|Home | About | Journals | Submit | Contact Us | Français|
Drs. Williams and Sudore: San Francisco Veterans Affairs Medical Center, 4150 Clement Street, 151R, San Francisco, CA 94121.
Ms. Hooper and Dr. Weithorn: Hastings College of the Law, 200 McAllister Street, San Francisco, CA 94102.
Mr. Sabatino: American Bar Association, 740 15th Street NW, Washington, DC 20005.
Advance directive law may compromise the clinical effectiveness of advance directives.
To identify unintended legal consequences of advance directive law that may prevent patients from communicating end-of-life preferences.
Advance directive legal statutes for all 50 U.S. states and the District of Columbia and English-language searches of LexisNexis, Westlaw, and MEDLINE from 1966 to August 2010.
Two independent reviewers selected 51 advance directive statutes and 20 articles. Three independent legal reviewers selected 105 legal proceedings.
Two reviewers independently assessed data sources and used critical content analysis to determine legal barriers to the clinical effectiveness of advance directives. Disagreements were resolved by consensus.
Legal and content-related barriers included poor readability (that is, laws in all states were written above a 12th-grade reading level), health care agent or surrogate restrictions (for example, 40 states did not include same-sex or domestic partners as default surrogates), and execution requirements needed to make forms legally valid (for example, 35 states did not allow oral advance directives, and 48 states required witness signatures, a notary public, or both). Vulnerable populations most likely to be affected by these barriers included patients with limited literacy, limited English proficiency, or both who cannot read or execute advance directives; same-sex or domestic partners who may be without legally valid and trusted surrogates; and unbefriended, institutionalized, or homeless patients who may be without witnesses and suitable surrogates.
Only appellate-level legal cases were available, which may have excluded relevant cases.
Unintended negative consequences of advance directive legal restrictions may prevent all patients, and particularly vulnerable patients, from making and communicating their end-of-life wishes and having them honored. These restrictions have rendered advance directives less clinically useful. Recommendations include improving readability, allowing oral advance directives, and eliminating witness or notary requirements.
U.S. Department of Veterans Affairs and the Pfizer Foundation.
Advance directives allow people to designate a health care agent and specify health care preferences for future medical situations (1). Advance directive laws were created in response to high-profile “right-to-die” legal cases, such as those of Nancy Cruzan and Karen Ann Quinlan, to protect patients’ rights of self-determination to decline life-sustaining treatments (1, 2). However, a fundamental tension exists between advance directive law and clinical practice.
In the clinical setting, advance directives often are used in conjunction with other forms of verbal or written communication of patients’ wishes. In contrast, advance directive law takes a strict, legal-transactional approach to advance care planning that is akin to signing a will. This approach has resulted in many legal requirements and restrictions to execute an advance directive. Moreover, each U.S. state and the District of Columbia has adopted its own statutes that govern advance directives, living wills, and durable powers of attorney for health care, resulting in profound variability (3).
The attempt to safeguard a patient’s right of self-determination through a legally driven process may have unintended consequences for patients (4). Although advance directives may stimulate discussions and reduce the stress of surrogate decision making (5–8), well-documented controversy exists over their clinical effectiveness, including their inability to affect clinicians’ and families’ understanding of patients’ preferences and the type of care received (9). Without a health care agent, the absence of an advance directive may result in undertreatment or overtreatment, yet advance directives often are not completed, especially among minority and disenfranchised populations (10). Advance directive legal requirements may actively impede people from engaging in advance care planning (11). Because of ongoing efforts at both state and federal levels to improve advance directive law (12–15), the unintended consequences of the current law on the clinical effectiveness of advance directives must be considered.
We describe unintended consequences of advance directive law that may prevent patients from making or communicating their end-of-life care preferences or having their preferences honored. We also discuss vulnerable patient subgroups who may be most affected by unintended consequences of specific laws, explain the effect of advance directive law on clinicians, and suggest changes to current law to improve the clinical effectiveness of advance directives.
