For patients living in the “Dementia Gap”,a newly defined substage of late-stage Alzheimer's disease,with intolerable suffering: Can sedation to unconsciousness be moral?

Past history

Sarah is a “Person with Incapacity from Neurodegenerative Etiology” (PINE). She is one of a huge number of patients including progressive Alzheimer's disease, related dementias, and those with stable brain disorders such as stroke and persistent vegetative state. She received the diagnosis of dementia eight years ago. For the last two years, she has been spending all her days in the stark hallway of her memory unit, strapped into but slumped over in a wheelchair, receiving quetiapine for sedation, but little social stimulation. Staff mostly ignores her unless her episodic difficult-to-manage agitation requires extra attention to calm her down. This behavior increased her monthly fee a year ago. Escalating expenses are draining her family's finances rapidly, which will limit her family's educational and business opportunities. How long she remains in this state may partly determine whether her family will face the threat of medical bankruptcy—which she would never want.^1–3

Sarah is suffering

Despite appearing peaceful or asleep most of the time, Sarah could be experiencing profound suffering that remains largely non-observable, difficult to detect and therefore goes untreated, or primarily affects her loved ones. ⁴ Most of her observable suffering is untreatable. Dementia caused the loss of all her important relationships since she no longer can recognize loved ones or close friends, communicate coherently and meaningfully, or recall memories of the peak life experiences they shared. Some of the consequences are difficult-to-detect loneliness and loss of sense of self that her previous relationships helped define. Dementia also disrupted her life narrative, including her roles as wife, mother, friend, and attorney. The consequence: she has lost most of her joy and meaning in life.

Sarah's loved ones are suffering

Janson et al. found that caregiver suffering includes “stress and adjustment disorders, depression, anxiety, sleep disorders, diseases of the spine and back, pain conditions, and lower quality of life.” ⁵ Sarah's loved ones also suffered from their “long goodbye.” As dementia progresses, her personality changed and her mind slowly died, making it impossible to enjoy her. Instead, they feel helpless as they experience profound “bi-directional empathic suffering” that acknowledges Sarah would be horrified if she had the ability to view her current condition with capacity that still understood her lifelong values. ⁶

To this incomplete list of sources of suffering is Sarah's haunting dread that her survivors will retain mostly tarnished memories of her final years during which she lacked dignity according to her former self's judgment. Example: caregivers must now change her diapers due to fecal incontinence.

The dismissive non sequitur fallacy

If casual observers (including healthcare professionals) conclude, “She's not suffering; she's just sitting there,” they will have wrongly extrapolated from her external state of apparent calm that her inner emotional state is mostly peace (except for her occasional episodes of agitation). This assumption could diminish the efforts of professionals and friends and relatives to investigate and treat her possible non-observable suffering (N-O-S).

Recently, Sarah's behavior changed

About a month ago, she began to repeatedly cry out, “Help me. Help me. Help me.” These vocalizations recently escalated in frequency and intensity. Increasing her dose of quetiapine and prescribing acetaminophen—first without, and then with codeine—did not help. At the institution's staff meeting, all concurred that (A) Sarah's suffering had progressed to the point where its intensity has become intolerable, and (B) non-sedating modalities of treatment cannot provide effective relief.

During advance care planning (ACP), Sarah completed an advance directive specifying Natural Dying (ND) if someday she reached an irreversible condition causing severe suffering. ND includes two physician orders: (A) cease assisted oral feeding and hydration, and (B) always place food and fluid within reach. The latter makes this treatment of last resort legal, ethical, and possible to view as moral. ⁷ But her current clinical state presents a challenge: ND cannot be implemented because she can still eat and drink independently. Withholding food and fluid is not an option; it is euthanasia, which is illegal in the U.S. Also, many view it as immoral.

To summarize, in the past Sarah was not quietly at peace; and now, her suffering is likely intolerable. Relief requires a treatment of last resort other than ceasing assisted feeding. She meets the criteria of living in a substage of dementia, a therapeutic desert, that has previously been rarely recognized, which this article refers to as the “Dementia Gap.”

Scenarios similar to Sarah's occur frequently across America's memory care units, skilled nursing facilities, and private homes. This article's Supplemental Material includes the estimate that about 500,000 patients are living in the Dementia Gap. Worldwide, the number is about 6 million. Therefore, characterizing advanced dementia as “a cruel, burdensome, and prolonged terminal illness” understates both individual depth and scope of challenges dementia presents to patients and their loved ones.

Fear of Alzheimer's disease among retirees ranks higher than COVID-19, cancer, heart attack and stroke” ⁸ despite the public's limited (if any) awareness of the additional challenges presented by the Dementia Gap and the phenomenon of N-O-S, which this article reveals.

The physicians’ role is to reduce patients’ suffering. So, if their orders prolong suffering, or fail to mitigate suffering, are physicians causing suffering?

The American Medical Association's Code of Ethics, Section 1.1.1, begins by reflecting this primacy ²⁸ : “The practice of medicine, and its embodiment in the clinical encounter between a patient and a physician, is fundamentally a moral activity that arises from the imperative to care for patients and to alleviate suffering.” If these two imperatives conflict, AMA's Code of Ethics Opinion 2.20 provides wise authoritative guidance: “The social commitment of the physician is to sustain life and to relieve suffering. Where the performance of one duty conflicts with the other, the preferences of the patient should prevail.” ²⁹ That medicine arises from the imperative to alleviate suffering is also the spirit behind the ARISE Movement this article proposes later.

If physicians write or continue orders that prolong the dying process of suffering patients who expressed the opposite preference, physicians thereby violate the bioethical principle of nonmaleficence (Do No Harm) ³⁰ —not by creating new harm, but by prolonging existing harm. Failure to prevent suffering violates nonmaleficence whether through action or inaction. Physicians are responsible for harms they cause by omission if they have (A) a duty to act (to relieve suffering), (B) the ability to act (available sedative medications), but (C) failed to act (not ordering effective treatment), thereby continuing or worsening patient suffering. ³¹

This guidance is consistent with: “The healthcare team members need to do as much as possible to benefit the patient (beneficence) while taking the necessary precautions not to prolong the dying person's suffering (non-maleficence),” and, “When the treatment provided by the healthcare team is medically futile or prolongs patient suffering, terminating the care and allowing the patient to die is necessary.” ³² “Physicians Obligations” include, according to Varkey (among others): “Avoidance of harm to the patient in the course of care.” ³³

AMA's Opinion 2.20 is unequivocal: patients’ preferences must prevail. Therefore, physicians must implement treatment that aligns with patients’ known wishes—whether learned from patients’ current capacitated expression, their legally designated agents’ current instructions, or past living wills. Opinion 2.20 can justify two physicians’ orders for treatments of last resort: (A) cease assistance with oral feeding and hydration, and provide adequate sedation if non-sedating methods of treatment failed to provide relief. In 2023, Gilbertson et al. wrote: “Dying patients…should not be left to suffer.” ³⁴ This sentiment echoes Freshwater's 1999 statement: “Patients must never be left to suffer so that doctors ‘stay out of trouble.’” ³⁵

The relevance of considering life expectancy when considering MA

In 2007, de Graeff et al. reviewed the controversial practice of continuous deep palliative sedation. Their recommended practice standards included: “Death should be expected within hours to days. ⁴³ Gilbertson et al. later extended life expectancy to six months. ³⁴

This article posits two points to argue that specifying any duration of time is arbitrary: First, several European countries and Canada use the criterion of unbearable suffering for medical aid in dying (MAiD), usually, with no other path available for relief that is tolerable, to which both patient and physician agree. ⁴⁴

Second, a precedent was established for the right of non-terminally ill capacitated patients to express their legal right to refuse food and fluid. Elizabeth Bouvia had cerebral palsy from birth. When she was 28, Bouvia challenged her physicians and hospital who were keeping her alive against her verbal, capacitated expressed will by force feeding her through a nasogastric tube. A 1986 California Court of Appeals set the precedent for the legality of refusing food and fluid. ⁴⁵ The ruling explained, “The fact that she is young and not terminally ill does not negate her right to refuse medical treatment; it only lengthens the time during which she must endure the very suffering and indignities she seeks to avoid if treatment is imposed against her will.”

