or committed persons judged mentally ill.  When the proposed medication would violate a person’s validly executed DPOA, a 45-day evaluation period would be established to “test” the provisions of the DPOA.  If the court decided at the end of this period that there was no significant clinical improvement in his or her mental state, then the court could decide to medicate the patient involuntarily with no further regard for his DPOA.

After the execution of her DPOA, Nancy Hargrave challenged the legality of Vermont Act 114 on the grounds that the provisions for abrogation of DPOAs executed by patients who have been committed violate Title II of the ADA as discriminatory towards those with mental illness.  The decision of the court was that Vermont Act 114 is discriminatory under the ADA for distinguishing action on the basis of mental illness (Hargrave v. Vermont, 2003).

Legal and Ethical Issues

            The decision in the case, Hargrave v. Vermont, establishes that the state cannot exclude involuntarily committed psychiatric patients from the statutory provisions of its medical advance directives legislation (Appelbaum, 2003) which prior to Act 114 and now again provides that the only means to override a validly constructed DPOA are revocation by the subject who created the directive or a court order in conjunction with the appointment of a guardian (Hargrave, 2003).  At the present time, this law applies only in Vermont; in other states, the law generally allows physicians to override psychiatric advance directives when in their professional judgment it is necessary to do so – for example when a patient is involuntarily committed to a facility or when a patient’s wishes are in conflict with what may be deemed the current standard of care.  The issues of when physicians and other clinicians can supersede a PAD and when a patient can revoke one remain the two most controversial issues regarding psychiatric advance directives (Harvard, 2007).

            The results of a recent survey indicate that 66 to 77 percent of over 1,000 mental health consumers from five U.S. cities indicated that they would complete PADs if they were given the opportunity and assistance to do so. The extent to which clinicians can override PADs, the situations in which they can override them, their reasons for doing so, and how they communicate these reasons to patients and their family members, have the potential either to significantly undermine PADs or to actually help to implement them more broadly (Swanson, 2006a).

            In January, 2005, Pennsylvania enacted Act 194, a mental health advance directive and powers of attorney law.  This law is consistent with most other similar state laws enacted through the 1990’s.  The PA law allows competent individuals to specify their mental health treatment wishes before becoming incapacitated and to appoint an agent to carry out their wishes during future periods of incapacity.  The law also reinforces the Patient Self-Determination Act of 1990 by requiring facilities and providers of mental health care to ask patients upon admission whether they have a PAD and to inform all patients about the availability of PADs as part of discharge planning.  Treatment facilities are also required to place a copy of the mental health advance directive in the patient’s mental health record (Swanson, 2006a).

            Pennsylvania Act 194 contains three specific sections devoted to guaranteeing that physicians may override PADs with few if any consequences.  The law states, first, that it should not be viewed to affect in any way the ability to admit a patient to a facility under either the voluntary or involuntary commitment provisions of the Mental Health Procedures Act.  Second, the compliance section contains a clause allowing a physician who cannot in good conscience comply with the instructions of an appointed agent because the instructions are contrary to accepted clinical practice and medical standards to refuse to comply so long as he or she makes every reasonable effort to assist in the transfer of the patient to another provider who is willing to comply.  Third, the law states that a physician who acts in good faith in overriding an advance directive may not be subject to criminal or civil liability or be disciplined for unprofessional conduct for refusing to comply with a PAD when in the judgment of the provider the provisions of the PAD violate accepted clinical standards of care (Mental Health Declarations Act, 2004).

            The Pennsylvania law is typical of other state laws and so the question arises, does the ease with which physicians can override the psychiatric advance directives of their patients promote patient autonomy at all; do they actually reinforce the physician’s autonomy?  Before addressing this question it is useful to review the results of a statewide survey of 164 psychiatrists in North Carolina where there is evidence of sharp division of opinion on the question of whether physicians should follow or override PADs that refuse treatment. Forty-seven percent of psychiatrists surveyed indicated that they would override a valid, competently-executed PAD that refused hospitalization and medication in the case of a non-violent psychotic patient in a hospital emergency room; they would use the discretion allowed by state law and involuntarily admit the patient.  The other 53% of psychiatrists surveyed indicated that they would follow the directives of the PAD and not admit the patient (Swanson, McCrary, 2006b).  In the opinion of Swanson et al., many clinicians would consider the risk of discharging an acutely psychotic patient without treatment far too great against the benefit of upholding his autonomy:  “Releasing the patient would seem to require quite strong support for the value of patient autonomy over beneficent paternalism in a situation where these principles conflict. Such a robust endorsement of patient autonomy is surely not universal among physicians and may be the exception rather than the rule.” (Swanson, McCrary, 2006b).

