Chapter 5. Legal Issues in Critical Care

• In order to be considered legally effective, consent to medical treatment must meet three tests: (1) it must be voluntary; (2) it must be adequately informed or knowing; and (3) it is given by an individual with adequate mental capacity and legal authority.
• Whereas a variety of state laws in the United States pertain to decisional capacity and surrogate decision making, in most cases the physician relies on family members as the surrogate decision makers for an incapacitated patient, even in the absence of a specific statute, advance directive, or court order empowering the family.
• An increasing percentage of patients, especially as the population ages, lack available, willing relatives or friends to act as surrogate decision makers. In such cases physicians should seek guidance from living wills or other forms of advance directives, when these exist.
• Neither a patient nor his or her family has a legal right to, nor does the physician have the obligation to provide, medical treatment that would be futile—that is, nonbeneficial—for that patient.
• As directors of critical care units, critical care physicians should participate in creating institutional protocols that delineate operational principles of the critical care team and the individual physician's supervisory responsibilities.

The making, implementation, and documentation of treatment decisions in the practice of critical care medicine raise a host of potential legal implications. This chapter briefly outlines some of the more salient issues and suggests avenues for their management and further exploration. Figure 5-1 depicts these issues as they arise within and impact upon the physician–patient relationship in ICUs.

It must be noted that what is described here are the evolving legal ramifications particular to critical care practice in the United States. For comparison, in critical care as practiced outside the United States, legal issues may impact much less on day-to-day decision making. It may be argued that the U.S. approach encourages greater sensitivity to and protection of the rights of patients. In contrast, the legal systems of many other countries, explicitly or by default, have elevated the role of the critical care physician as unquestioned advocate for the critically ill patient. A tangible outcome of this difference is that intensivists in the United States usually are more cautious in giving medical advice—especially advice with difficult ethical connotations—to their patients or surrogates compared to Australasian, European, or even Canadian intensivists, all of whom assume a more commanding presence at the physician–patient interface regarding critical care decision making. It is not uncommon to encounter U.S. critical care decision making suspended, awaiting the input of family or an institutional ethics committee that has yet to be apprised of the patient's condition or need for a decision; in contrast, quite similar situations are decided autocratically by the attending intensivist in other countries. Because critical care is evolving rapidly throughout the world, the contributions of the legal system—positive and negative—to health care and to society as a whole are not yet clear. It is clear, however, that legal issues become less dominating and intrusive, and ethical concerns become more central, when intensivists and their critical care teams initiate and continue sensitive, complete, consistent, and honest communication with their critically ill patients and significant others.

See ref. 1. The well-established legal doctrine of informed consent is based on the ethical principles of autonomy (personal self-determination) and beneficence (doing good for the patient).2 Early lawsuits growing out of medical interventions conducted in the absence of informed consent were predicated on a battery theory (i.e., touching the person without their consent). But most such cases today are framed as negligence actions, in which the alleged unintentional wrong is the physician's violation of the fiduciary or trust duty to inform the patient adequately as part of the permission process.

In order to be considered legally effective, consent to medical treatment must meet three tests (in the absence of an exception such as an unforeseeable emergency). First, consent must be voluntary—that is, not coerced—in nature. The patient (or surrogate decision maker) must retain the ultimate power to accept or reject the available interventions.

Second, consent must be adequately informed or knowing. A slight majority of U.S. jurisdictions still enforce a physician-oriented standard of information disclosure, inquiring whether the amount of information shared by the physician with the patient or surrogate was consistent with the usual practice of other prudent, competent physicians in similar circumstances. Almost half the states, however, have adopted a more patient-oriented standard, requiring physicians to disclose all the information that a reasonable patient would want to know under the circumstances. This latter approach is also termed the materiality standard, since it compels the disclosure of information that would be material—that is, information that might affect the decision making—for a reasonable patient.

Under either a physician- or patient-oriented approach, several particular kinds of information need to be disclosed in understandable lay language. These include: the nature of the patient's medical problem; prognosis with and without treatment; nature of the suggested intervention; likely benefits of treatment; reasonable alternatives; and foreseeable risks of the various alternatives. Additionally, many people today are interested in knowing the probable financial ramifications of their medical decisions.