Table 1, a decision matrix, clarifies key terms. The middle row covers typical “dementia-specific” living wills for patients who require assisted feeding. The bottom row covers patients in the Dementia Gap who retain independent eating and drinking ability.

OPEN IN VIEWER

Table 1.

Treatment selection based on suffering severity and eating/drinking ability for PINEs (Persons with incapacity from neurodegenerative etiology).

	Mild suffering	Irreversible suffering (including non-observable)
PINE depends on assistance with feeding and hydrating	Careful hand feeding or comfort feeding only	If severe: Natural Dying = refuse assisted oral feeding and hydrating
PINE can eat and drink independently	No intervention needed	If intolerable: Moderate Anesthesia with no artificial nutrition and hydration

“Success”: a behavioral definition for advance care planning

ACP can be considered successful if, after patients reach their previously judged “qualifying condition(s),” their treating physicians promptly write and implement orders for treatment that, in patients’ opinion, delivers goal-concordant end-of-life care.

During ACP, planning principals are responsible for judging suffering that 36 conditions would cause them or their loved ones—based on their lifelong values and current treatment preferences. Example: “If I reach a condition where I cannot interact with other human beings, even at my own birthday party, then I want either Natural Dying or Moderate Anesthesia—depending on whether I can still eat and drink without help.”

Case 1, “BB”: Moderate Anesthesia for intolerable suffering

Case 2, “AK”: Polling AK's PDC members to decide on MA. Establishing moral justification when the Rule of Double Effect could not apply. Understanding the source of her husband's unexpectedly profound grief

As I entered her home, AK was gracefully poised on a tufted leather couch. Her physical presence had changed little in the five years since I helped her complete ACP. As I approached, she made eye contact and smiled, but with no hint of recognition. She fluidly rose and extended her hand. Yet each time I tried to let go, she held on and shook again. After five shakes, I exerted enough effort to withdraw my hand.

Her husband's exhausted eyes were surrounded by deep creases, not surprising as he reported fourteen hours of daily vigilant caregiving. When I remarked on how well AK appeared, he shared that she still walked four miles a day. In other circumstances, this fact would be celebrated as a sign of vitality. But here, it starkly warned that physical longevity could prolong her existence with prolonged and devastating loss of cognition and suffering.

AK's first signs of dementia appeared seven years earlier. Her deterioration accelerated in the past year. Her husband requested my reevaluation for two reasons. First, their three children and several close friends spontaneously shared that AK would not want to prolong her dying in her current mental state due to loss of dignity and imposing burdens on others. Second, authorities at two recently visited memory units refused to implement Natural Dying if she became dependent on assisted feeding in the future. While they admitted this option was legal, they refused to allow it on their watch.

Non-sedating treatments, including antidepressants and antipsychotics, provided little relief from emotional lability that sometimes manifested as combativeness toward caregivers. Several weeks ago, she eloped: She managed to sneak out of their home, find the keys hidden in their truck, and take a “joy drive.” —Shocking, amazing, but dangerous. A repeat performance might cause serious harm to herself and others.

AK's memory deficit was so profound that she could not state her lifelong career had been nursing. Her once-precise medical vocabulary and passionate political discussions had deteriorated into strings of words lacking logical connection or meaning (word salad). While she could still grab a sandwich from across a table, she remained silent unless someone talked to her. Then, her verbal responses were incoherent; she could not engage in responsive, relevant, meaningful conversation. She rarely recognized anyone except perhaps her husband on occasion, although she was usually calmer if sitting next to him. She was in CDR-3 ²¹ and FAST Stage 7. ²²

During their visit to two memory care units, AK became agitated—an unusual response for her. Her husband heeded the advice of her experienced caregiver: AK would not tolerate institutional care well. She would need to give up the comfort and personal attention of home to live among strangers. Yet her elopements demanded vigilance: she was no longer safe at home.

Polling PDC members: do AK's current conditions qualify her for MA?

PDC members’ polling resulted in unanimous agreement: 6 current conditions were causing intolerable suffering and thus qualified her for MA, as shown in Table 3 and Figures 2–7.

OPEN IN VIEWER

Figure 2.

Loss of ability to recognize close relatives and friends.

OPEN IN VIEWER

Figure 3.

Depends on others to change diapers due to fecal incontinence.

OPEN IN VIEWER

Figure 4.

Cannot communicate—even to say “Yes” or “No”.

OPEN IN VIEWER

Figure 5.

Fighting caregivers.

OPEN IN VIEWER

Figure 6.

Elopement risk.

OPEN IN VIEWER

Figure 7.

Loved ones’ suffering includes “ambiguous loss” for patients who can self-feed and self-drink—who are living in the Dementia Gap.

OPEN IN VIEWER

Table 3.

Polling patient decision committee members’ substituted judgment for conditions AK would have judged are currently causing her intolerable suffering.

Fig. #	Condition: what can be observed; inner feelings	Husband/agent	Caregiver	Child 1	Child 2	Child 3
2	Loss of ability to recognize close relatives and friends confused; lonely	Y	Y	Y	Y	Y
3	Depends on others to change diapers due to fecal incontinence; dreaded dependency, loss of dignity	Y	Y	Y	Y	Y
4	Cannot say or indicate “Yes” or “No”; frustrated, cannot get what she wants without help	Y	Y	Y	Y	Y
5	Fights with caregivers; confused, afraid, feels attacked	Y	Y	Y	Y	Y
6	Elopement risk; scared and desperate	Y	Y	Y	Y	Y
7	Loved ones’ suffering includes “ambiguous loss”: loss of cognitive functioning is devastating but can self-feed and self-drink; is living in the Dementia Gap	Y	Y	Y	Y	Y

Comments: Figure 1, readers will recall, shows the Dementia Gap. Each condition was presented with verbal descriptions written at grade-three level reading comprehension. While AK and members of her PDC were originally shown black and white line drawings, this article presents the 2026 color upgrades.

These illustrations demonstrate how high-resolution color images help planning principals express, and PDC members and physicians understand and appreciate, the intensity of suffering that conditions cause: (A) facial expressions that convey recognizable emotions with immediacy, and (B) thought bubbles containing words that reveal the inner experiences of patients and others affected. Together, these elements enable those concerned to identify sources of N-O-S.