Our primary data sources included state statutes dedicated to advance directives, living wills, and durable power of attorney for health care in all 50 U.S. states and the District of Columbia (Appendix Table 1, available at www.annals.org) and legislative summaries from the American Bar Association through August 2010. We also searched LexisNexis and Westlaw for state and federal legal cases by using the search terms advance directives, living wills, durable power of attorney for healthcare, surrogate decision maker, and physician immunity from 1966 to August 2010. Finally, we systematically searched MEDLINE and Lexis-Nexis for English-language articles by using the search terms advance directives, durable power of attorney for healthcare, advance care planning, barriers, health disparities, physician immunity, and state advance directive law and related terms from 1966 to August 2010.
Two investigators independently identified eligible state legislative statutes among all U.S. state statutes. Using the aforementioned search terms, we also independently identified 128 articles and selected a subset of 20 that specifically addressed advance directive law. All disagreements were resolved by consensus.
Three lawyers independently selected 105 federal and state legal proceedings by using the aforementioned search terms. All disagreements were resolved by consensus. We excluded references to court-appointed guardians, agents’ authority for pregnant patients, and psychiatric directives.
We analyzed advance directive statutes and related materials for their potential to prevent patients from making or communicating end-of-life preferences. Three lawyers (including one of the authors) and 2 of the investigators also independently analyzed federal and state case law for how courts have interpreted and applied advance directive and physician immunity statutes.
Using content analysis, 2 of the investigators independently grouped legal and advance directive content-related barriers into overarching categories (16). The coding schema was revised by rereading the statutes and recoding until reviewers reached more than 95% agreement. A third investigator helped to resolve discrepancies.
Because advance directives have been shown to be written above a 12th-grade reading level, we hypothesized a priori that poor readability would be a barrier (17, 18). We then assessed how poor readability and other identified legal barriers may specifically affect disenfranchised populations. We focused a priori on socially isolated or institutionalized older persons and patients with limited literacy, limited English proficiency, or both because these groups report difficulty with medical forms and decision making (19–24). We also focused on patients with physical and intellectual disabilities, minority populations, homeless, and migratory patients because these groups perceive a high degree of discrimination (25–28).
This study was supported by the U.S. Department of Veterans Affairs and the Pfizer Foundation. The funding sources played no role in the searches, selection, data extraction, analysis, or interpretation of the findings.
We identified 5 overarching legal and content-related barriers: poor readability; health care agent restrictions; execution requirements (steps needed to make forms legally valid); inadequate reciprocity (acceptance of advance directives between states); and religious, cultural, and social inadequacies. Appendix Table 1 summarizes these barriers, related case law or statute examples, and vulnerable populations particularly affected by these barriers. Appendix Table 2 (available at www.annals.org) describes individual barriers for all 50 U.S. states and the District of Columbia. We also identified physician immunity statutes that may profoundly affect patients and their families.
Legal, precise language has been used in an attempt to minimize ambiguity. For example, some states, such as Ohio, require mandatory language to describe life-sustaining treatment and a 1700-word disclosure statement with warnings to patients (3). Other states, such as Oregon and Wisconsin, require the entire advance directive form to adhere to mandated legal language (3) (Appendix Tables 1 and and22).
Although the Institute of Medicine recommends that health-related materials be written at or below a 6th-grade reading level (19), most advance directives are written above a 12th-grade level (17, 18) and are unavailable in many patients’ native languages (26). Furthermore, many directives contain ambiguous language, such as forgoing treatment if a condition is considered terminal or irreversible. Physicians, much less patients and their families, have been shown to have difficulty deciphering the meaning of these terms (29).
Approximately 40% of the U.S. population reads at or below an 8th-grade level, and the mean reading level of older persons is a 5th-grade level (19). Being unable to read or understand advance directive forms threatens patients’ ability to understand their health care choices and may limit their ability to communicate their preferences. In fact, limited health literacy explains some of the racial or ethnic variability demonstrated in end-of-life preferences (30).
In an attempt to prevent coercion, many states restrict who may serve as a health care agent (Appendix Tables 1 and and2),2), including primary clinicians, caseworkers, and persons working for the patient’s clinician or the care facility in which the patient resides. These restrictions occur in 37 states and the District of Columbia, but most states allow exceptions for immediate relatives.