Third, informed consent is sufficient only when it is given by an individual with adequate mental capacity and legal authority. This element of consent may be especially problematic in critical care and is discussed in the next section.

A note of caution regarding the legal value of written consent forms must be added here. Contrary to the usual mistake that health care providers make of treating the two things as synonymous, a written consent form is not the equivalent of legally effective informed consent. Informed consent is the process of mutual communication and ultimate patient or surrogate choice3 as described earlier. The written form is only tangible documentation or evidence that the communication process occurred. This evidence may be very important to the physician in defending against a lawsuit claiming medical intervention in the absence of informed consent, by creating a legal presumption that consent was properly obtained. However, where inadequate communication actually took place, the presumption created by the consent form may be successfully rebutted or overcome by the plaintiff. While broad, blanket consent forms used in many hospitals are sufficient to authorize routine noninvasive medical treatments, more specifically tailored and informative forms are preferable for interventions that are nonroutine or invasive, and certainly for treatments that might be characterized as experimental or innovative.4 One recent study4a of informed consent in critically ill patients suggested that early identification of patients at risk for needing invasive procedures facilitates an earlier discussion of risk-benefit relationships, fewer procedures performed as “emergent," and a possible reduction in surrogate consent.

Application of these general principles is somewhat complicated in critical care by the fact that in the ICU, interventions can be nearly continuous for the most unstable patients. Informed consent is necessary any time the doctor proposes doing anything to a patient (i.e., any “intervention”). For quasi-continuous interventions, informed consent must certainly be obtained at the initiation of the intervention, and the patient or surrogate must be afforded the opportunity to withdraw that consent at any later point.

Most interventions in critical care may be categorized as risky or invasive, that is, an intervention involving a higher degree of risk or intrusion for the patient than he or she would ordinarily face in normal, everyday life. However, this categorization is relevant only to the question of the extent of advisable documentation; the requirement of an informed consent process itself applies regardless of an intervention's level of risk or invasiveness.

Minor children (defined in almost every state for medical purposes as those younger than 18 years old) are presumed legally to be incapable of making their own medical decisions, and in the absence of an exception based on the minor's “emancipated” or “mature” status, the natural parent or court-appointed guardian is legally empowered as the medical decision maker. Adults, however, are presumed legally to be decisionally capable.5 In critical care, many (although certainly not all) patients will be so ill and debilitated that they are not presently capable of making and expressing treatment choices resulting from a rational thought process of digesting and weighing information about the benefits and risks of different alternatives.6

Although in theory the same degree of mental capacity is necessary whether the patient is accepting or declining the recommended medical intervention, in practice, a formal inquiry into patient capacity usually occurs only in the case of patient refusal. Ideally, a judgment about patient decisional capacity, which may fluctuate widely over time because of a variety of natural and iatrogenic factors, is being made at least implicitly by the attending physician each time the patient is seen. Such an inquiry ought to be inherently built into every physician–patient encounter. Where decisional capacity is seriously questioned, a more focused examination needs to be conducted. In especially ambiguous cases, documentation in the patient's medical record by consultants who have evaluated the patient's decisional capacity is good risk management for the physician proposing an intervention, who is the one ultimately responsible for ensuring informed consent; the legal system affords psychiatrists a great deal (arguably an excessive amount) of deference as consultant capacity assessors.

For patients who in fact lack the ability to engage in a rational decision making process, the presumption of capacity is rebutted or overcome. This does not, however, mean that the physician may dispense with informed consent prior to initiating particular interventions. Consent remains necessary for treatment of incapacitated patients, but the consent must be obtained from a surrogate or proxy acting on the patient's behalf. Several approaches to surrogate medical decision making have developed.

Over half the states in the past three decades have enacted “family consent statutes,” which specify relatives (ordinarily in a priority order) who may legally make medical decisions for an incapacitated family member.7 Advance proxy directives (see next section), especially the durable power of attorney, may be used by currently capable persons to designate their own surrogates in the event of subsequent incapacity. A formal guardianship or conservatorship (precise terminology varies among jurisdictions) proceeding may be initiated, in which a probate court finds the patient (the ward) to be decisionally incapacitated (the legal term usually employed is incompetent) and appoints someone else (the guardian or conservator of the person) as the surrogate decision maker.