Case 3, Terry Matthews: Rapid induction of deep sedation to unconsciousness for intolerable pain caused by advanced cancer

Terry Matthews was a 31-year-old wife and mother who died of advanced cancer more than a quarter-century ago. Her treating physician, Ira Byock, published her sad story in his 1997 book. ⁵⁷ Today, it still inspires ethical and moral debates.

Terry's fight with cancer began six years before her death. After initial therapeutic success, she believed (hoped) she was cured and adopted a child. Sadly, her cancer recurred. Toward her end, despite increasing pain, she refused to accept that dying was inevitable, let alone imminent. Instead, she tenaciously held onto life through her relationships with her husband and children. The price she paid became unbearable pain, which Byock noted was the worst he had seen in his then-fifteen years of providing palliative care.

Terry's oncologist bluntly confronted her by predicting she would die within three weeks. This was likely his attempt to motivate her to accept her grim prognosis so she would consider available treatments of last resort to relieve end-of-life suffering. But Terry took no action until she characterized her pain as “paralytic” and a “crisis” that caused her to shriek, “I feel like I’m on fire.” After she extraordinarily depleted local pharmacies’ supplies of IV Dilaudid and found 900 mg per hour of morphine ineffective to provide relief, Terry finally admitted to Byock that her pain had become unbearable. About that time, Terry's oncologist called Byock to suggest implementing a treatment of last resort that a group of local physicians had thoroughly discussed but used only once: administer the rapidly acting anesthetic thiopental IV to eliminate Terry's conscious awareness of pain.

Case 4, Brittany Maynard: Her decision on timing Medical Aid in Dying

After months of headaches, newlywed Brittany Maynard received a diagnosis of incurable glioblastoma. Adamant about dying on her own terms, she relocated her family from California to Oregon—then one of three states where “Death With Dignity” was legal. She ultimately ingested a lethal dose of physician-prescribed medications. She died shortly before her thirtieth birthday.

Brittany became a volunteer advocate with Compassion & Choices, the largest right-to-die organization in the U.S. With their support, she created a goodbye video ⁶⁰ that 12 million people viewed. It stated Oregon's law required “unbearable suffering.” Yet Brittany admitted, “I still laugh and smile and enjoy walks,” which her video demonstrated.

Brittany's main worry was waiting too long to die: “If I wait too long, my autonomy may be taken away from me.” This concern parallels the Dementia Fear. Her suffering was mostly anticipatory: fear of developing morphine-resistant pain, personality changes, verbal, cognitive and motor loss, and intractable seizures. None of these conditions had yet occurred, but they do afflict some glioblastoma patients. Even so, they might have been treatable for several months.

Brittany ended her life while she still enjoyed living. Her dying was reported to be peaceful with full support from her husband and mother. They continued to advocate to legalize Medical Aid in Dying.

Brittany thus prioritized preserving her autonomy over maximizing her time to enjoy living. Conceivably, supportive psychotherapy combined with realistic counseling about glioblastoma's symptom trajectory and the likely effectiveness of palliative care could have extended her life by months. Her anticipated worst-case scenarios might not have materialized, or might have been manageable for a while. But if she lived into her thirties, as others have pointed out, her story might have been less poignant and perhaps less effective. California passed its End-of-Life Option Act a year after she died. As predicted, she became the new face of the American right-to-die movement. ⁶¹

Case 5, Gladys Wilson: Skillful, compassionate human connection awakened her human and spiritual passion

Gladys Wilson, an 87-year-old institutionalized PINE had not spoken in six months. She exhibited repetitive behavior—hitting the arm of her chair—apparently without purpose. Then, Naomi Feil interviewed her using Validation Therapy, the therapeutic approach Feil founded. ⁶²

Feil began the six-minute video excerpt of their interview with a clinical, educational statement:

“When no one enters their world and they are just sitting there, they will withdraw inward more and more, and their desperate need for human connection is all inside.” ⁶³

Technically, Feil combined empathetic listening, sustained and intense eye contact, and therapeutic touch. She cupped Gladys’ head in her hands and sang hymns that she knew would be familiar to Gladys, given her strong religious background.

First, Gladys shed one small tear. Then she changed her repetitive banging behavior to match Feil's speaking volume and tempo that reflected her changing mood, sometimes calm, other times passionate (as Feil changed). At one point, Gladys stopped banging and reached over to touch Feil's shoulder—an action consistent with appreciating Feil's contact. As Feil sang the next hymn, Gladys broke her long silence—first in whispers, as expected for someone who had rarely vocalized—and then more loudly. When Feil asked: “Do you feel safe with Jesus?” and, “Do you feel safe with me?” Gladys answered “Yes” to both, with conviction.

A speculation about Gladys's repetitive arm banging: While several neurological etiologies are possible, it might have been her only available means to communicate feeling ignored, lonely, and desperate for human contact. If arm banging signaled such distress, dismissing it as meaningless, as part of the “just sitting there” fallacy, profoundly dishonors her and deeply increases her suffering.

Some of Gladys's previous casual observers might be astonished if they viewed this video since Feil's skilled intervention demonstrated Gladys's ability to connect personally and spiritually.

Three haunting questions remain:

Was the intensity of Gladys's suffering during her silent months proportional to her passion when Feil made contact?

Case 6, Henry Dryer: Awakening his hidden passion by personal investigation

Casual observers dismissed the withdrawn behavior of 94-year-old nursing home resident Henry Dryer as typical for patients living with dementia. For endless hours, he remained slumped over in his chair, head down, hands clasped. Some described him as “disconnected from his environment.” They were proven wrong when a recreational therapist discovered, through research and experimentation, that Henry responded to jazz music of his era. ⁶⁴

When headphones playing jazz were placed over his ears, Henry had a dramatic awakening. He played jazz drums on dishes with silverware as he moved his body to the rhythm of the music. And for those willing to listen, he clearly, coherently, and meaningfully expressed how important jazz was to him. Example: “The Lord came to me and made me holy, giving me these sounds.” ⁶⁵

The passion that music evoked in Henry ⁶⁵ was memorialized in the aptly named documentary, “Alive Inside.” ⁶⁴ The intensity of his response invites this question: Was the intensity of Henry's suffering during his months of silence proportionate to the passion that music now awakened? If so, how many other patients sit in institutions while others dismiss their withdrawal simply because no one has dedicated the skills and time to discover areas of special interest that might awaken them?

The cases of Gladys and Henry inspire hope that professionals and family members will not make dismissive assumptions about patients “just sitting there,” but instead will try to discover their passion and awaken their capacity to engage. As the database of successful awakenings increases, it may become easier to motivate clinicians to increase their therapeutic efforts and to inspire an educational movement grounded in this universally accepted principle: Every patient deserves treating clinicians and caregivers who recognize their unique humanity, who seek to understand their individual needs, and who work diligently to connect with them—especially when patients no longer can advocate for themselves.

Discovering the Dementia Gap—a huge, hidden humanitarian crisis; and appreciating the magnitude of non-observable suffering (N-O-S)

This article is likely the first to describe a newly defined substage of dementia, which it refers to as the “Dementia Gap,” and which is characterized by four clinical criteria: Patients have (1) reached late-stage Alzheimer's disease or related dementia (which implies they have lost decision-making capacity); (2) retained their ability to eat and drink independently; (3) experienced intolerable suffering from combined observable and non-observable sources; and (4) failed or would likely fail to respond to any non-sedating treatment.

