His dimly lit room is quiet except for the hum of the monitor. “Mr. L” is alone, lying flat on his back in a slightly inclined bed, eyes squeezed shut, hands raised in the air as immobile fists. He has not spoken in weeks. He has not moved much voluntarily, aside from the hand positions that are reminiscent of his passion for boxing. The medical team calls it catatonia, a severe psychiatric syndrome that has rendered him mute, immobile, and medically fragile. The psychiatric care team has recommended comfort measures, and our hospice team is ready to accept him.

But Mr. L cannot leave the hospital and cannot receive full comfort care measures. Not because he is too sick. Not because his family objects, as he has no family here. Mr. L cannot access hospice because an overworked stranger appointed by the court has not yet returned the hospital’s calls. This is a guardian, a professional fiduciary who has never met Mr. L, who must sign the papers first. So we wait for the court to hold a hearing to approve the order. And Mr. L, still, silent, and suffering, waits with us.

We came to Mr. L’s room from different roles. Dr. Kirsten Engel is an emergency and palliative care physician who cares for patients near the end of life. I was observing in the supportive care unit as a graduate student in Harvard Medical School’s Media, Medicine, and Health program, studying how stories, illnesses, systems, and medical decisions collide. What we saw was not a failure of our compassion, but a failure of legal infrastructure.

Across the United States, patients who cannot make their own medical decisions are trapped in legal machinery that delays or prevents access to hospice care in terminal illness. The machinery is called guardianship, and it is in place to support vulnerable people. It can protect patients with serious mental illness, intellectual disability, homelessness, estranged families, social isolation, and no legally recognized surrogate from a harmful decision being made while they cannot consciously participate. But for patients like Mr. L, the process can take days or weeks, and we felt how protection becomes delay.

Even when everyone involved is acting in good intentions, the stranger-guardian relationship can feel dehumanizing. Mr. L’s guardian is a kind person, I am sure. But they probably do not know what makes Mr. L who he is. Endless questions even plagued my mind as I met him in his late stage of life. Does he have loved ones somewhere? What did boxing mean to him? Did he want to die in a hospital? Does he have anything he would have wanted to share on his deathbed if he were able? Surely he did, as we are all full of trapped stories much richer than what a chart reads. The legally appointed guardian cannot be expected to know these things, and they are not at fault. A flawed structure at the systemic level spreads them far too thin, so we linger as all they can do is read the medical chart and follow the law in the order and pressed timing they are able.

An estimated 5 to 10 percent of hospitalized adults lack both decision-making capacity and an available surrogate. For psychiatric patients like Mr. L, who are more likely to be socially isolated or estranged from family, the numbers are even higher. The same is true for people with intellectual disabilities, and for LGBT individuals who may lack legally recognized next-of-kin. These patients are invisible in the system until they are dying. And then, suddenly at the end of their life, the legal proceedings demand a new decision-maker. But the decision-maker it provides is a stranger armed with a court order and a liability concern, not a lifetime of shared history that could discern what a patient’s desires would be.

We would never ask a stranger to choose a patient’s career, or spouse, or faith. But we do ask strangers to choose how a patient will die.

The counterargument is logical. Guardianship exists to protect vulnerable patients from premature withdrawal of care. Psychiatric patients are at risk of being “written off” by clinicians who harbor stigma about mental illness, so the concern is legitimate. The literature demonstrates that people with severe mental illness face diagnostic overshadowing and are less likely to receive palliative care consultations.

But the current system does not solve this problem; it compounds it. Protection cannot mean prolonged suffering, as that is inherently paradoxical. For a patient dying of multi-organ failure, a second court hearing should not be the rate-limiting step between a dying patient and relief from pain. The system overcorrects by treating every end-of-life decision as a potential abuse, every request for hospice care as suspect. And so it builds walls that trap patients in a purgatory between curative care they do not want and comfort care they cannot access.

Most state laws are silent on whether guardians can authorize withdrawal of life-sustaining treatment or hospice care. Some states require additional court hearings, while others demand “clear and convincing evidence” of the patient’s prior wishes, which is an impossible standard for patients who never articulated preferences. In Massachusetts, for instance, guardianship law does not allow end-of-life care decisions to be made easily; the guardian must go back to court. Studies of guardianship and hospice transitions have documented the same pattern: patients waiting weeks for approval, clinicians struggling to reach guardians, and burdensome treatments continuing while legal permission catches up to clinical reality.

Meanwhile, the patient waits in limbo. To be sure, no one intends for a patient to suffer at the end of life, but this structural wait time is enormous. We have seen patients remain on ventilators they do not need, and receive dialysis that prolongs dying without restoring function. Those who have closely worked in hospice like us regard this current navigation as hardship for the patients, for the loved ones when they do not have legal autonomy, and for the clinicians who watch their patients die in slow motion.

What must change is specific and actionable. State legislatures should give trained and vetted guardians explicit authority to authorize hospice enrollment and comfort-focused care without automatically requiring a second court hearing. When judicial review is necessary, courts should use an expedited pathway for terminally ill patients. Hospitals, especially those without ethics consult teams, need clear protocols so access to comfort does not depend on whether a patient happens to be dying in a well-resourced institution.

Some states have already begun to move in this direction. Only five states currently permit guardians to make end-of-life decisions independently, while the majority of state statutes offer no guidance at all. In 2020, the New Hampshire Supreme Court ruled in In re Guardianship of L.N. that court approval is not required for a guardian to withdraw life support, calling judicial involvement “neither necessary nor warranted” when all parties agree. Minnesota’s Supreme Court reached a similar conclusion, citing the risk of “extended suffering” if guardians were required to seek judicial review for every such decision. These courts recognized what many statutes still do not: Delay can be its own form of harm.

We watched Mr. L die days after the guardian was appointed. He was never able to enroll in hospice or leave the hospital for the comfort-focused care our team believed he needed. The guardian, after consulting his attorney, requested a court hearing to authorize comfort care. The hearing was scheduled for the following week. But Mr. L died before it could take place, still in the hospital, still on advanced curative medications, still waiting for permission to be comfortable at home.

I think about him often. I think about the silence in his empty room, the rigidity of his body, the way his fists were up in the air, silently fighting the way his treatment was ensuing. I think about the fact that we knew what he needed and were stuck in being able to provide it.

The law failed Mr. L, and it fails patients like him every day.

Medha Tripathi is a health writer and incoming Fulbright Research Scholar whose work examines psychiatry, global mental health, medical anthropology, and public-facing health communication. She earned her B.S. in neuroscience and biophysics from the University of Michigan and her M.S. in media, medicine, and health from Harvard Medical School, where she is affiliated with the Department of Global Health and Social Medicine.

Her research has included cardiovascular biomarkers, culturally adapted mental health interventions, caregiver-mediated autism care, ethical uses of AI in psychiatry, and health equity. As a 2026 to 2027 Fulbright Research Scholar in Munich, Germany, she will study machine learning-based relapse prediction in psychosis at Ludwig Maximilian University of Munich.

Through research and storytelling, she hopes to make complex health issues feel less distant from the people living through them. She shares updates on LinkedIn.

Kirsten Engel is a palliative medicine physician.