In most cases, however, the physician relies on family members as the surrogate decision makers for an incapacitated patient, even in the absence of a specific statute, advance directive, or court order empowering the family. This informal process of making do using next of kin, even without explicit legal authority, works well in the vast majority of situations in which the family members agree on a course of conduct both among themselves and with the physician, and where they appear to act consistently with the patient's values and preferences (the “substituted judgment” standard) or with the patient's best interests.

See refs. 8, 9. Some patients retain a sufficient degree of capacity to make and express their own medical decisions even after admission to a critical care unit (CCU). Legal precedent is very clear that, except in cases in which the welfare of a third party such as a minor dependent is jeopardized, a properly informed, capable patient has the right to make personal medical decisions, including a decision to refuse even life-prolonging treatment.

If the patient is decisionally incapacitated, a more difficult scenario may confront the physician. A surrogate decision maker may be identified (see previous section) under a state family consent statute, the patient's prior execution of a durable power of attorney (see Appendix 1 for a model) naming a surrogate or health care agent, a court appointment of a guardian/conservator, or informally relying on available, willing family members.

The surrogate is legally expected in most states to make decisions consistent with what the patient would choose if he or she were presently able to make and express choices personally (substituted judgment standard, or “donning the mental mantle” of the incapacitated person). If there is no reliable indication (some states, e.g., Missouri10 and New York,11 require proof by “clear and convincing evidence”) of the patient's preference under the circumstances, the surrogate must act in the patient's best interests, considering—from the patient's perspective—the proportionality of likely benefits and burdens associated with available medical alternatives.

An increasing percentage of patients, especially as the population ages, lack available, willing relatives or friends to act as surrogate decision makers.12 In such cases, there are several potential sources of guidance for the physician.

The patient may have executed a living will document while still decisionally capable. This type of advance directive, which is authorized by statute (frequently called a natural death act) in almost every jurisdiction, permits a capable adult to make a record of personal preferences regarding future medical treatment in the event of subsequent incapacity and critical illness (see Appendix 2 for a model). Although usually this directive is used to indicate a preference for the limitation of future medical intervention, the directive could be employed to request maximum medical intervention if there is any likelihood of benefit to the patient. However, neither a patient nor his or her family have a legal right to, nor does a physician have the obligation to provide, medical treatment that would be futile—that is, nonbeneficial—for that patient.13,14

A physician who complies in good faith with the patient's expressed voluntary, informed wishes to limit life-prolonging treatment is on firm legal ground.15 For the physician who for ethical reasons chooses not to follow a living will's directive to limit treatment, there is an ethical and legal obligation to notify the patient or surrogate (if one is present), and to make a reasonable attempt to transfer the patient to another physician who is willing to comply with the living will, or at least not to prevent such a transfer.

In many states, a public guardianship system has been created to make surrogate decision makers available, either through a government agency or through a private agency under government contract, for incapacitated patients who lack available, willing relatives or friends.16 There are also demonstration projects in some locales that use privately funded agencies to act as surrogate medical decision makers for incapacitated patients.17 The physician should consult hospital legal counsel to determine acceptable sources of surrogate decision making for the incapacitated patient without relatives or friends.

The most controversial and complicated issue in the treatment limitation arena is the status of artificial feeding and hydration.18 The courts have been unanimous in holding that feeding tubes are merely another form of medical intervention, which could be withheld or withdrawn under the same circumstances that would apply to the withholding or withdrawal of any other type of medical intervention.19 The American Medical Association20 and other groups21 endorse this analysis. However, some argue that feeding and hydration, even where accomplished through tubes surgically or forcibly inserted in the patient's body, are fundamentally different and more elemental than medical treatment, and therefore ought to be continued as long as they can keep the patient alive.22 A number of state legislatures have embodied this argument in living will or durable power of attorney statutes that purport to severely restrict the ability of surrogate decision makers to authorize the removal of feeding tubes. Both the wisdom and the constitutionality of these purported restrictions are extremely questionable.23,24

The case of Nancy Cruzan remains the only case decided by the U.S. Supreme Court, on June 25, 1990, that sheds light on the issue of discontinuing life-prolonging treatment. Cruzan was an automobile accident victim who was kept alive in a permanent vegetative state within a state-operated (Missouri) long-term care facility, through the use of feeding and hydration tubes. Her parents asked that this intervention be discontinued, a request they claimed was consistent with the patient's previously expressed (although not documented) wishes. The physicians refused this request, and the Missouri Supreme Court upheld the lower court decision and refused the parents' request to discontinue treatment.