            And what of the other 53% survey respondents who would honor the PAD, would follow the patient’s documented wishes, and would not admit the patient? The profile of these psychiatrists who would honor a PAD refusal of hospitalization and medication for a psychotic patient are “(1) less likely [than the admitting psychiatrists] to actually work in a hospital emergency department or crises center; (2) less concerned about patient’s lack of insight into their illness or the potential for patient violence; (3) less legally defensive; and (4) less inclined to see coercion as intrinsically necessary given high-quality services, while being more sanguine about community-based alternatives to hospitalization for such patients” (Swanson, McCrary, 2006).

            In considering when, if ever, a patient’s advance directive should not be honored, Dan W. Brock first raises questions about whether an advance directive always accurately reflects what a patient would have wanted.  Since advance directives typically require individuals to predict what they would want well in advance of the use of the directives in treatment decision making, directives may not accurately reflect what a patient would want in contemporaneous circumstances.  This may be further reinforced when directives are written in vague language making it unclear what a patient actually wants.  A second case would be when current interests of a patient may be significantly different from conditions when the advance directive was created, i.e. the patient has suffered such severe cognitive changes that there are doubts that the personal identity of the person who executed the directive and the present person are significantly the same.  Finally, Brock offers as a third category for when advance directives may be overridden, situations in which the interests of others may warrant it.  These “others” could be family members, physicians or society in general (Brock, 1991)..

            Brock concludes his commentary rejecting the notion that advance directives should always be binding.  He proposes that institutional and judicial procedures be established as safeguards to reduce the risk to patient autonomy to “tolerable levels.” He suggests that these procedures may require going to court in some cases and consulting with ethics committees in others.  He holds that advance directives should be binding in the vast majority of cases and should be set aside only after careful consideration and by following procedures adequate to limit abuse  (Brock, 1991).

            Returning to Swanson, McCrary et al., they apply the three cases of Brock specifically to psychiatric advance directives.[iii]  They hold that severe and persistent mental illness can impair a person’s baseline ability to make and communicate reliable decisions about their health care but that this would be truer for medical situations than for psychiatric ones.  For patients with a history of psychiatric treatment, PADs may reflect treatment preferences quite accurately to the extent that these preferences have been shaped by previous experiences with the treatments in question.  A patient who has been repeatedly hospitalized and treated medically for bipolar disorder would bring much knowledge, personal experience and some authority to advance directives relating to future treatment for this condition.  A contrasting situation would be if the patient were including in a PAD directives refusing treatment of which he/she has no knowledge, such as electroconvulsive therapy (ECT).  Here a psychiatrist, familiar with the use of ECT in treating a condition such as life-threatening depression, may want to override a directive refusing a course of treatment that, in the opinion of the clinician, offers perhaps the best and only option left to the patient (Swanson, McCrary, 2006a).  For a highly controversial treatment such as ECT surrounded by other uncertainties – the patient’s informed consent and best interest – there still seems to be a question about the override of a PAD directive.

            In Brock’s second scenario, a change in the personal identity of the patient through the course of a chronic mental disorder, Swanson and McCrary et al. believe that a PAD may be susceptible to override.  Cognitive impairments are associated with major psychopathology and these impairments may be permanent.  In such cases, if the “identity” of the patient has been altered, then his/her agency with respect to the advance directive could certainly be called into question as well.  Some ethicists have opposed PADs altogether because of this potential conflict between the prior and present “self” and between patient autonomy and safety and survival issues (Swanson , McCrary, 2006a).

            Brock’s third scenario deals with instances where it is questioned whether the interests of others justify not honoring advance directives.  Some “interests of others” that could come into play here are those of family, physicians and/or society in general. This could be the case when physicians are “legally defensive” and the override of a PAD directive becomes a strategy to guard against legal liability.  Cases of involuntary commitment also fall under this category of promoting the interests of others with every state jurisdiction allowing a PAD override in such circumstances. Swanson and McCrary et al. suggest under this category, situations where physicians may assume some responsibility for past ineffective treatment and who wish to rectify or have a second chance to improve on past situations.  An example of such a scenario would be overriding an existing PAD directive to refuse medication in order to prescribe newer and more effective drugs that minimize some of the unpleasant and dangerous side effects of drugs used with the patient in the past. The psychiatrist might reason that mental health care professionals or the system itself are partly responsible for the current state of the patient, that the patient’s medication refusal is misinformed, and that therefore under these circumstances it would be irresponsible  or even negligent to honor a patient’s directive that refuses all medication (Swanson, McCrary, 2006).