Furthermore, if a durable power of attorney for health care is not executed, most states also restrict who may serve as a default surrogate. Many states typically authorize next of kin in the following order of priority: spouse; adult children; parents; siblings; sometimes, the nearest living relative; and last, in about 20 states, a “close friend” (Appendix Tables 1 and and22).
The decision-making status of a patient’s domestic or same-sex partner is more complicated. Only 8 states (Arizona, Maine, Maryland, New York, New Mexico, Nevada, Oregon, Wisconsin, and the District of Columbia) recognize domestic partners in their health decisions law (31). However, same-sex partners may be recognized as valid default surrogates through marriage (in California, Connecticut, Iowa, Massachusetts, New Hampshire, Vermont, and the District of Columbia) or explicit authorization in state statutes (in Arizona, Maryland, and New York). In other states, a domestic or same-sex partner may be considered a “close friend” at the end of the default priority list.
In most states, designated or default surrogates also have limited authority to consent to withdrawal or withhold life-sustaining treatments unless the patient is deemed to be in a terminal or persistent vegetative state (32). A few states, such as Oklahoma, have additional restrictions on withdrawal of artificial nutrition or hydration (Appendix Tables 1 and and22).
Restrictions on who may serve as a health care agent have unintended consequences for the estimated 20% of homeless and uninsured persons (28, 33–35) and the 3% to 4% of older nursing home residents who cannot or do not wish to name a surrogate decision maker (24, 36). Homeless, institutionalized, socially isolated, disabled, or migratory patients often lack legally appropriate health care agents and frequently prefer to name their trusted case managers, social service providers, or physicians as surrogates (28, 35, 37). Health care agent restrictions may leave many isolated patients without advocates and potentially in the hands of court-appointed conservators who do not know these patients and may request treatment that is not in the patients’ best interests.
In addition, not recognizing same-sex or domestic partners in states’ health care decisions laws has unintended consequences for the 6 million Americans within these partnerships (38, 39). If these patients require medical care outside of the states that grant same-sex unions or recognize domestic partnerships, their partner may not be allowed to make medical decisions on their behalf. Because a patient’s same-sex or domestic partner often has greater knowledge of the patient’s values than statutorily recognized decision makers, such as estranged family (40, 41), presuming that the partner will “act in the patient’s best interest” (40) is reasonable. In this regard, having a patient’s same-sex or domestic partner act as his or her surrogate is morally and ethically justifiable and disregarding the partner’s input is unethical (40, 42).
Furthermore, restrictions on the types of decisions that agents or surrogates can make also may prevent patients from having their wishes honored. Some states require documentation of the exact decisions that an agent can make, which may affect patients who prefer not to write attestations of their wishes in advance, such as racial or ethnic minority populations and people with limited literacy (43–45).
Patients must navigate many execution requirements, such as attainment of signatures and witnessing, for advance directives to be considered legally valid. Only 16 states recognize oral advance directives (Appendix Tables 1 and and2),2), and many of these states have additional execution requirements, such as witnessing. Furthermore, several states (for example, Missouri) have separate statutes governing living wills and durable powers of attorney for health care, resulting in separate forms and requirements and potential conflicts (Appendix Table 1). In addition, nearly all states require 2 witnesses to make advance directives legally valid, with 18 states permitting notarization as an alternative. However, North Carolina and West Virginia require both 2 witnesses and notarization. Furthermore, most states prohibit a patient’s appointed agent, spouse, relative, health care provider, or an employee of the provider from being able to act as a witness. Some states even exclude anyone considered to be an heir from acting as a witness (Appendix Tables 1 and and22).
Improperly executed advance directives have caused patients’ documented wishes to be invalidated (Appendix Table 1). Even if physicians document a patient’s verbally expressed wishes, those wishes may not be honored because oral advance directives are not universally accepted. However, substantiated oral declarations are considered to be clear expressions of patients’ wishes and may provide the best evidence of their health care preferences (46). Ignoring such declarations may be legally justifiable but ethically problematic (40, 41).