After our 2022 critical review of the then-available “dementia-specific” advance directives, ⁶⁶ continued monitoring of newly published living will offerings has, to date, yielded none that recognizes the existence of the Dementia Gap and its additional challenge to reduce patients’ suffering. If so, then defining the “Dementia Gap” is a truly new discovery.

In turn, this discovery provokes a three-fold educational event: first, the Dementia Gap exists; second, it affects an estimated 500,000 or more Americans; and third, a proposed treatment to relieve suffering is challenging since it must be both effective and morally justified.

Living wills that omit the Dementia Gap risk this dire result: All the diligent work planning patients invest in completing their living wills can fail; surprise and disappoint family; and leave patients ineffectively treated for as long as they remain in the Dementia Gap.

This Discussion strives to answer four crucial questions: Why has the Dementia Gap not been previously recognized? What treatment would be effective? Can it be morally justified? What other obstacles need to be overcome so the proposed treatment can be widely accepted and thereby reduce the suffering of these vulnerable, inarticulate, incapacitated patients?

Why has the “Dementia Gap” substage previously not been recognized?

The pool of patients who can sound the alarm of living in a therapeutic desert is small, even though the total number of Dementia Gap patients is large. First, they must manifest this clinical disparity: great loss of cognitive functioning but minimal loss of self-feeding and drinking abilities. Second, patients must endure suffering to desire a treatment of last resort sufficient to accept the two risks of sedation: permanent loss of communication ability and possible earlier dying. Third, their agents and family must be impacted by this stark discovery: their existing living will—despite being touted as “dementia-specific” by its drafters—cannot reduce patients’ suffering until the patient becomes dependent on assisted oral feeding and hydrating—a disease progression that might take months, even years.

Non-Observable Suffering: a greatly needed, enhanced clinical insight

Interest in this article's second clinical contribution, N-O-S, was inspired in part by a healthcare provider sharing (in 2015) the story ⁶⁷ of a memory care unit patient who seemed to be “sitting there peacefully” but as all eventually learned, she was not just withdrawn; dementia had evidently destroyed her ability to speak and move. Hot coffee had been accidentally spilled in her lap, but she was unable to complain and did not writhe in pain. Yet the resulting second-degree burns required weeks of rehabilitation. She suffered untreated in silence for the first two post-burn days, until staff discovered the pathognomonic welts during routine continence care. In retrospect, when she most needed palliation, she received none.

In 2024, we broadened the concept of suffering in an article ⁴ whose Table 1 lists five general categories and examples of hard-to-observe or non-observable suffering: physical pain; psychic or emotional pain; existential suffering; disruption of life narrative, and burdening loved ones.

N-O-S is more frequently the source of severe to unbearable suffering than previously recognized. It hides in plain sight, masking itself as withdrawn behavior that clinicians and caregivers often mistake for calmness that some clinicians describe as: “She's not suffering; she's just sitting there.” Sometimes, N-O-S appears alongside observable suffering, but accounts for far more of the patient's total suffering than observable sources. Finally, N-O-S may exist entirely on its own, with nothing visible to betray its presence, such as living without jazz music, singing religious hymns, or enduring pain from second-degree hot coffee burns. To reduce how much and how long patients and family must endure suffering requires a living will that deliberately strives to identify these sources so they can be treated—otherwise, a huge population of patients will not be identified as needing treatment for their suffering. This article describes an illustrated patient decision aid (IPDA) designed to accomplish this goal.

The relationship between Dementia Gap and N-O-S: overlap and differences

In common, both are greatly underappreciated, and no existing living will, to my knowledge, includes ways to recognize, investigate, or treat the suffering each can cause.

They differ because the Dementia Gap is a duration of time that warrants recognition as a substage of late-stage dementia, based on patients fulfilling its clinical criteria. In contrast, N-O-S is a specific category of suffering that causes patients to endure severe, intolerable, or unbearable experiences that can be as intense as or greater than observable suffering, whether or not any observable suffering is present, but its source is either hidden or difficult to discover. Physicians do not treat what they cannot observe.

These two terms have some overlap. Both describe situations that rank among the most difficult to treat. The Dementia Gap thwarts the effectiveness of “dementia-specific” living wills that do not address patients who retain the ability to feed and hydrate themselves. N-O-S can occur at any stage of dementia, but its incidence increases as patients gradually lose more of their abilities to complain and manifest behavior consistent with the experience of pain or suffering, such as writhing or agitation. The greatest challenge emerges when their behavior becomes withdrawn.

The iceberg phenomenon

The extent of PINEs’ suffering in the Dementia Gap fits the epidemiological “iceberg phenomenon:” the observable burden represents only the tip. ⁶⁸ The exact N-O-S to observable ratio is unknown. The studies quoted in Supplemental Material lead to a conservative estimate that more than 500,000 Americans currently live in the Dementia Gap with intolerable suffering. The number who experience severe but not intolerable suffering must be much greater. Applying my estimate of 12:1 global-to-U.S. ratio for all advanced dementias from several sources including Cavaco et al., ⁶⁹ approximately 6 million patients worldwide live in the Dementia Gap. These numbers may double by 2060. ⁷⁰ Thus, the scale of this crisis is huge.

Multiple lines of evidence validate the existence of Non-Observable Suffering

(1) Planning principals implicitly endorse the existence of N-O-S as they comply with this primary task of ACP. They manifest no hesitation when asked to judge how much suffering a condition would cause themselves and loved ones (choosing from five intensities—mild, tolerable, severe, intolerable, or unbearable). The patient decision aid does not ask them to specify what types of N-O-S or what proportion of suffering they versus their loved ones would bear; however, these nuances are frequently discussed during video-memorialized ACP interviews, which are often valuable to help guide decision-makers who review them later.

Pascal's logic depends on both outcome probabilities and magnitudes of what is at stake. Table 4 shows that if clinicians wrongly assume “She's not suffering” when N-O-S exists, the consequence can be catastrophic (like Hell): intolerable suffering for months to years until death. If clinicians assume N-O-S might exist andthey thoroughly investigate it but fail to “awaken” the patient, then the cost of failure—including time, money, and frustration—is modest (like being a good person living a good life who eventually learns Heaven does not exist). Pascal's logic applies to all inarticulate patients, including post-stroke, ALS, and locked-in syndrome patients.

OPEN IN VIEWER

Table 4.

The N-O-S moral imperative: a clinical application of Pascal’s Wager.

Clinical Assumption	YES: N-O-S Conditions Exist	NO: N-O-S Conditions
N-O-S might possibly exist → Try to awaken → Rigorously investigate → Attempt to make contact → Treat hidden sources	Optimal outcome Detect suffering Relieve suffering Restore human connection Prevent premature sedation Honor patient’s humanity	Minor cost Investigation time Staff resources Family emotional energy But no harm to patient
N-O-S does not exist → “Just sitting there” → No investigation → Dismiss withdrawal	Catastrophic outcome Patient suffers in silence Suffering intensifies over time No relief until much later death Profound dishonor to humanity Moral failure	Assumption correct Appropriate comfort care No unnecessary burden Doubt lingers; grief worse

A history lesson: advocating for voiceless sufferers—the anti-vivisection movement

Research animals in vivisection experiments share four devastating characteristics with PINEs:

Verbal inarticulacy: Animals cannot describe pain in human language; some PINEs can neither verbally nor physically complain.