            Up to this point we have been considering the override of PADs by persons other than the patient. A related issue is the revocation of provisions of an advance directive by the patient/ agent him/herself. By definition and as presented in all instructive material with respect to advance directives, the competent adult may appoint an agent and may update, change or revoke an advance directive, general or psychiatric, at any given time and as often as desired in accordance with the applicable state law (Harvard, 2007).

            A discussion of advance directives would not be complete without a consideration of the so-called “Ulysses Contract” or “self-binding contract.”  The name Ulysses refers to Homer’s example of Ulysses instructing his crew to bind him to the mast of his ship before they sailed past the irresistible sirens and to ignore his requests for release.  Thus he was able to enjoy the beautiful singing of the sirens without suffering the disastrous results that would normally have followed (Widdershoven, 2001).  With respect to health care directives what is fundamental is that an individual anticipates how he is likely to act at some future time and, being dissatisfied with the prospect of that action, makes plans in the present to try to increase the likelihood that he will decide to act (or be treated) in some other way in the future time (Davis, 2008).  John Davis presents a thorough analysis of the concept of pre-commitment to treatment through Ulysses Contracts and proposes the Diachronic Respect Principle as a guide for deciding when it would be consistent with the agent’s wishes to supersede his advance directive.  This principle states: “To actively respect an agent’s autonomy in cases of parity pre-commitment, act so that the circumstances of the agent’s life are consistent with his or her concurrent intentions over the longest period of time, qualified by reference to the intensity of the desires and the severity of the conflict.  This justification is diachronic and prospective, rather than synchronic and retrospective[iv]” (Davis, 2008).

Medical Issues

            Thus far in our consideration of PADs, we have focused on the value of patient autonomy and seen how this value may be compromised when the directives of PADs are too easily overruled by others or by the patient him/herself.  An additional goal of the use of PADs is the positive contribution they can make in supporting the mental health improvement and recovery of the patient.  When used in time of crises, PADs have the potential not only to direct and inform the crises service, but to shift the role of the patient to that of expert regarding his/her own care, thereby increasing both self-determination and respect.  Kim reports from a qualitative study of individuals who had created PADs the experience of one individual upon presenting his/her PAD at the hospital: “The doctor didn’t treat me like a nut case because some hospitals do. You know what the doctor said to me? [He said,] ‘You’ve got rights and it’s great that you know you have them’” (Scheyett, 2007).

            The very act of creating a PAD has the potential of supporting recovery by building alliances between individuals and their case managers, thus increasing perceptions of the patient that treatment needs were being met, and by creating a sense of control over one’s treatment thus increasing understanding of and compliance with treatment plans (Scheyett, 2007). 

The National Consensus Conference on Mental Health Recovery and Mental Health System Transformation identifies many ways in which PADs support the fundamental components of recovery.  PADs are tailored to the specific needs of individual; they provide individuals with an opportunity to take responsibility for their own care during crises; they use past crises as an opportunity for learning, helping the patient to identify preventive actions, coping skills and self-management techniques to be used in the future; they empower the individual while decreasing the authority of the traditional provider role.  They focus on respect, hope, peer support and responsibility, all essential components of mental health recovery (Center for Mental Health Services, 2006).  PADs also support better medication adherence and thus effectiveness, fewer and shorter hospitalizations, less undesired treatment, early recognition of a crisis, and the availability of surrogate decision-making by a trusted person.  In addition to the benefits for the patient, PADs provide a patient’s family with knowledge and awareness of the patient’s own wishes and therefore a greater ability to support recovery (Swanson, 2000).

            To increase the likelihood that care will be consistent with advance directives and thus promote recovery, having a surrogate decision maker is desirable.  Only about half the persons who complete PADs appoint a surrogate decision maker and surrogates are involved in only about 31% of crisis events in which directives are accessed (Srebnik, 2007).  There is no data available that provides an insight into this lack of involvement of surrogates.  Perhaps it is a desire on the part of the patient to maintain control or perhaps a hesitation on the part of the potential surrogate to involve himself in circumstances that could become tense or conflictual.

