Section 2: The Core Obligations of Treatment Providers

The Provider's Fiduciary Responsibility to Clients

There are several core ethical principles that govern the relationship between treatment provider and client. The first is that the practitioner is a fiduciary, whose primary obligation is to act in the client's best interest. This principle is embedded in the ethical standards of the behavioral health care professions. For example, the Code of Ethics of the National Association of Social Workers notes that "social workers' primary responsibility is to promote the well-being of clients" (National Association of Social Workers 1996, Code of Ethics, Standard 1.01).

This relationship creates a responsibility for the provider to refrain from certain types of actions, for example, those that exploit the client or do harm to the client. The fiduciary responsibility to act in the client's best interest also creates affirmative obligations. The practitioner, acting within the scope of his or her competence, must seek to provide the treatment he or she believes is best suited to the client's needs. This obligation exists even if the economic interests of the treatment provider are at odds with the interests of the client. The Code of Ethics of the American Medical Association reminds physicians that "[a] physician has a duty to do all that he or she can for the benefit of the individual patient" but that if the "economic interests of the hospital are in conflict with patient welfare, [then] patient welfare takes priority."

The fiduciary obligation to the client is not absolute. The interests of third parties may take precedence in some situations. For example, if the practitioner suspects that his or her client has abused a child, that suspicion must be reported, despite the principle of confidentiality. If the practitioner concludes that the client presents a real harm to third parties, many States require or permit the practitioner to take steps to protect the third party (and the ethical principles of each of the professions contemplates such an exception to confidentiality as well). However, in such cases, society has deemed the interests of a third party to have greater importance than those of the client; the interests of the treatment provider rarely, if ever, take precedence over those of the client, and never do if the interest at stake is the economic interest of the provider.

The importance of autonomy

Dr. Lewis, treating Mr. Darwin for depression, prescribed an antidepressant and suggested that he come for an hour of therapy once a week. Dr. Lewis also gave Mr. Darwin her home phone number and pager number and told him on several occasions to call her if he was depressed, anxious, or simply needed someone to talk with. Mr. Darwin, feeling much better after three weeks of this regimen, quit taking his medication and canceled two appointments with Dr. Lewis. Shortly after Thanksgiving, he attempted to kill himself. He suffered permanent brain damage in the attempt, and later sued Dr. Lewis for negligent treatment. Dr. Lewis wants to know whether Mr. Darwin's failure to follow her treatment recommendations will absolve her of all or partial liability in the case.

In this vignette, Dr. Lewis has prescribed the care she believes to be in her patient's interest. However, the patient has not acted on her recommendations, and then has attempted suicide. In such a case, the patient may claim that Dr. Lewis should have prescribed different treatment, more or less of it, or taken steps to get him back into treatment. However, Mr. Darwin's failure to follow treatment recommendations made by a competent treatment provider probably will reduce significantly the amount of damages he might win in a malpractice case, and in the vignette presented, would greatly reduce the prospect of any recovery, particularly given that Dr. Lewis, in addition to the treatment she proffered, gave the patient both her phone number and pager number and told him to call her anytime he wished to talk.

At the same time, Mr. Darwin had the right to reject Dr. Lewis's advice, because the ethical and legal principle of autonomy holds that competent individuals as a general rule have control over health care decisions. The treatment provider may suggest, may attempt to persuade, may cajole the patient in an effort to ensure that the patient follows his or her advice. However, as long as the patient is competent, except in certain situations where the law permits the use of coercion, the treatment provider may not force the client to act on the treatment recommendations.

One of the first recognitions of autonomy as a legal principle occurred in Schloendorf v. Society of New York Hospital (105 N.E. 92 (N.Y.1914)). New York's highest court wrote that "[e]very human being of adult years and sound mind has a right to determine what shall be done with his body." Historically, the right of people with mental illness to exercise autonomy in decisionmaking has been limited because of a categoric assumption that mental illness robbed individuals of the capacity or competence to make decisions. However, in the last 30 years that categoric assumption has eroded, and people with mental illnesses generally are assumed until proven otherwise to have the same right to autonomy in health care decisionmaking as that enjoyed by other citizens.

While the principle of autonomy has been extended to people with mental illness, the law still permits the use of coercion in at least two situations. First, all States permit involuntary civil commitment, based on a combination of mental disorder and behavior (typically defined as either danger to self or others or an inability to care for oneself). While commitment on the basis of mental illness and behavior is permitted in each State, a 1996 review found that only 31 States and the District of Columbia permitted civil commitment on the basis of alcohol or substance abuse. In addition, the commitment of children in many States may be accomplished through parental rather than judicial decision.

Coercion also is permitted in some situations to administer treatment over the objection of an individual, most commonly in cases where the person does not wish to take psychotropic medication. Medication may be administered without a court hearing in an emergency (usually defined as a situation where harm to self or others is imminent without the use of medication). In other situations, most States require a judicial hearing into the person's competency; individuals who are competent and refuse medication have the right to do so, even after civil commitment.

Informed consent as a
mediating principle

Informed consent has both legal and ethical roots. It assumes that individuals cannot exercise autonomy and choice in health care decisionmaking unless they have adequate information and can make a reasoned choice. There are three elements to informed consent: It must be voluntary, it must be competent, and it must be knowledgeable or informed.

Informed consent must be voluntary. Legal consent cannot be the product of deceit or duress. For example, a contract entered into under duress will not be enforced by the law. Similarly, a health care decision made under duress may invalidate the patient's consent. As the discussion below suggests, the question of "voluntariness" may take on new importance in some types of managed care settings.