The U.S. Supreme Court held that a mentally capable person has a fundamental constitutional right, under the liberty provision of the Fourteenth Amendment's due process clause, to make personal medical decisions, even regarding life-prolonging treatments, including artificial feeding and hydration. For incompetent patients, though, the court ruled that a state's interest in preserving life was strong enough to permit a state, if the state so chooses, to require “clear and convincing” evidence of the treatment preference the patient would hold if currently able to make and express an autonomous choice, before the state must comply with a surrogate's instructions to withdraw life-prolonging medical care. Presumably, a written declaration made while the patient was competent would suffice as evidence of treatment preference in the event of subsequent incapacity. Under the Cruzan decision, states are also free to set lower standards of proof than “clear and convincing” evidence for incapacitated patients, such as proof by a preponderance of the evidence (i.e., greater than 51% likelihood). In the Cruzan case itself, Missouri by statute imposed the “clear and convincing” evidentiary standard, and since Cruzan had never clearly documented her treatment wishes while still competent, the proof was inconclusive, and the state was permitted to err on the side of continued treatment.

One form of treatment limitation around which there is a high degree of current consensus is the do-not-resuscitate (DNR) or no code order, which instructs caregivers to refrain from initiating cardiopulmonary resuscitation (CPR) for a patient who suffers an anticipated cardiac arrest. There have been very few legal cases in this arena, but the well-accepted rule is that a decisionally capable patient has the right to refuse CPR, and that surrogates may elect to forego CPR for a patient if the likely burdens of this intervention to the patient would be disproportionate to any benefits (e.g., mere continued existence until the next arrest) that might be derived.25 As is true for all medical decisions, a DNR order should be made only after a thorough consultation with the patient or surrogates and should be clearly documented in the medical record.

Where the patient or surrogate has declined aggressive, technologically oriented interventions, the physician still retains the legal obligation to continue basic palliative (comfort and pain control) and hygiene measures. Failure to do so could constitute negligence or form the basis for professional disciplinary action.26

The other side of the coin on treatment decision making is presented when the patient, or more usually the family, insists on initiation or continuation of medical treatment (“doing everything possible”) that the clinician judges to be futile in terms of benefit to the patient.27 On the very rare occasions that courts have been involved prospectively with the futility issue, their holdings have been confusing, inconsistent, and often poorly reasoned.28 No court has ever imposed liability for failure to begin or perpetuate futile interventions for a critically ill patient, even in the face of family insistence on doing everything possible. In practice, clinicians usually seem to take the path of least resistance in such circumstances and “treat the family,” often out of misapprehension of potential liability exposure. In the vast majority of cases, better physician–family communication, in which the realistic implications of “doing everything possible” are spelled out clearly, can obviate serious disagreement over how to proceed.29

Most hospitals have adopted written policies and procedures concerning patient admission to, retention in, and discharge from CCUs. The ability of physicians to consult institutional guidelines generally leads to better, more consistent decisions that are easier to defend against later claims of impropriety.30 Clear protocols facilitate communication and cooperation among members of the health care team, decreasing both inadvertent mistakes and interpersonal tension. Institutional protocols are also essential as inevitable discussions of health care rationing take on increasing urgency.31

The development and dissemination of institutional protocols regarding critical care, especially DNR orders, is required by the federal Patient Self-Determination Act (PSDA)32 and by some states.33 It is also required for approval by the Joint Commission on Accreditation of Healthcare Organizations (JCAHO).34

Critical care physicians must be completely familiar with their own institutions' formal policies and procedures, and must assure familiarity with them on the part of nurses and other staff members. Ideally, members of the medical staff should contribute to the drafting, continuing re-evaluation, and revision of institutional protocols.35 Questions regarding the meaning or implementation of these protocols should be addressed in a timely fashion (i.e., before a crisis erupts) to hospital legal counsel.