Execution barriers may most affect patients from culturally and socially diverse backgrounds or patients with limited literacy or English proficiency. These populations often prefer not to or cannot document their treatment preferences and may be unable to navigate all required paperwork (33).
Witness restrictions may particularly affect socially isolated, unbefriended older adults, as well as patients who are homeless, institutionalized, or migratory and who often lack appropriate witnesses. Many disenfranchised patients also may not understand what a notary public is; know how to find one; or be able to pay for such services, which cost $0.50 in Wisconsin (47) and $10 in California (48).
In addition, a form of identification is required to use the services of a notary, which prevents patients without proper documentation from executing an advance directive. Finally, although some hospitals provide notary services, these are often unavailable during outpatient visits, potentially resulting in the completion of advance directives without a clinician’s input (11).
Reciprocity refers to whether an advance directive executed in one state will be accepted in another. Excluding Kentucky, Michigan, and Wyoming, 47 states and the District of Columbia have reciprocity laws. However, reciprocity laws ensure only that out-of-state directives be considered validly executed but do not ensure that the advance directive will be interpreted exactly the same way because of varying mandatory language, restrictions, and differences in how state statutes are interpreted (3). For example, a durable power of attorney for health care designated under Massachusetts law (49) includes the authority to withdraw life support (including artificial nutrition or hydration) and to consent to long-term care placement; whereas Wisconsin law requires that this authority be expressly documented (Appendix Tables 1 and and22).
Lack of reciprocity between states may most affect frail elderly people who live with various caregivers in different states, migrant and transient homeless patients, and persons who have the financial means to travel. Although court cases have shown that out-of-state advance directives may be enforceable as a clear expression of patients’ wishes, the acceptance of these directives is clinician-dependent and may require the financial, educational, and social means to retain legal counsel (Appendix Tables 1 and and22).
Although the importance of providing culturally competent care is well recognized (50), standardized advance directive forms do not permit the expression of alternative religious, cultural, or social preferences. For example, almost all durable power of attorney for health care statutes anticipate the appointment of a single agent rather than a family unit, and most statutes require completion of a written directive instead of an oral one. In addition, cultural, religious, or social preferences, such as death rituals or religious health care beliefs, have not been in the purview of advance directive statutes and therefore are not incorporated in most advance directive forms. Lack of inclusion of alternative religious, cultural, or social preferences represents a Western, autonomy-oriented, educated cultural bias.
The U.S. population is becoming increasingly diverse (39), yet differing views on autonomy and shared decision making among religious and cultural groups are often not considered (50–52). Most advance directives ignore cultural differences regarding autonomy (such as the concept of family decision making) and individual preferences (such as whether patients want to be aware of their prognoses) (51, 53). Western cultures generally view autonomy as self-empowering, whereas members of other cultures, such as Latinos, Native Americans, and Asians, may view it as a burden, resulting in loss of hope (50, 52, 54). Cultural or religious rituals about body preparation after death also differ. Yet, advance directive forms do not include such choices, which are as important to patients as treatment preferences (55). Decisions based on culture, religion, or spirituality often reflect a patient’s inherent values and should be elicited and honored (40).
Advance directive forms also discourage documentation of alternative preferences, such as the desire of isolated older and homeless persons to document their willingness to go to a nursing home or an institution and to discuss the ethical disposal of their body after they die (24, 56). Furthermore, homeless patients often wish to document personal characteristics (for example, tattoos) that may aid health care providers in body identification, preventing an anonymous death (35). In addition, patients with physical disabilities often prefer to document what constitutes good quality of life and their concerns about premature withdrawal of life-sustaining treatment (27, 57).
Because advance directive statues do not include cultural, religious, or social preferences, we identified few legal cases addressing these issues. The most prevalent cases involved adult Jehovah’s Witnesses whose wishes to refuse blood transfusions were not honored, even with legally executed directives (Appendix Table 1).
In addition to legal and content-related barriers of advance directive law, legal protection for clinicians may substantially affect patients and their families. Many states have provisions that enable physicians to presume the validity of an advance directive in the absence of actual knowledge that the directive is invalid (Appendix Table 3, available at www.annals.org). Most states also have cumulative clauses that consider advance directives as only 1 method to express patients’ preferences and may allow for other expressions, such as oral directives. Furthermore, all states have immunity statutes that protect physicians from criminal or civil liability or disciplinary action if they are acting on information in an advance directive in “good faith.”