The anti-vivisection movement demonstrates both the possibility and difficulty of protecting voiceless populations. Advocates challenged dismissals of animal suffering beginning in the 1860s in the UK with mass demonstrations including the “Brown Dog” riots. Yet a century later, a new law was still needed: the UK's 1986 Animals Act. ⁷⁵ The lesson: those opposed to protecting the voiceless vulnerable can require a sustained movement across generations.

There is one crucial difference: Suffering inflicted in animal research has decreased considerably, but the number of PINEs experiencing N-O-S will double by 2060. Without an effective movement, millions of humans will suffer in silence.

The ARISE Movement: Advocate by Recognizing and Investigating Suffering's (hidden) Existence

This article calls on clinicians, educators, and institutions to ARISE—to fulfill their moral obligation to vulnerable, inarticulate patients by serving as their Advocates, Recognizing that Non-Observable Suffering exists, and Investigating Suffering's hidden Existence by applying the required skills, time, and compassion for withdrawn patients who are NOT “just sitting there.”

Comment: “(hidden)”—in Italicized font and parentheses in the definition and acronym signals that the suffering on which the movement focuses is hidden. While “h” is not included in the acronym, hidden is implied by focusing on “all” suffering.

Can Moderate Anesthesia be moral?

“ Yes,” if physicians first diligently attempt to relieve patients’ suffering using non-sedating treatments, and second, increase the dose of sedative medications proportionately, reserving high doses to patients requiring dissociation from awareness of suffering. MA can help patients avoid unwanted, non-beneficial treatment that prolongs refractory suffering.

Regarding the moral argument, “foreseeability equals intended”

Critics could argue: sedated patients cannot eat or drink, and no artificial nutrition and hydration is provided; therefore, dying is inevitable—not merely a foreseeable possibility. How then can earlier death not be intended? Here are two counterarguments:

First, if light sedation does relieve suffering, the protocol's next step is to allow patients to resume eating and drinking or, if then needed, provide oral assisted feeding and hydration.

Second, this “inevitability objection” subjects MA to a higher ethical standard than withholding or withdrawing other life-sustaining treatments from terminally ill patients. Clinicians and ethicists rarely argue that withdrawing ventilatory support, dialysis, or pacemakers intentionally cause death. Instead, this is the prevailing ethical position: withholding and withdrawing non-beneficial, futile treatments for terminally ill patients—especially if such treatment prolongs severe suffering—is ethical, legal, and a patient “claim right” (a right that imposes a correlated duty on others).

How did surrogate decision-makers decide to implement treatments of last resort and can their decisions be considered moral?

Designated agents for BB (who had moderately severe dementia) and for AK (advanced dementia) decided on MA; Terry Matthews gave verbal consent herself. Table 5 compares how these decisions were made and morally justified. Additional highlights and Comments are below.

OPEN IN VIEWER

Table 5.

Clinical decision-making and moral justification for three patients administered Moderate Anesthesia.