Similarly, the physician must be knowledgeable about the operation of his or her hospital's institutional ethics committee (IEC). The past three decades have seen a proliferation within health care institutions of entities designed to provide education, formulate policies and procedures, and offer advice regarding particular cases and issues with serious bioethical implications.36 JCAHO standards require that hospitals have in place a mechanism for carrying out these functions.37 Although the emphasis of IECs is, and ought to be, on better ethical decision making, salutary legal by-products may also result from their activities. Effective use of an IEC may preclude resorting to the judicial system, initiated either by the family or by a health care team member who feels his or her opinion has been inadequately taken into account, for definitive resolution of a case. Furthermore, in the relatively unlikely event of the informal decision making process breaking down and court involvement being invoked, utilization of an IEC may act as powerful evidence of the provider's good faith and appropriate concern for patient autonomy and welfare.38

One inescapable aspect of critical care medicine with important legal implications is the determination and declaration of when a patient has died. Traditional standards of death based on cessation of cardiorespiratory functioning are no longer sufficient by themselves in light of modern medical technology that frequently can maintain the human organism almost indefinitely. Questions relating to the discontinuation of medical intervention and the harvesting of organs for transplantation have demanded new approaches to the legal definition of death.

In response to these questions, almost all states have adopted, either by statute or court decision, a version of “brain death.” This definition provides, either as an alternative to or a replacement for the traditional heart-lungs approach, that a person is legally dead when there is irreversible cessation of all (including brain stem) brain function.39 Brain death should be confirmed clinically according to the Harvard criteria,40 as they have been periodically updated.41,42 Once a patient has been declared dead, there are no longer any treatment decisions to be made (although autopsy and organ donation issues may remain). There is neither a legal duty nor a right to continue medical intervention on a patient who has become a corpse.43

Critical care medicine is an interdisciplinary team enterprise, and the manner in which members of the team relate to each other and to the patient and family carries legal consequences. Under the old doctrine of “captain of the ship,” a physician in charge of a critical care unit automatically was held legally responsible for any negligently caused patient injury occurring in the unit, regardless of the physician's personal ignorance of or lack of involvement with the particular error or omission. The captain-of-the-ship doctrine has been abandoned by the courts over the past several years in recognition of the increasing complexity of health care delivery.

However, a physician still may be held responsible, under a vicarious liability rationale, for patient injuries proximately (directly) caused by negligent errors or omissions committed by nurses or other providers over whom the physician has supervisory power. The key inquiry in potential vicarious liability situations is not whether the physician actually was exercising supervisory power at the time of the supervisee's negligence, but whether he or she had the authority and opportunity to supervise properly if he or she had chosen to exercise that power.44

Thus, the vicarious liability doctrine has significant legal ramifications for the interdisciplinary team's conduct in rendering critical care. The physician and other team members must understand their legal relationships to each other, and their implications with regard to assignment of tasks, oversight, reporting, communication, and problem resolution. The physician must take their obligations as legal leader of the team seriously, without acting autocratically and thereby negating the benefits of broad interdisciplinary contributions to patient care.

Institutional protocols should delineate operational principles of the team and the individual physician's supervisory responsibilities. Where there are multiple consultants on a particular case (as is the norm), medical staff bylaws must spell out the continuing coordination and monitoring obligations of an identified attending or primary care physician; failure to do so unambiguously will increase the liability exposure of all involved clinicians and the hospital in the event of a bad clinical outcome. Consultants who are not hospital employees must be credentialed to practice within the hospital according to criteria contained in the bylaws. Hospital policies and procedures must designate their ICUs as either “closed” (in which case the patient is transferred to an intensivist who functions as the primary care physician) or “open” (in which case the original primary care physician retains ultimate authority and responsibility, but is permitted or required to consult with a critical care physician on the hospital's staff).

Similarly, methods of triaging patients into and out of the ICU should be delineated within the hospital's written policies and procedures. Included should be a specification of ultimate responsibility for a patient's admission or discharge.

As a general principle, when there is a question concerning allocation of responsibility for decisions or actions, development of an institutional policy may be advisable. Courts ordinarily grant hospitals broad leeway in the development and enforcement of the sort of institutional protocols discussed in this chapter, as long as policies and procedures appear to ensure patient care within currently acceptable medical standards. As noted earlier, accreditation standards of the JCAHO also set permissible parameters for internal institutional protocols.