Informed consent must be competent. There has been significant controversy regarding the question of competency and how it should be measured. Behavioral health care providers sometimes are accused of judging a client competent as long as the client adheres to treatment recommendations, while finding the client incompetent when the client declines treatment. Practitioners often express frustration with the law's insistence that individuals, even an individual who has just been civilly committed, are considered competent by the law for the purposes of making treatment decisions until specifically found to lack competence.

There is also not always consensus on what "competence" means. Roth and his colleagues, in a review, found five tests that have been suggested or employed:

1. Expression of a choice. Under this test, if the client expresses any choice about the treatment decision, he or she is judged competent. This test has found few adherents (and none among courts) because it ignores completely the client's underlying mental status and processes.
2. The treatment decision is rational. The client is considered competent if the decision he or she reaches is the decision that would be reached by a "reasonable" client. This test had significant support historically, but in recent years has lost favor because it makes the exercise of autonomy subject to a definition of reasonableness imposed by others.
3. The process by which the decision was reached was unimpeded by mental or emotional illnesses or other factors, for example, substance abuse. This test, in common use today, examines the underlying processes by which the client reached his or her decision. It tolerates choices that others might consider irrational, so long as the client's decisionmaking processes were not affected by illness or mental disorder and so long as the patient had "rational reasons" for the decision. This test is biased toward treatment adherence, because whether the patient had "rational reasons" for his or her decision may turn on the treatment provider's views on the wisdom of the patient's ultimate decision. The decision not to accept treatment may be viewed less favorably than the decision to accept it.
4. The client had the ability to understand the choices he or she had and the consequences of the choice he or she made. This test, in the view of Roth and his colleagues, is most consistent with the law of informed consent. It examines whether the patient had sufficient ability to comprehend those things deemed important by the law in making decisions about treatment. It does not concern itself with whether the patient then reaches a decision that is different from the decision that the treatment provider or another person might make.
5. The client had actual understanding. This may be the most difficult test to assess, because it requires something more than assessing the client's intellect and cognitive abilities. It requires an assessment as to whether the patient truly understood information conveyed, choices to be made, etc. However, because it is difficult to measure what constitutes actual understanding, this test is rarely, if ever, applied in practice.

Competency is also a fluid concept, both substantively and temporally. As a general rule, an individual must be "more" competent the more intrusive or consequential the treatment proposed. For example, more competency is required to consent to electroconvulsive therapy than to take aspirin. Also, an individual may be competent one day, but incompetent the next, and then competent again the third, depending on the type of illness and its effects. Because of this, there has been a movement in the recent past to encourage client choice in anticipation of incompetency, for example, through living wills.

Finally, recent research done under the auspice of the MacArthur Foundation's Research Network on Mental Health and the Law has resulted in the development and testing of a number of new instruments for assessing competency to make treatment decisions (Grisso et al. 1995a, 1995b).

Informed consent must be knowledgeable. An individual must have adequate knowledge to give informed consent. As with competency, the definition of what constitutes "adequate knowledge" has changed as deference to the principle of autonomy has increased. As a general rule, a client must be informed of the treatment proposed, its risks and benefits, the risks and benefits of forgoing treatment, and the risks and benefits of alternative treatments. In the past, a treatment provider had broad discretion to withhold information if the provider believed that disclosure might make it unlikely for the client to undergo needed treatment. While many State laws continue to recognize some variation of this principle, most courts now ask whether a person in the patient's position would want the information. This perspective expands the amount of information made available, because courts generally assume a patient would want all information pertinent to the decision, even if a risk might be remote or the information might unduly heighten the patient's anxiety if revealed.

If the client is unable to give consent, because of incompetency or age, then consent if not presumed must be given by another person.

Situations when consent is presumed

Consent will be presumed in an emergency when the client is unable to give consent. For example, if a patient is presented to an emergency room unconscious after an automobile accident, treatment staff has the legal authority to assume that the patient would consent to the interventions necessary to address the emergency. However, consent is presumed only for the duration and extent of the emergency; in the course of providing emergency surgery, the physician cannot without some legal risk expand the scope of the surgery to address a problem unrelated to the emergency.

In other situations, consent may not be required if there is an immediate risk of harm to the patient or others. For example, on a psychiatric ward, medication to control behavior may be administered without patient consent if the individual's behavior creates an emergency situation on the ward threatening the safety of the patient, other patients, or staff.

Situations when substitute consent must be obtained

If the client lacks capacity to give consent in a nonemergency, then consent must be obtained from another party. All States have processes for obtaining substitute consent. A common example is guardianship law. Many States now recognize "health care proxies" as well. This means that a competent individual may designate another individual as his or her "proxy" for making health care decisions if the individual becomes incapacitated. In other situations, courts may be the designated decision maker--for example, in States requiring a judicial order before a patient's decision to refuse psychotropic medications may be overturned.

Consent by and on behalf of children is a more difficult issue. Historically, children, like people with mental illness, were assumed categorically to be incapable of providing informed consent. However, research has shown that even young children in many situations may have decisionmaking capacity equivalent to that of adults, and the age at which children are assumed legally to be incapable of making decisions has dropped. Today, most States permit individuals 16 years of age to make at least some types of health care decisions without parental consent; however, some types of decisions, for example, to admit oneself to a psychiatric hospital, may require parental approval before or shortly after the admission. Because State laws vary on the question of parental consent, the treatment provider must know the laws of the particular State in which he or she practices.

Historically, informed consent dealt with clinical, not economic, information. However, as the discussion below suggests, the emergence of managed care is causing a rethinking of what types of information should be discussed with the client in obtaining informed consent.