Understanding Resistance to Honoring VSED Advance Directives in Dementia Patients: A Cross-Sectional Provider Survey

In the United States, patients with Alzheimer’s and other dementias make up more than 40% of the residents in care communities, hospices, residential nursing homes, and other care facilities [1]. These institutions collectively hereafter referred to as long-term care facilities (LTCs) and the practitioners within them are legally and morally charged with protecting these patient-residents from harm while respecting their right of autonomy. Capacitated patients may execute an advance directive (AD) to achieve a death in line with their own values and goals, and to avoid prolonged suffering at the end of life. The conceptual foundation of an AD is respect for autonomy.  When a person with decision-making capacity provides instructions to guide treatment decisions in future states of cognitive incapacity such as dementia, an AD serves as a valuable tool to ensure respect for the values of the earlier self [2,3]. 

The right to refuse life-sustaining interventions via an AD has been recognized as a legal, ethical, and moral right for decades. It is concerning that a quarter of providers are unwilling to honor an AD limiting life-sustaining interventions for a patient with dementia [4]. The act of voluntary stopping of eating and drinking (VSED) is a legally permissible option to hasten death for competent adults who find life with a progressive, irreversible disease unendurable. The right to voluntary stop eating and drinking is both a palliative and gerontological care issue. Some legal scholars cite the acceptance of VSED in capacitated patients as the precedent for allowing refusal of food and water by AD [5]. Recognizing that a patient with dementia has an equivocal right to refuse assisted oral feeding via AD is a recently emerging trend that pushes the boundaries of widely accepted medical practice at the end of life. This is known as voluntary stopping of eating and drinking via advance directive, or VSED-AD. 

The premise of a VSED-AD is much the same as any other directive made: an individual who wishes to avoid unwanted prolonged dying can specify a situation in which they would not want assistance or intervention, such as with eating and drinking if they could no longer feed themselves [6]. There remain conceptual disagreements as to whether feeding assistance is a basic personal care need like bathing and dressing, or whether the complex activity required feed a patient with advanced dementia is more advanced than basic care and therefore represents a kind of treatment [7,8]. Legal and ethical scholarship have found no distinction between the limits of bodily integrity from spoon feeding versus feeding tubes, or of the withdrawal of artificial feeding and the withdrawal of hand feeding [9,10]. 

The ongoing clinical and scholarly debates around VSED-AD entered public discourse in the recent highly publicized case of Susan Saran [11]. Diagnosed with frontotemporal dementia, a progressively fatal brain disease, Saran worked in conjunction with a lawyer to complete an AD that instructs caregivers to withhold hand feeding and fluids at the end of life. Her continuing care retirement community refused to honor the AD, incorrectly citing a federal requirement that they must provide her nutrition and hydration. However, that assertion was made in error. Medicare (CMS) requires that patient-residents are offered at least three meals per day at regular intervals and are provided help eating, if needed [12]. A patient is not required to consume the food. Misunderstanding the nature of offer versus obligation leads care facilities and providers to forcefully feed residents against a patient’s explicit written refusal of assistance out of fear that the facility will be issued an immediate jeopardy sanction [13]. Immediate Jeopardy (IJ) occurs when an entity’s noncompliance with one or more Conditions of Participation (CoP) has placed the health and safety of patient-residents at risk for serious injury, serious harm, serious impairment, or death. IJ noncompliance is among the most serious deficiency types and carries significant sanctions for providers and care-providing entities. 

We hypothesize that these persistent knowledge gaps and misunderstandings are directly affecting bedside care. There is some evidence that forgoing artificial or assisted hydration in patients with severe dementia is not associated with high levels of discomfort but there is minimal literature measuring the willingness of institutions to accept VSED-AD [14]. Previous studies of provider attitudes have been limited by a focus on hypothetical patient scenarios rather than real-world experiences [15]. 