For physicians who function as clinical instructors in teaching programs, residents and medical students may expose the attending physician to vicarious liability for negligent acts or omissions done in the course of the educational activity. The exercise of due care in the monitoring, supervision, task assignment, and evaluation of residents and students who are supposed to be under the physician's supervision cannot be overlooked.45

The courts have held that creating and maintaining accurate records of patient care is an integral part of the duty that a health care provider owes to a patient. Good documentation is imperative to providing competent patient care, and since avoiding unexpected bad outcomes is the best legal prophylaxis, is therefore wise legal practice. Furthermore, in the event of accusations of substandard care, the physician's best (and often only) defense will lie in the quality of documentation created to explain and justify decisions made and actions taken. In addition, institutional accreditation and third-party reimbursement turn heavily on information drawn from medical records.

The quality of medical records is especially important in critical care, where patient conditions are subject to rapid change, many different professionals may be involved in treating the patient, cost considerations are ever present,46 and decisions (such as limiting the application of life-prolonging technology) may be controversial. The watchwords of documentation are the same from the legal and medical perspectives: completeness, legibility (dictation is preferable, but dictated entries should be read and corrected before being signed), accuracy or truthfulness, timeliness, corrections should be made in a clear and unambiguous fashion, and objectivity.

A corollary to the subject of medical record keeping is confidentiality. In light of common law privacy principles and federal regulations effective in 2003 to implement the Health Insurance Portability and Accountability Act (HIPAA),47 the physician must guard against unauthorized disclosures of personal information about the patient. The party who has authority to give or refuse consent for medical treatment (i.e., the patient or her or his surrogate) ordinarily controls the release of identifiable medical information to third parties, absent a court order or countervailing government regulation. All questions about the release of medical information to third parties in specific cases should be directed to the institution's medical records department or legal counsel.

The counterpart to the right of the patient or surrogate to control the release of information to others is the patient's own right of access to the information contained in the medical record. This right of access is guaranteed, at least for in-hospital care, by the federal Privacy Act for federal facilities, and by the HIPAA regulations, state statutes, and JCAHO standards in most private and public institutions. Patients or their surrogates request access to records for a variety of reasons, ranging from curiosity to serious questioning of quality of care. A physician who is informed of a patient's or surrogate's request for access to records should offer to go through the record with the patient or surrogate, explain matters, and answer questions. In short, treating this circumstance as an opportunity to bolster or correct existing but perhaps unnoticed deficiencies in communication between the physician and patient or surrogate, rather than as a personal affront calling for defensive posturing, often can pay risk management dividends by preventing at an early stage potential misunderstandings that would otherwise eventually manifest themselves as legal action.

The hospital's overall program to identify, mitigate, and avoid potential problems that could result in legal, and therefore financial, loss to the institution (i.e., the risk management program), should incorporate specific activities designed to address legal risks prevalent in the delivery of critical care.48 Particular areas of attention in a critical care–sensitive risk management program should include the organization and administration of CCUs, the roles and responsibilities of the different professionals having contact with patients in those units, medical records, equipment maintenance, equipment modification, equipment records, analysis of equipment malfunctions, incident reporting, and trend analysis of unexpected incidents.

The physician should be knowledgeable about the institution's risk management program and cooperate with it to ensure appropriate sensitivity to critical care practices and potential problems and their avoidance or mitigation. The physician should view the risk manager as a partner in pursuit of the common goal of providing, and if necessary proving after the fact, quality patient care.

Perhaps the single most influential aspect of effective risk management is the fostering of a positive relationship between the critical care team, led by the physician, and the patient and his or her family. There is a demonstrated correlation between patient (or family) psychologic satisfaction with the quality of the physician–patient (or –family) relationship, and the propensity to file a lawsuit if a bad outcome occurs.49 Communicating openly and compassionately, especially acknowledging the vast uncertainty that pervades much of critical care medicine, is as important a tool in the malpractice prophylaxis arsenal as being proficient, up to date, and conscientious in knowing and applying technologic information and skills.50