	Factor or characteristic	“BB”	Terry Matthews	“AK”
A	FACTORS THAT LED TO DECIDING TO IMPLEMENT MODERATE ANESTHESIA (MA)
1	Diagnosis, stage	Dementia, moderately severe	Metastatic Renal Cell Carcinoma	Dementia, advanced
2	Range of life expectancy expected without any intervention	Weeks	Days	Years
3	Was the patient's intolerable suffering expressed/documented?	Yes, by self-report and observing her behavior	Yes, by self-report	Yes, by polling Patient Decision Committee members’ substituted judgments: Would AK have judged conditions that she previously judged as causing “severe” suffering (in her living will) as “intolerable suffering”—if this choice had then been available?
4	Had less-sedating methods of treatment been tried but failed to relieve her suffering?	Yes	Yes	Yes
5	Did the treating physician and patient face a forced choice between either suffering or possible unintended loss of consciousness?	Yes. After failed attempts to treat psychiatric symptoms while preserving consciousness, total unconsciousness became necessary to relieve suffering—foreseen, but not initially intended	Yes. Unconsciousness was intended as her treatment of last resort to dissociate her from being aware of her suffering	Yes. Unconsciousness was intended as her treatment of last resort to dissociate her from being aware of her suffering
6	Could most of her intolerable suffering be appreciated by direct, current observation of the patient?	Yes. Agitation and depressive statements	Yes. She had capacity to articulate her complaints	No. Non-Observable Suffering from dreading future tarnished memories, exhausting caregivers, loss of purpose/meaning in life, embarrassing behaviors, lost relationships/loneliness, confusion when helped, and her elopement may have been based on fear
7	Were there clinical factors that made reducing her consciousness compelling?	Yes: mental anguish, psychotic mood, agitated behavior	Yes: intractable, untreatable, unbearable pain; 5 inches of her Portacath wire migrated to her right atrium	Yes. See Table 2. Her clear and convincing living will stated she prioritized reducing suffering over sustaining her life since it would prolong her dying
8	Were family members and friends suffering? What intensity?	Yes, severe	Yes, severe to intolerable	Yes, severe
9	Did patient fear further loss of all her important relationships by dying?	Not as important as her primary goal, to get relief from her current suffering	Yes. This source of suffering was the main reason she endured pain (until it became unbearable)	Likely no since she already lost most of her loving relationships due to impaired memory, inability to recognize them, and to communicate
10	What type of memories will she likely leave her surviving loved ones?	Mostly good, since her final ordeal was short	Both bad and good: prolonged, unbearable pain like unforgettable; valiantly dedicated to remain engaged with family for as long as possible	Mixed but progressively worse since her ability to communicate meaningfully and to behave in socially accepted ways were deteriorating; some behaviors, such as elopement, were dangerous to all
11	What immediate precipitating factor led to considering implementation of Moderate Anesthesia (MA?)	Current observable behavior such as agitation, anxiety, and depression—all expressed with desperation	Intractable pain led the patient to beg her physician to stop it and help her enter Heaven	Living will's clear expression of past preferences combined with PDC members’ current substituted judgment—now that Moderate Anesthesia was available
12	Did all directly involved agree with implementing MA?	Yes	Yes	Yes
13	Were other clinicians consulted? Did they concur with implementing MA?	Palliative care physician and primary care physician agreed she qualified for hospice but MA was not discussed in our phone calls	Treating oncologist suggested palliative care physician implement sedation protocol	Two board-certified palliative care/hospice physicians considered treating physician's oral summary and then concurred with the treatment plan to implement MA
14	Did anyone later express regret using MA, including other MDs?	No (did not contact MDs as they were not involved)	Not reported, but likely no	No, in follow-up Zoom calls with each consulting MD: no problem in not requesting a formal ethics consult
15	In retrospect, was the decision to implement MA timely?	Yes, definitely. But managing her behavior and mood were difficult for longer than a day: it took that long to empirically reach effective doses of sedating medications	No, definitely. Physicians could have informed patient about sedation providing relief earlier—before escalating pain and exhausting all local pharmacies’ supply of IV Dilaudid	Likely yes. Postponing MA risked greater personal suffering, further burdening loved ones, and another more destructive elopement. She would not be happy residing in a memory unit
16	What ultimate condition was likely if the intervention had not been implemented?	A repeating cycle of agitation to exhaustion followed by temporary rest until she died of a heart attack, stroke, or infection	Dying in an even worse crescendo of increasingly unbearable pain	Reaching end-stage dementia, where she would lie immobile in the fetal position while her exceedingly severe suffering might not be detected and therefore not be treated.
B	TREATMENT COURSE AFTER IMPLEMENTING MODERATE ANESTHESIA
17	Was a trove of sedating medications immediately available?	Yes. Medications, including DEA-controlled medications and OTC comfort aids, were available.	No. “The 70 min it took to deliver felt like 7 years,” (Dr. Byock). Note: this case occurred in the early years of palliative care	Yes. Medications, including DEA-controlled medications and OTC comfort aids, were available.
18	What was the physician's intent/aim in implementing Moderate Anesthesia?	After low doses of sedatives failed to relieve her symptoms while preserving consciousness, high doses provided relief but caused unconscious. (Moral according to the Rule of Double Effect)	To render her unconscious in response to the patient's clear, capacitated request to end her unbearable pain and help her enter Heaven	To reduce the patient's conscious awareness of all sources of exceedingly severe suffering, including those sources of pain that were not currently observable by examining the patient
19	What was the cost of losing consciousness to relieve suffering?	Minimal: consciousness was overwhelmingly negative and provided little current benefit. Rendering her unconsciousness benefited the rest of her being and restored loved ones’ peace	Minimal: lost an unknown amount of time with family, but her unbearable pain made it impossible to enjoy them	Mixed: loss of occasional moments showing her loving personality. But her living will indicated she preferred loss of consciousness to relieve her suffering and burdening loved ones
20	Was rapid titration needed urgently to effectively relieve suffering?	Yes. Perhaps it could have been more rapid	Yes, definitely needed and accomplished as soon as possible (a few minutes)	No. Medication doses were increased parsimoniously and proportionately to the patient's observed suffering
21	Was informed consent obtained re: possibly dying earlier and not regaining consciousness before death?	Yes, executed by her proxy/agent; patient gave assent	Yes: Terry said she did not want to wake up and wanted to go to Heaven to which Dr. Byock agreed	Yes, using the almost identical consent for palliative sedation to unconsciousness that her husband, as her proxy/agent reaffirmed
22	Had the patient also clearly refused artificial nutrition and hydration (ANH)?	Yes	No. The decision to withhold medical (artificial) nutrition and hydration is generally not considered when dying is imminent (days)	Yes, explicitly refused ANH five years previously, when she completed her living will
C	DISTINGUISHING MODERATE ANESTHESIA FROM “SLOW EUTHANASIA”: MORAL ARGUMENTS
23	Can prescribing sedation be clearly distinguished from “slow euthanasia”?	Yes. Sedation was prescribed to relieve psychotic symptoms and agitated behavior —not to cause death, though death was a foreseen possibility	Uncertain. Sedation to relieve unbearable pain is not euthanasia, but agreeing to help the patient “enter Heaven” can be viewed as euthanasia	Yes. Sedation intended to reduce awareness of exceedingly severe suffering (including non-observable sources); refusal of life-sustaining treatment (ANH)—not sedation—was likely the cause of death
24	Could critics still object by arguing that the physician committed “slow euthanasia”?	Yes. The patient's plea was, “I want to die,” and the outcome was death	Yes. The patient's plea was, “I want to go to Heaven,” and she died in 30 hours	Yes. The patient's living will/surrogates’ substituted judgment clearly instructed to withhold nutrition and hydration if her suffering became intolerable . Sedation made eating/drinking impossible
25	What can the physician's counterargument be?	The Rule of Double Effect permits treatment if death is foreseen but not intended. Arguably, the patient's plea, “I want to die” reflects her desperate request for relief from continued living with intolerable suffering	The physician's role is to provide comfort if cure is not possible. Sedation can be an extraordinary act of compassionate end-of-life care. Most people prefer it to prolonged unbearable suffering—even if it hastens dying	Physician's intent was to reduce awareness of suffering—not to cause death. Sedation made eating/drinking impossible. Combined with her living will instructions for refusing ANH and her terminal diagnosis, sedation led to her unintended death by initiating a chain of events that overwhelmed the limits of the craft of medicine but physician's goal was attained when suffering was treated
D	PREVALENCE OF THE NEED FOR MODERATE ANESTHESIA
26	How commonly do terminally ill patients have intolerable suffering that requires MA?	Likely common and under-recognized. Agitated patients may receive sedation sufficient only to manage behavior—but not enough to relieve intolerable suffering—since sedation and advanced dementia can prevent patients from complaining about their ongoing distress	Rare. Modern palliative care has potent analgesics, though some patients’ suffering will remain refractory to less-sedating treatments	Extremely common yet neglected. Supplemental Material estimates 500,000 Americans are living with exceedingly severe suffering in the Dementia Gap, for whom Moderate Anesthesia may be their only effective treatment of last resort

Decision-making capacity: BB's agent gave legal consent on her behalf after BB gave her assent based on listening to me read aloud the descriptions of a dozen common, advanced dementia conditions. Her global preference to not prolong life was clear and consistent for all conditions; but she could not articulate her appreciation of the consequences of any specific condition or explain her reasoning.

AK met the clinical criteria for receiving MA based on combining her past judgments of severe suffering intensity for each condition memorialized in her living will, and the current unanimous consensus of substituted judgments of her PDC members who all agreed to this: If during ACP, AK had been asked if these six conditions would cause intolerable suffering, she would have judged yes. AK also consented to palliative sedation to unconsciousness (PSU)—a treatment of last resort with the same risks as MA, so her consent given during ACP sufficed for both.

Byock assumed Terry had sufficient DMC to give verbal consent for Rapid Sedation to Unconsciousness. While cognitively intact, her judgment might have been impaired due to being in a crisis state caused by unbearable pain and impending loss of family that caused depression. Consider this thought experiment: Had Byock given Terry enough IV Dilaudid to partially relieve her pain before asking if she were willing to consent to sedation that would begin in one hour, she might have declined. If so, she would have continued her previous priority and behavior: to remain with her family as long as possible.

Observability of Suffering: BB's psychiatric suffering manifested as observable agitation, depressive statements, and threatening behaviors. In contrast, AK's suffering included substantial non-observable dimensions: inner fear and sadness, burdening loved ones, and distress over leaving loved ones with tarnished and undignified memories of her. Terry's pain was clearly expressed: “I feel like I’m on fire.”

The Forced Existential Choice: All patients faced this choice: They can either retain consciousness that no longer offers meaningful experience but permits them to experience severe suffering, or they can eliminate consciousness to cease their awareness of suffering. The latter incurs the risks of permanent loss of communication and possible earlier death—sacrifices that could be deemed minimal compared to BB's depression, AK's cognitive impairment, or Terry's experience of pain.

Brittany Maynard's forced choice was based on anticipatory rather than current suffering. She could have continued living since consciousness still offered her meaningful and pleasant experiences. She stated unwillingness to risk future loss of capacity from progression of her brain disease that could prevent her from having sufficient capacity to make the decision to die later. So, she chose to exercise her current autonomy to die now, sacrificing weeks to months of somewhat satisfying living. To my knowledge, no one explained and then asked her to consider this third choice: Complete an effective strategic living will that would allow her to live until her condition deteriorated to a well-defined point where she judged her suffering would become intolerable.