This study intends to advance the current body of knowledge by surveying clinical providers about their actual practice experiences. We hypothesize that medical directors and the LTC institutions themselves do not commit to respecting a VSED-AD due to a misinterpretation of relevant federal regulations regarding immediate jeopardy and a misunderstanding of the interplay between federal regulations, state statutes, and case law [16]. The aim of this cross-sectional survey study is to identify and quantify the range of reasons given for (dis) honoring a patient’s VSED-AD, to clarify misinformation, and to offer actionable policy solutions to address the patient care barriers experienced by long-term care providers.

The survey was designed following a review of the literature and discussion among the research team to rank important topics. Survey responses were anonymous and respondents provided informed consent. The Columbia University Institutional Review Board granted approval for this study.  The survey tool was distributed via email link to the entire membership of The Society for Post-Acute and Long-Term Care Medicine (AMDA) in February 2022. Reminder emails were sent over the subsequent three weeks. Respondents were limited to those with specific professional roles at long-term care facilities based in the United States (Table 1). The survey consisted of a total of 66 questions. Fourteen questions were conditional and were only presented to a respondent based on their answer to a preceding question. Respondents could decline to continue the survey at any point. 

Table 1: Survey Respondent Professional Characteristics.

We received 291 survey responses and 138 respondents proceeded to full questionnaire based on their primary professional role (response rate 7.16%). Respondent characteristics are outlined in Table 1. All respondents are members of AMDA. Additional professional society memberships were reported to the American Geriatrics Society (n = 50, 36.23%), the American Academy of Hospice and Palliative Medicine (n = 25, 9.16%), and the Gerontological Society of America (n = 6, 2.19%). Eighty-four percent practice in some clinical capacity (n = 116) while 76.09% (n =105) serve as the medical director in their primary or secondary role. Most work alongside less than 10 patient-facing practitioners at their facility. No facility had greater than 50 patient-facing practitioners. 

A diverse range of LTC facility types were represented in the study sample (Table 2). The median bed count was 179 (range 24-500). Municipally owned LTCs (n = 5, M = 314.00, SD = 185.42) had on average more beds than non-profit (n = 47, M = 205.30, SD = 139.73; p = .05), for-profit (n = 67, M = 152.61, SD = 90.82; p = .004), and other facilities (n = 9, M = 167.89, SD = 146.96; p = .029). For-profit facilities had fewer beds than non-profit facilities (p = .021). 

Table 2: Facility Demographics. 

Patient-Resident Population at Intake 

CMS is the primary patient insurer at 72.33% (n = 81) of LTCs. Additional payer distribution is described in Table 3. A mean of 53.57% of patients enter the LTC with dementia (range 15-100%) while an additional 41% of patients develop a diagnosis of dementia while under the care of the LTC (range 0-98%). 

The majority of LTCs (95.69%, n = 89) asks a patient or their family about an advance care plan (ACP) at intake or shortly after admission. This is a requirement of the federal Patient Self Determination Act for all healthcare agencies receiving CMS reimbursement [17]. ACPs include ADs, living wills, and other documents specifying care wishes. Fewer than half of facilities (48.05%, n = 37/77) discuss oral nutrition and hydration at intake. Fifty-seven percent (n = 53/93) of respondents indicated that they or other staff at their facility have made use of CMS Code 99497 to reimburse for services related to voluntary ACP meetings. Sixty percent (SD = 24.32, range 5-100) of patients enter the LTC with an ACP document and 9.01% of documents mention VSED. Forty-five percent of respondents (n = 42/93) were “familiar” or “very familiar” with VSED, 35.56% (n = 32) were “a little” or “somewhat” familiar, and 20.43% (n = 19) were not at all familiar. 

Table 3: LTC Bed Size and Resident Payer Distribution. 

Regulation Familiarity 

More than half of respondents (n = 46/82, 56.10%) reported high levels of familiarity with IJ regulations. Only a small percentage (n = 8, 9.76%) reported no familiarity with the CMS regulations. There was no correlation between bed size or faith affiliation and familiarity with IJ regulations. Eighty-one percent (n = 66/81) correctly identified that an IJ finding should not be found in cases where a patient declined assisted oral feeding through an AD. Fifty-eight percent (n = 47) indicated that an IJ finding was not a concern for honoring a VSED-AD. Twelve respondents (14.81%) indicated that a potential finding of IJ was their primary concern and would stop the implementation of a patient's request to withhold nutrition or hydration. Approximately one-fifth of respondents (n = 17/74; 22.97%) indicated that a regulatory accusation of patient neglect was a significant concern. 