Family and friends are generally the best choice for substitute agents, but in their absence, local support groups often have trained advocates to assist in this capacity.  Adherence to a patient’s treatment wishes is more likely when a PAD is made available to care-givers and it is more likely to be made available when a surrogate is involved (Srebnik, 2007).

Discussion

            The use of advance directives and the underlying premise of the Hargrave v. Vermont case are about the rights of individuals to care for themselves and to exercise free will in the direction of that process.  As noted in the introduction, advance care directives are the efforts of individuals to take good or proper care of themselves both in living and in dying.  General health care advance directives which are more commonly known and invoked than psychiatric advance directives represent the efforts of individuals to assure that their existence and well-being will unfold according to their most deeply held beliefs about life, beyond the time when they are competent to direct this unfolding and as they approach the end of life.  Psychiatric advance directives usually take the insights and wisdom surrounding and gained from painful, often dehumanizing experiences when the individual is powerless to control the situation, applying them to promote a better quality of handling these life experiences should they recur.  And so, individuals record their desires and ask others to implement their desires for them and we have objects known as living wills, advance directives and durable powers of attorney for healthcare.

            In this paper we have returned quite frequently to the concept of autonomy or self determination as a value and a goal.  In caring for oneself physically and psychologically, there are all manner of challenges to autonomy and self-determination and there are legitimate boundaries to them – I cannot do anything I want, whenever I want without respect for the rights of others.  

In the case of Hargrave v. Vermont, Nancy Hargrave’s legitimate right to refuse medication was honored once by virtue of the determination that she was competent to decide for herself, but in a second instance, medication was forced upon her because her autonomy was overruled by her mental incompetence.  Legitimate authority of the state stepped in and attempted to choose her best interests for her – we would hope with great respect for the dignity of her person and only to protect her from her own potentially poor choices in her state of diminished capacity.

Subsequent to the mental health hospitalizations, Nancy executes a DPOA in which she refuses all anti-psychotic, neuroleptic, psychotropic or psychoactive medications as well as electroconvulsive therapy in her future mental health care and she appoints a guardian to act on her behalf.  She then legally challenges a recent law that would add a third mechanism to two existing provisions for overriding DPOAs.  The court ruled in her favor thus restoring a more restrictive protocol for superseding a person’s advance directive in Vermont, one of the most restrictive in the country.  The decision of the state of Vermont thus establishes strict standards, respectful of the dignity and autonomy of the individual, for supporting the right of self-determination in general and mental health care.  It would seem that the existing statutes are sufficiently broad to protect both the individual and society in general from harm that could occur.  Additionally, the statute in question, Act 114, was directed towards a rather small, restricted class, which made it discriminatory to those in the class – those who were committed by virtue of mental illness.

Protecting the right to make one’s own health care decisions is a form of self-preservation, a natural inclination of man.  The boundaries towards which this right must be held responsible are harm to self and harm to others.  In some of the subsequent material in this paper, the individual’s autonomy and self determination seem to be in jeopardy.  In the survey of the North Carolina physicians, it is alarming to me that those surveyed were split almost evenly on whether to follow the directives of a PAD to refuse medication and hospitalization for a non-violent psychotic patient in the emergency room.

The support for the completion of PADs among mental health consumers in a recent survey – 66% to 77% – is strong. PADs are promoted as a means to preserve autonomy and self-determination in the event of mental health incapacity.  These directives have value in their composition as well as in their application.  Even if the owner of a PAD never had to invoke this document, the very process of creating one has advantages; these would include collaboration with clinicians in determining future care choices and involvement of family in care evaluation and planning to name two. 

While general health care directives are often created with the help of an attorney, psychiatric advance directives are very often completed in collaboration with health care clinicians and often with family members as well.  The participation of all these parties towards the responsible treatment planning for the subject of the PAD promotes the individual’s autonomy and empowerment in managing treatment to progressing towards recovery.  If mental health consumers thought that mental health providers were willing to set aside the provisions of the advance directives lightly – as there is evidence that some do – perhaps this valuable tool would receive less support.

Both Brock and Swanson, McCrary et al. provide a clear and instructive analysis of situations in which advance directives might be “trumped.”  This type of thorough analysis is what would be hoped to ensue in any consideration of the override of advance directives.  Finally, Davis’ analysis and conclusions in the consideration of  Ulysses Contracts is an excellent example of seeking the best interest of the patient when confronted with an ethically challenging situation that might call for setting aside an individual’s advance directives to actually promote not only his/her best interests but his/her prospective desires.

Conclusion