However, nearly all states grant clinicians the right of refusal based on conscience or other objections (Appendix Table 3). The criteria for refusal vary considerably. Nineteen states and the District of Columbia have entirely open-ended criteria that permit providers to decline to comply with patients’ wishes for any reason. Most states acknowledge that providers are not required to act contrary to the standard of care, whereas other states variously permit noncompliance on the basis of one’s conscience and personal, moral, religious, and philosophical beliefs (58). If a provider invokes a conscience objection, states require the provider or institution to notify the patient and permit his or her transfer to another provider. Obligations range from merely refraining from impeding a transfer (Kentucky) (59) to requiring clinicians or institutions to transfer the patient or comply with his or her wishes within a specified interval (Florida) (60).
Immunity and conscience objection provisions may serve patients if surrogates are abusing their authority but also may give clinicians license to ignore patients’ wishes. Even if they receive advance notice of a clinician’s conscience objections, patients (especially those who are disenfranchised or live in rural areas) may have limited options. Thus, advance directive statutes meant to protect patients’ right of self-determination may instead better protect physicians from punitive action. Indeed, we identified only a few cases in which a physician faced punitive action. Our findings may reflect the inability of most disenfranchised patient populations to procure legal counsel, leaving many individuals without a voice.
Empirical evidence suggests that changing the focus of advance care planning from a legal-transactional approach to a relationship- and communication-based one would result in care consistent with patients’ goals (4, 7, 35, 61). In this paradigm of advance care planning, written advance directives are considered only 1 piece of information to be evaluated when a decision must be made (61).
Therefore, for advance directives to be clinically useful, advance directive laws must allow flexibility. Patients have differing needs, learning styles, and preferences for information when engaging in advance care planning. Patients also exist within a complex web of culture, religion, relationships, and experiences and have dynamic clinical courses that may change from moment to moment. Therefore, a one-size-fits-all legal-transactional model is not clinically effective.
Some experts suggest that merely designating a surrogate or adopting a universal advance directive form can remedy unintended consequences of advance directive law. However, these approaches do not address the legal restrictions on who may serve as a surrogate or witness; acknowledge that many patients may not have a surrogate (24, 36); or improve the readability, cultural inadequacies, or myriad execution requirements of advance directive forms. Others suggest using the Physician Orders for Life-Sustaining Treatment paradigm for patients with advanced progressive illness (15). However, these orders do not encourage discussions about the patient’s wishes before he or she develops an acute or terminal illness or address cultural, religious, or values-based preferences.
As advance care planning continues to evolve toward a flexible, relationship- and communication-based model, any advance directive tool or discussion could help to guide clinical care. The American Bar Association already has advocated a flexible approach to advance care planning through the Uniform Health-Care Decision Act (UHCDA), which was drafted in 1994 but only adopted by a few states (62). The UHCDA attempted to simplify state legislation by combining living will and durable power of attorney for health care statutes, allowing oral directives, and decreasing execution requirements. Building on the UHCDA, specific changes to advance directive law may help mitigate its unintended consequences on clinical care, particularly for vulnerable patient populations (Table).
To address readability barriers, we recommend eradicating mandatory legal language; writing advance directives at a 5th-grade reading level, which is the mean reading level of older persons in the United States; and, when possible, offering advance directives in patients’ native languages (19, 63). To address the problem of health care agent restrictions, clinicians should encourage socially isolated patients to discuss or document their health care wishes (28).
For patients who are unable or unwilling to do this and would prefer to designate a health care agent, clinicians should direct extensive effort toward helping patients to connect with social networks, distant family, or religious leaders. If a health care agent cannot be identified, we suggest that vulnerable patients with no other options be allowed to designate a professional who is not directly responsible for administering medical care, such as a social worker or case worker.
We also recommend eliminating health care agent or surrogate authority limitations, such as not allowing withdrawal from artificial nutrition and hydration. Clear documentation that the patient made these choices using his or her own judgment can prevent potential abuse (64). Internal ethics reviews also may be considered in such cases.