More than 91% of respondents (n = 74/81) were not familiar with a Ulysses Clause. A Ulysses Clause is a unique provision a patient can incorporate into an AD request for withholding care even when the patient forgets or objects after they have lost decision-making capacity. This clause serves as a prospective self-binding agreement to safeguard the patient's “authentic” care wishes from future “inauthentic” changes of the mind [18,19]. 

Staff Trainings and Site Policy 

Seventy-one facilities (86.59%) have specific practices or procedures for providing assisted oral feeding to patients with dementia. For-profit facilities were more likely to have institutional policies about IJ specific to assisted oral feeding compared to non-profits (p = .008) and other facility types (p = .039). Less than one-third (n = 23/81, 28.40%) of respondents had ever held or received staff meetings, trainings, or briefings about avoiding IJ in clinical practice. Approximately 70% of respondents (n = 56/81) had never had an opportunity to discuss concerns related to a patient who has declined assistance with eating and drinking in relation to a finding of IJ. Insufficient or absent staff training on patient care issues was a commonly reported experience amongst respondents (Table 4). 

Table 4: Respondent Familiarity with Policies and Trainings. 

Influence of Personal Beliefs on Care Provision 

Of the ninety respondents provided more information regarding their personally held convictions about patient and end-of-life care issues, 90% (n = 81) agreed with the statement: “Advance directives are useful for protecting patient wishes and values.” Although 73.33% (n = 66) felt that respecting a VSED directive fit into their idea of what “care” entails at the end of life, slightly less (n = 63) considered VSED to be a viable and an appropriate option for patients at the end-of-life. Fifty-nine respondents (65.56%) would personally feel comfortable with a patient choosing VSED in their facility while fourteen (15.56%) were neither comfortable nor uncomfortable. Support for VSED in patients with dementia increased from 55.56% (n = 50) to 72.22% (n = 65) if that patient had a VSED-AD. Although 62.22% (n = 56) felt their facility could accommodate patient wishes for VSED, one-fifth were unsure (n = 17). 

Sixty-nine respondents provided information about their own future end-of-life care planning. The majority reported that they personally would be “somewhat” or “very” willing to consider VSED for themselves in the event of terminal illness (n = 58, 84.06%) or in the case of late-stage dementia (n = 56, 81.16%).

Relation to Other Studies 

Our findings affirm those of prior research in this field while adding new quantitative data measuring the willingness of LTC practitioners to honor previously expressed treatment wishes by patients, including those with dementia, at the end-of-life. Four-fifths of respondents in this survey reported familiarity with VSED. This is consistent with previously reported rates among providers [20,21]. 

Regulatory Misinterpretation and Policy Uncertainty 

Our survey did not measure the depth or veracity of self-reported familiarity. One respondent who indicated they were “somewhat familiar” with VSED incorrectly stated that: “CMS guidelines will not allow us not to offer food for at least pleasure feedings, so VSED is not an option in a long term care/SNF” [sic, skilled nursing facility]. This response among others supports our hypothesis that providers are misinterpreting the guidance around offering nutrition. Honoring a VSED-AD in clinical practice does not prohibit an LTC from offering a meal to a patient as required by CMS. Although several states limit the withdrawal of oral nutrition and hydration regardless of the wishes expressed in a directive [22] the salient difference is that VSED is not denying, withholding, or withdrawing food or fluid to a patient. Rather, it is honoring a patient-expressed direction to caregivers. VSED represents part of a broader treatment plan whereby a patient can refuse any unwanted intervention. In 1984, a justice of the Supreme Court of the State of New York in Syracuse ruled that a nursing home was not obligated nor empowered to force-feed a competent man who wished to hasten his death with VSED. 

This study found that there was a 16% increase in provider support if a patient’s VSED wishes were expressed in an AD. Increased provider comfort with honoring VSED wishes that are expressed and documented when a patient had decision-making capacity suggests that a VSED-AD may be viewed more as an effective legal defense against a claim of IJ or negligence than as a means of protecting patient wishes and values. This confirms the presence of regulatory misinterpretation as VSED is not an IJ deficiency. 