For all four patients, the forced-choice perspective helps justify their treatment decisions as moral. In the case of Brittany Maynard, clinicians and bioethicists could reasonably ask whether the informed consent she was provided was adequate.

Is dying by VSED by capacitated patients moral or suicide? Is it moral for advance directives to conditionally request ceasing assisted feeding and hydrating, or sedation—be it Moderate Anesthesia or Rapid Sedation to Unconsciousness?

Note: Brief answers to these two questions were inspired by Farr Curlin's opinions embodied in the text of some of the arguments he presented during a “Point–Counterpoint Session on Dementia-Specific VSED Directives” held on 2026 Apr 30 at the American Geriatrics Society Annual Scientific Meeting. (Personal communication by email, May 29, 2026.)

In contrast to physicians’ obligation to respect and respond to capacitated autonomous decisions to withhold or withdraw burdensome treatment, Curlin argued that physicians remain responsible to care for their current patients through basic care of assisted feeding and hydrating, which he considered an “expression of ordinary human solidarity.” Curlin could well have then quoted Matthew 25:35 (which I confirmed in our email exchange).

Referring to VSED requests, Curlin stated living wills “that request actions fundamentally contrary to a patient's basic interests—especially those that are self-destructive in nature—do not bind clinicians or surrogates in the same way as refusals of medical treatment.” He particularly noted that requests that “amount to self-harm, in this case by starvation and dehydration, fall outside what a reasonable person standard would require others to carry out.” (Out of context, this opinion does not consider either current or future harm patients can experience due to longer and greater suffering.) Curlin concluded that even where VSED by advance directive is legally permitted, physicians and family members are not ethically obligated to implement it.

I favor comprehensively expanding the concept of suffering as long as the requested implemented interventions are moral.^4,7 My fundamental belief is that a primary societal function of physicians is to reduce suffering, a priority I strive to fulfill as I treat my patients and how I want to be treated myself. The imperative to reduce suffering can become relevant, compelling, and sometimes urgent when two conditions converge. First, an incapacitated patient may no longer understand that continued feeding and hydrating will prolong suffering as their terminal illness progresses. Second, a terminal illness causes suffering from multiple sources, some of which are non-observable, so that, when combined with observable suffering, patients can experience a suffering intensity that is severe, intolerable, or unbearable, even when the act itself—careful hand-feeding by another person—contributes no additional suffering.⁴

Curlin has revealed that he recognizes suffering can be severe enough to warrant palliative sedation to unconsciousness. When he wrote about Terry Matthews in 2018, ⁹ he stated: “I propose that physicians should act to relieve suffering insofar as that suffering is generated by and contributes to insults to health.” Terry died 30 hours after beginning sedation—in the gray zone that makes it difficult to decide whether sedation itself could be deemed the proximate cause of death.

On reflection, Curlin and I likely agree that the critical moral distinction is not whether death follows treatment, but whether physicians’ intent was to relieve suffering or to cause death. On that criterion, implementing a living will's instructions for a patient who can no longer understand that continued feeding will prolong and likely increase suffering—whether this patient lives in the Dementia Gap or endures currently undetected, Non-Observable Suffering-may reach an intensity similar to what Curlin endorsed as sedation-appropriate in the final clinical treatment of Terry Matthews’ terminal cancer.

Should ethics consultations be mandatory before implementing MA?

Capacitated patients can legally express their autonomy to VSED. In contrast, laws protect vulnerable, incapacitated patients when others make the same decision on their behalf. To protect them, Supreme Court Chief Justice Rehnquist warned, in his 1990 Cruzan ruling: “Close family members may have a strong feeling—a feeling not at all ignoble or unworthy, but not entirely disinterested, either—that they do not wish to witness the continuation of the life of a loved one which they regard as hopeless, meaningless, and even degrading. But there is no automatic assurance that the view of close family members will necessarily be the same as the patient's would have been had she been confronted with the prospect of her situation while competent.” ⁷⁸ This court established the cautionary precedent of “clear and convincing evidence” as the highest standard for surrogate decision-making in end-of-life care.

Similarly, vulnerable incapacitated patients (Type II and Type III), need protection that ethics consultations can provide.

Table 6 states the author's opinions that ethics consultations are mandatory for Type II and Type III patients; but optional for Type I and IV patients since they have capacity to give informed consent for themselves.

Comment on prevalence with respect to Gaster's “dementia-directive” ⁷⁹ : While Gaster's directive has likely been downloaded more than all others combined, it does not offer the option, Natural Dying/VSED by AD. This living will cannot be effective for many patients; therefore, Table 6 classifies patients using this living will as Type III.

Why I did not request an ethics consultation for AK

I consulted two board-certified palliative care physicians before I implemented the MA-POLST for AK, but I did not request a full, independent ethics consultation, which Table 6 states is “mandatory.” I mulled over my reasons for months, which I summarize and assess below.

I was still learning. AK was my first Dementia Gap patient. While her treatment can be deemed successful (see Table 5), I ultimately decided on the recommendation to mandate ethics consultations in the protocol. My assessment of this “developmental” excuse: inadequate.

I consulted with Katherine Hammond, the daughter/agent of Margot Bentley, who died in 2016 after enduring nearly four years of court-ordered, forced spoon-feeding. Many consider the order overrode her living will's statement of reaching a condition for which she wanted “No nourishment or liquids.” The court accepted the argument she limited her wish to the more commonly requested refusal of artificial nutrition and hydration. ⁸⁰ I speculate that her tragic end-of-life suffering might have been prevented had she added just one word to her living will, “oral,” between “No” and “nourishment.” (Bentley created her own living will before British Columbia or other sources provided guidelines and good examples that suggested adding “oral.”)

Hammond testified to Canadian Parliament that her mother was “forced to exist in the very state that she tried so hard to avoid…[that] her suffering prolonged by others.”^81–83 One judge ruled the behavior of opening her mouth and swallowing what others put in, as reflecting her remaining DMC that expressed her continuing desire to be fed. Comment: This is an astonishingly low standard for DMC; I personally know no bioethicist who agrees.

Bentley died from slow starvation. The illustrated patient decision aid ⁸⁴ that planning principals use to generate ND-Living Wills ⁸⁵ includes a horrific photo after her long starvation, where she resembles a Holocaust victim. Force feeding likely caused suffering that included hunger, wasting with weakness, and a multitude of N-O-S, many of which will never be fully known—that in combination may have been a torturous experience.

Summary assessment: The lesson is clear: any authority, including a well-intentioned ethics consultant, could become an unintentional barrier to the timely relief of suffering. AK's husband and I welcomed ethics consultations, but we opposed placing AK in a situation where authorities could wield their power—whether intentionally or not—to prolong her suffering and dying. I take ultimate responsibility for deciding not to request a formal ethics consultation. Going forward, I consider such consultations mandatory, as long as new guidelines can give confidence they will not become barriers to the timely relief of suffering.

A challenge: how to motivate Type IV patients to upgrade?