To address execution barriers, we recommend the universal acceptance of oral advance directives and the eradication of witness requirements. We also recommend that all states adopt nonrestrictive reciprocity laws, regardless of the location or type of advance care planning tools used. To address cultural and social insensitivity, we recommend allowing and encouraging patients to document their values, cultural traditions, and other socially or culturally important information and to consider group or shared decision making. By allowing flexibility to include religious, cultural, and social beliefs, clinicians and surrogates may determine a clearer course of action when deciding on the best treatment for an incapacitated patient.
At the bedside, many clinicians choose to use all available information from patients and their families to help provide the best possible clinical care (61). Many, but not all, states seem to accept such flexibility, given the immunity and cumulative clauses. A flexible rather than a legal-transactional approach to advance care planning may help to ensure that patients’ voices are heard even at the end of their lives.
This review has several limitations. First, advance directive laws are dynamic, and this assessment may not reflect the most up-to-date statutes. Second, most available case law information is found at the appellate level, which means that the cases have gone to trial and been appealed to a higher court. Because most disenfranchised populations lack the means to appeal, there may be many more cases that we were not able to ascertain. To elucidate the full scope of advance directive legal barriers, further studies should attempt to identify cases that have progressed only to an ethics review board or lower state courts. Finally, future advance care planning interventions should focus on eliciting individualized patient preferences and promoting discussions rather than solely encouraging advance directive completion.
In conclusion, unintended negative consequences of legal restrictions and requirements related to poor readability of advance directives; health care agent restrictions; execution requirements; insufficient reciprocity; and lack of attention to religious, cultural, and social issues may prevent all patients, and particularly vulnerable patients, from making and communicating their end-of-life wishes and having them honored. In an attempt to safeguard patient autonomy, legal restrictions have rendered advance directives less clinically useful. In addition, advance directive laws seem to protect physicians more than patients.
We recommend improving the readability of advance directives; eliminating surrogate restrictions; accepting oral and out-of-state advance directives; eradicating witness and notary requirements; and encouraging documentation of religious, cultural, and social beliefs. These changes could help to restore the clinical effectiveness of advance directives and ensure that all patients’ wishes are heard and honored.
The authors thank Desa Sanders, JD, and Heather Landis, JD, for their assistance with legal research and analysis.
Grant Support: Ms. Castillo and Dr. Sudore are supported by a Veterans Affairs Career Development Award and a Pfizer Fellowship in Clear Health Communication. Ms. Hooper is supported by the University of California, Hastings College of the Law.
Potential Conflicts of Interest: Dr. Williams: Grants received (money to institution): National Institutes of Health and University of California, San Francisco, Hartford Foundation Center of Excellence; Consultancy: Independent Medical Monitor of Michigan and Disability Legal Rights Center of Southern California; Expert testimony: University of Denver Student Law Office, Hunton & Williams LLP, and Holland & Knight LLP. Ms. Hooper: Consulting fee or honorarium (money to institution): Department of Veterans Affairs; Payment for writing or reviewing the manuscript (money to institution): Department of Veterans Affairs. Dr. Sudore: Grants received (money to institution): Pfizer Foundation. Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum = M10-1985.
Author Contributions: Conception and design: L.S. Castillo, R.L. Sudore.Drafting of the article: L.S. Castillo, B.A. Williams, S.M. Hooper, C.P. Sabatino, L.A. Weithorn, R.L. Sudore.
Critical revision of the article for important intellectual content: L.S. Castillo, B.A. Williams, S.M. Hooper, C.P. Sabatino, L.A. Weithorn, R.L. Sudore.
Final approval of the article: L.S. Castillo, B.A. Williams, S.M. Hooper, C.P. Sabatino, L.A. Weithorn, R.L. Sudore.
Obtaining of funding: R.L. Sudore.
Administrative, technical, or logistic support: L.S. Castillo, S.M. Hooper, R.L. Sudore.
Collection and assembly of data: L.S. Castillo, S.M. Hooper, R.L. Sudore.