For-profit facilities were more likely to have institutional policies about IJ and assisted oral feeding compared to non-profits and other facility types. This is likely due to more elaborate risk management programs in the for-profit sector. Regardless of funding structure, there was widespread confusion among respondents over permissions and a lack of understanding about how to apply facility standards and relevant policies in clinical and bedside practice (Table 4). 

Table 5: How long-term care providers can help manage end-of-life issues for patients, including those with dementia. 

^↓Responses to survey questions: (a) “What would assist your facility in managing end of life issues for those of your patients with dementia?” and (b) “Do you see room for improvement in how end of life care is managed at long term care facilities in general?” 

Provider Biases 

Respondents were asked about their individual knowledge and views on VSED to specifically understand whether personally held convictions were affecting the implementation of patient ADs. The results show that respondents were more likely to consider VSED an appropriate end-of-life option for themselves than for their patients. There was no statistically significant relationship between personal willingness to consider VSED for oneself and one’s primary professional role, facility type, funding structure, IJ protocol, or prior receipt of IJ trainings. Respondents were comparatively likely to report a willingness to consider VSED for themselves as they were to support their terminally ill patient who chooses VSED (84% vs 78%, p = 0.23). There was a significant difference among providers who would support a VSED directive from a patient with dementia compared to one’s own willingness to consider VSED for oneself in the event of late-stage dementia (55% vs 81%, p = 0.004). 

This suggests the presence of an action-bias paradox in long-term care practices. This bias refers to a general preference for action over inaction even if acting is likely to result in poorer outcomes, and can lead physicians to recommend treatments for their patients that are different from those they would choose for themselves [23,24]. 

Recommendations 

It is imperative to identify, quantity, and address areas of practice uncertainty. Uncertainty about the validity of medical treatment options leads to over diagnosis, overtreatment, and increased costs [25,26]. This is counterintuitive to the goals of geriatric medicine to optimize the care of older people by preserving function and maintaining autonomy as long as possible [27]. Over-action to appease situational uncertainty may appear to provide a safeguard for physicians from litigation, but it does not improve bedside care or patient outcomes. 

There have been repeated calls in the literature and in clinical practice for the development of policies and trainings to address LTC patient-residents who present with potential end-of-life care concerns like VSED-AD and symptom management [2,15,28]. Our research affirms the need for clear policies and improved staff trainings. Up to 41% of respondents were unaware of their facility’s policy or practice guidance on specific patient-care and end-of-life issues. More than two-thirds reported that their facility did not holding trainings on care topics likes oral nutrition and hydration. 

These findings can be translated through a model of preventative ethics into clear, actionable processes to assist LTCs in reducing uncertainty when caring for advanced age or cognitively impaired patients. Research-backed examples of implementable improvements include promoting daily staff huddles to provide updates and address specific case-related concerns, and providing more 1:1 engagement between supervisors and staff [28]. 

There were common themes that respondents provided (Table 5) when asked where they see room for improvement in how care is managed at their LTCs. Additional qualitative analyses, focus group interviews are planned to further assess, and quantity these responses. Until then, these themes can function as an initial framework to recognize, prioritize, and address systemic gaps in bedside care. Proactively formulating relevant policies helps individual clinicians make good decisions and equips staff with the relevant knowledge, skills, and guidelines to appropriately address future patient care requests or concerns that arise [29].

This study had several limitations. First, the response rate was lower than other survey studies of medical directors [30]. Response rates can be limited by office and facility policies to not participate in any surveys [31]. This survey was administered during the third wave peak of the COVID-19 pandemic, which we theorize lowered the response rate as clinical practitioners were navigating increased caseloads. Second, due to the anonymous nature of this survey, we cannot assess whether two providers in separate roles at one facility responded. However, the majority of respondents were medical directors and a facility will have only one medical director. Each IP address could only respond once to the survey tool. We feel the risk for such duplication is minimal. Third, the survey was provided to those with AMDA membership and eligibility was limited based on professional role. The sample is not representative of all providers but rather a sampling of those who, in their primary roles as medical directors and lead administrators, should be most informed of the relevant federal regulations, state statutes, case law, and clinical practices surrounding palliative, hospice, and end-of-life patient care issues. Although survey responses reflect the attitudes of a limited professional society, engagement in a professional society indicates that practitioners are engaged in quality of the care provided to gerontological patients in LTCs.