Type IV patients represent a new generation of planning principals: they have already completed a “Natural Dying-Only” (“VSED by AD”-only) Living Will, unaware that it will be ineffective if they reach the Dementia Gap. This applies to all patients who completed ACP with Caring Advocates prior to 2023 (like AK), plus many more who completed living wills drafted by others: perhaps tens of thousands need to upgrade. Motivating them could be challenging: it requires education that alerts them to the additional challenge presented by the Dementia Gap at a time when they may not be suffering and have lived thinking they already completed ACP. Since they will lose the opportunity to upgrade when they lose DMC, finding ways to motivate them is a research imperative.

Why this article deferred considering PINEs enduring “severe” suffering by first focusing on the smaller number enduring “intolerable” suffering

It is easier to justify a needed treatment of last resort for intolerable suffering than for “only” severe suffering, so others might more readily accept such treatments in two steps.

Summary: Beginning the discourse with 500,000 is large enough to get attention. After considering the five points above, professionals may be more likely to accept a larger number. The Supplemental Material discusses why “500,000” may be viewed both as large and as small.

Comparing two protocols to judge intensity of suffering: single physician assessor versus three sources

In Quebec, Advance MAiD Requests (AMRs) recently became legal for patients diagnosed with neurocognitive disease. ⁸⁶ To implement MAiD, the patient's current suffering must be judged intolerable. The question is, how to judge suffering intensity in inarticulate, incapacitated patients?

In 2026, Holland and Moreau (H&M) presented what I call a single “clinical assessor” protocol. ⁸⁷ Example for a patient receiving incontinence care: If she “grimaces and cries while being assisted with toileting while appearing content at other times, the clinician can infer they are suffering” and execute MAiD; if not, “the clinician may infer that the person has accepted this loss of independence” and not execute MAiD. This protocol has potential errors of observation, interpretation, and inadequate information about values. Current Quebec forms do not capture values information; video is not routinely used to memorialize patients’ expressed values; and assessors are not asked to consider Non-Observable Suffering or to observe patients over several hours. H&M themselves identify the core vulnerability: “By requiring the assessor to determine the presence of unbearable suffering, the assessor's values may be introduced into the decision-making process.”

Terman's Three-Source Concordance protocol addresses these limitations. Building on a 2024 “two-task, two-person protocol performed at two times,” ⁴ it adds PDC members’ consensus of substituted judgment to the treating physician's final determination, providing confidence for all concerned.

Source 1: the patient's own voice. During ACP, patients judge how much suffering each condition would cause as trained ACP counselors encourage capacitated patients to elaborate on their lifelong values and treatment preferences, to explain why, using a 36-condition Illustrated Patient Decision Aid (IPDA). Each condition is depicted as a high-resolution color image revealing inner emotions through facial expression, body language, and thought bubbles. For the incontinence condition, the patient grimaces and thinks, “I feel so ashamed.” Yet the introductory narrative asks patients to also consider whether joy can be experienced at other times during the day. If so, suffering during incontinence care may be relatively brief and therefore tolerable. Interviews are memorialized on video and transcribed. Obtaining direct information from capacitated patients is more credible than inferring patients’ values after they have lost capacity.

Source 2: the Patient Decision Committee. Three or more surrogate decision-makers, Trusted Persons, chosen for their deep knowledge of the patient, are polled during a clinician-facilitated discussion. They contribute personal knowledge of the patient's lifelong values, sources of joy, and past experiences with dying—information treating clinicians infrequently possess. They also alert the clinician when a qualifying condition may have been reached, and confirm through substituted judgment that disease progression has brought the patient to a condition previously judged intolerable during ACP.

Source 3: the treating physician's current assessment and review. Physicians consider all available information—their own clinical assessment plus the patient's living will responses plus the PDC polling of substituted judgment, local statutes, and generally accepted medical practice. Their function therefore extends beyond merely matching current conditions to AMR conditions. Physicians are ultimately responsible for determining the clinical appropriateness of deciding that implementing a treatment of last resort is patient-centered rather than self-serving.

Three-Source Concordance relieves physicians from bearing sole responsibility for judging suffering intensity in incapacitated patients. In making the ultimate existential decision, it considers the patient's own advance decisions and PDC members’ input. The result can be confidence in implementing MAiD or a Future POLST, including one for Moderate Anesthesia. Its motivating principle is the operational respect of patient autonomy in the context of a humanitarian crisis: every patient in the Dementia Gap enduring severe chronic suffering without access to an effective treatment of last resort represents a preventable failure of end-of-life care.

The “just sitting there” fallacy: final comments

Many dementia patients appear to be “just sitting there”—not suffering but withdrawn, unable to enjoy meaningful interaction. Observers assume they cannot recognize relatives, chant religious hymns, or “dig” jazz. This assumption could be wrong; if so, it is dangerous. Dismissing patients as incapable of human connection without researching their personal interests and using advanced interviewing skills can deny their fundamental humanity, create self-fulfilling prophecies that allow severe suffering to continue undetected until death, deceive caregivers into believing apparent calm equals inner peace, and deprive patients of goal-concordant end-of-life care.

The Dementia Gap crisis combined with N-O-S

As dementia progresses, an increasing percentage of patients will experience Non-Observable Suffering they cannot articulate. For those living in the Dementia Gap—able to eat and drink independently despite profound cognitive impairment—Moderate Anesthesia could be their only effective treatment of last resort. It is essential to distinguish patients who have potential to interact meaningfully if successfully awakened from withdrawal, from those who need sedation to relieve suffering. Progress cannot begin until clinicians stop assuming: “She's not suffering; she's just sitting there.”

This article argues that Moderate Anesthesia can be clinically appropriate, legal, ethical, and moral. It takes this bold stand: prognosis or life-expectancy is irrelevant. A trial of Respite Sedation and the use of proportional sedation that is not intended to hasten death may establish a sound basis to view MA as moral. Widespread acceptance could prevent prolonged suffering for an estimated 500,000 Americans in the Dementia Gap; extend lives by reducing premature deaths driven by “Dementia Fear” and “Loss of Autonomy Fear”; reduce attempts to die at “five minutes to midnight”; enable appropriate treatment for middle-stage patients rather than viewing illness as an “opportunity” to avoid late-stage dementia; provide legal, moral alternatives for families considering mercy killing; and stop wasting scarce resources on treatments that patients would refuse and cannot appreciate.

A proposal: begin the ARISE movement this humanitarian crisis demands

ARISE—Advocate by Recognizing and Investigating Suffering's (hidden) Existence—can be the rallying framework to recognize the humanitarian crisis this article calls the “Dementia Gap.” It aims to inspire Advocates to attend to individual patients’ needs and beyond: to investigate hidden suffering broadly and pursue the systemic changes this crisis demands. This can include expanding end-of-life curricula, developing skilled interviewing techniques, supporting effective PDC functioning, and ensuring that 500,000 to 1,000,000 patients in the Dementia Gap are not abandoned to suffering that goes undetected, unnamed, and untreated. Perhaps a “Registry for Dementia Gap Experiences” (RIDGE) will be established to transform individual caregiving experiences into collective wisdom. Family members and clinicians can share their experiences, adding to the clinical data that supports many more patients than we appreciate are actually “alive inside.”

Researchers must address critical unanswered questions: How long do patients remain in the Dementia Gap? What predicts loss of independent eating? How prevalent is N-O-S, and how satisfied are survivors after MA? Can assessment scales make N-O-S observable? Policymakers and professional organizations must convene expert panels to develop consensus guidelines with safeguards preventing premature sedation while ensuring severe suffering is not prolonged.