Our study reflects the knowledge variability among clinical providers at long-term care facility about care measures like VSED and how to honor advance directives. Directives are a beneficial communication tool but they must be properly implemented and upheld. There are common misinterpretations about the relevant federal regulations on assisted feeding and imminent jeopardy concerns [32]. This survey study finds that failures to honor a VSED-AD is often a multi-faceted response which begins with a fear of regulatory sanction and is inaccurately perpetuated due to procedural uncertainty and a lack of training. By characterizing common provider experiences in real-world clinical practice, this research identifies areas for direct and indirect training interventions that can improve palliative and end-of-life patient care in all practice settings. 

We suggest various approaches to improve gerontological care, specifically among patients with dementia. Gerontological and primary care providers must begin discussions early with their patients about nutrition and hydration issues that may arise at the end-of-life so that advance care plans contain actionable and intentional care goals that can be discussed if a patient does enter a long-term care facility. 

Care facilities would benefit from integrating the perspectives of relevant stakeholders who hold differing expertise in the fields of ethics, law, and medicine when creating standard operating procedures and policies. This diversity of stakeholder input ensures alignment with both professional and clinical best practice standards during the design and regular review of policies. Staff education and trainings on these topics must be prioritized. 

The discrepancies found between medical directors’ personal willingness to consider VSED compared to their lack of willingness to allow patients to pursue VSED should invite self-reflection and consideration on why providers hesitate to honor patient wishes regarding care decisions they do not conscientiously object to. Without clear guidance on policy, provider training, and actionable protocols on how to honor patient wishes as expressed in an advance directive, individual doubt and uncertainty will continue to unjustly affect patient care.

Conflict of Interest 

The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. 

Funding 

This work was funded by the Columbia University School of Professional Studies FY20 Dean's Applied Research Award and a grant from the Faith Sommerfield Family Foundation Inc. 

Sponsor's Role 

The sponsors had no role in the design, methods, subject recruitment, data collection, analysis, or preparation of this article.

1. Melekin A, Lu Z, Singh P, Lendon JP, Caffrey C, et al. (2022) Long-Term Care Providers and Services Users in the United States, Residential Care Component: National Post-acute and Long-term Care Study, 2020. National Center for Health Statistics. PP: 1-21.
2. Kollisch DO, Santulli RB, Bernat JL (2021) The Limits of Advance Directives in Maintaining Autonomy in Patients with Advanced Dementia. Am J Med. 134: 963-967.
3. Porteri C (2018) Advance directives as a tool to respect patients' values and preferences: discussion on the case of Alzheimer's disease. BMC Med Ethics. 19: 9.
4. Schoene-Seifert B, Uerpmann AL, Gerfs J, Herr D (2016) Advance (meta-) directives for patients with dementia who appear content: learning from a nationwide survey. J Am Med Dir Assoc. 17: 294-299.
5. Pope TM (2019) Whether, when, and how to honor advance VSED requests for end-stage dementia patients. Am J Bioeth. 19: 90-92.
6. Terman SA, K McGovern (2011) Peaceful Transitions: Plan Now, Die Later-Ironclad Strategy. Life Transitions Publications. PP: 1.
7. Ijaopo EO, Ijaopo RO (2019) Tube Feeding in Individuals with Advanced Dementia: A Review of Its Burdens and Perceived Benefits. J Aging Res. 19: 7272067.
8. Menzel PT, Chandler-Cramer MC (2014) Advance directives, dementia and withholding food and water by mouth. Hastings Cent Rep. 44: 23-37.
9. Vacco v. Quill, 521 U.S. 793 (1997)
10. Truog RD, Cochrane Tl (2005) Refusal of hydration and nutrition: irrelevance of the "artificial" vs "natural" distinction. Ann Intern Med. 165: 2574-2576.
11. Aleccia J (2020) Diagnosed with dementia, she documented her wishes for the end. Then her retirement home said no. The Washington Post. Available in Online.
12. HHS (2018) U.S. Centers for Medicare & Medicaid Services. Available in Online.
13. Hoffman DN, Schwarz J (2020) Can Patients Choose to Stop Eating – Even If They Have Dementia and Can Health Care Facilities Get Paid for Taking Care of Them? Ethics and Reimbursement at the End of Life. Health Lawyer Association Conference. PP: 1-20.
14. Pasman HRW, Onwuteaka-Philipsen BD, Kriegsman DMW, Ooms ME, Ribbe MW, et al. (2005) Discomfort in Nursing Home Patients With Severe Dementia in Whom Artificial Nutrition and Hydration Is Forgone. Arch Intern Med. 165: 1729-1735.
15. Volicer L (2016) Fear of Dementia. J Am Med Dir Assoc. 17: 875-878.
16. Miller DH (1984) In the Matter of the Application of the Plaza Health and Rehabilitation Center. Supreme Court, State of New York County of Onondaga.
17. Patient Self Determination Act of 1990, H.R. 5067, 101st Cong, 2d Sess (1990)
18. Quill TE, Menzel PT, Pope T, Schwarz JK (2021) Voluntarily Stopping Eating and Drinking: A Compassionate, Widely-Available Option for Hastening Death. 1st Ed. Oxford: Oxford University Press.
19. Brenna CTA, Chen SS, Cho M, et al. (2023) Steering clear of Akrasia: An integrative review of self-binding Ulysses Contracts in clinical practice. Bioethics. 37: 690-714.
20. Stängle S, Schnepp W, Büche D (2020) Family physicians' perspective on voluntary stopping of eating and drinking: a cross-sectional study. J Int Med Res. 48: 300060520936069.
21. Stängle S, Schnepp W, Büche D (2020) Long-term care nurses’ attitudes and the incidence of voluntary stopping of eating and drinking: A cross-sectional study. J Adv Nurs. 76: 526-534.
22. State Statutory Provisions Related to Orally Provided Food and Fluids and Comfort Care (2023) American Bar Association Commission on Law and Aging. Available in Online.
23. Zikmund-Fisher BJ, Sarr B, Fagerlin A (2006) A matter of perspective: choosing for others differs from choosing for yourself in making treatment decisions. J Gen Intern Med. 21: 618-622.
24. Tarrant C, Krockow EM (2022) Antibiotic overuse: managing uncertainty and mitigating against overtreatment. BMJ Qual Saf. 31: 163-167.
25. Ooi K (2020) The Pitfalls of Overtreatment: Why More Care is not Necessarily Beneficial. Asian Bioeth Rev. 12: 399-417.
26. Sekhar MS, Vyas N (2013) Defensive medicine: a bane to healthcare. Ann Med Health Sci Res. 3: 295-296.
27. Kotsani M, Kravvariti E, Avgerinou C (2021) The Relevance and Added Value of Geriatric Medicine (GM): Introducing GM to Non-Geriatricians. J Clin Med. 10: 3018.
28. McGilton KG, Escring-Pinol A, Gordon A (2020) Uncovering the Devaluation of Nursing Home Staff During COVID-19: Are We Fuelling the Next Health Care Crisis? J Am Med Dir Assoc. 21: 962-965.
29. Hester DM, Schonfeld T (2012) Guidance for Healthcare Ethics Committees. Cambridge: Cambridge University Press.
30. Williams JAR, Vriniotis MG, Gundersen DA (2021) How to ask: Surveying nursing directors of nursing homes. Health Sci Rep. 4: e304.
31. Wiebe ER, Kaczorowski J, MacKay J (2012) Why are response rates in clinician surveys declining? Can Fam Physician. 58: e225-e228.
32. Core Guidelines for Determining Immediate Jeopardy (2019) U.S. Centers for Medicare & Medicaid Services. Available in Online. PP: 1-22.