When Physicians Opt Out of Medical Aid in Dying, Who Is Left to Say Yes?

 

When Physicians Opt Out of Medical Aid in Dying, Who Is Left to Say Yes?

A Stanford study finds most oncologists asked to participate in MAID decline — leaving a small group of physicians carrying most of the caseload, and many patients navigating end-of-life care with a stranger.

Source: Ward-Lev E, Cvitanovic M, Tabor HK. Physician opt-out and patient pathways to medical aid in dying at an academic medical center. JAMA Network Open. Published August 19, 2026.

Medical aid in dying (MAID) sits at the intersection of law, ethics, and clinical relationship — and a new retrospective study out of Stanford University puts hard numbers behind something many end-of-life clinicians have long suspected: physicians asked to participate frequently decline, and the resulting caseload concentrates on a small number of willing prescribers.

California legalized MAID in 2016. To qualify, a terminally ill patient must make two requests to an attending physician and one to a consulting physician; physicians and institutions may opt out entirely or evaluate requests case by case. Researchers used electronic health records and ethics consultation data from Stanford Health Care to build a case series of 130 patients who made a first MAID request between June 2016 and June 2021, whether or not they ultimately completed the process.

By the Numbers

Metric	Finding
Total patients in study	130 (California, June 2016–June 2021)
Patients with a cancer diagnosis	115 (88.5%)
Cases with a written MAID prescription	85 (65.4%)
Patients who ultimately took the medication	45 (34.6%)
Attending physicians who were oncologists	17 of 49 (34.7%)
Physicians who opted out overall	16, of whom 12 (75%) were oncologists
Cancer patients whose primary oncologist served as attending	52 of 115 (45.2%)
Cancer patients who pursued MAID with an unfamiliar physician	48 (41.7%)

 

The overwhelming majority of patients in the study — 115 of 130 (88.5%) — had a cancer diagnosis; 10 (7.7%) had neurologic disease and 5 (3.8%) had cardiopulmonary disease. Of 49 attending physicians involved, oncologists made up the largest single specialty at 34.7%, followed by primary care (32.7%), neurology (16.3%), and palliative care/hospice (8.2%).

Caseload distribution was sharply uneven. Eighteen Stanford physicians (46.2%) served as attending for only a single patient, while just 4 physicians (10.3%) accounted for 49 patients (37.8%) of the total caseload. One oncologist alone served as attending for 24 patients — half of whom had no prior relationship with him.

Why Physicians Opted Out

Sixteen physicians in the study opted out of participating in MAID, and 12 of them (75%) were oncologists — despite oncologists also representing the largest attending specialty overall. Three primary care physicians and one palliative care/hospice specialist also opted out; no neurologists or physicians in other specialties did.

Documented reasoning was available for 24 cases, 23 of which (95.8%) involved cancer patients. The most common reason, cited by 9 physicians (53.6%), was simply not endorsing or not wishing to participate in MAID. Three physicians (18.8%) believed the patient would likely live longer than the 6-month prognosis threshold the law requires. One cited being new to the patient relationship, and one cited ethical concerns directly.

The Downstream Effect on Patients

Less than half of cancer patients in the study — 52 of 115 (45.2%) — had their own primary oncologist serve as MAID attending. The remaining 48 patients with cancer (41.7%) pursued the process with a physician who had no prior relationship with them at all. The study authors were direct about what that means in practice: high opt-out rates force patients to locate a new attending, potentially delaying access at a point when they are already seriously ill, and the resulting relationship may lack the longitudinal context that typically supports end-of-life decision-making.

The researchers also flagged a structural risk in the current system: concentrating prescribing responsibility on a small number of high-volume physicians — often without additional compensation — creates a setup where the departure or unavailability of even one or two of those clinicians could meaningfully disrupt patient access. The authors noted their dataset likely understates the scope of the problem, since it only captures patients who ultimately found an attending and may miss opt-outs that went undocumented.

 

■ CLINICAL BOTTOM LINE •     In jurisdictions where MAID is legal, know your organization's opt-out policy and referral pathway before a patient raises the request — not in the moment. •     A physician's choice to opt out is protected and does not need to be defended to the patient, but a timely, warm handoff to another attending materially affects how much delay and burden the patient experiences. •     Patients whose own oncologist or primary clinician declines to participate may end up navigating an intensely personal decision with someone who doesn't know their history — anticipate that gap and document goals-of-care context thoroughly enough that it can travel with the patient. •     Be aware that MAID caseloads in many systems concentrate on a handful of willing clinicians; if you are one of them, this study is a reminder to advocate for institutional support and compensation structures, not to absorb the burden silently.

 

■ CASE FROM PRACTICE An NP working alongside an oncology team is following a patient with stage IV pancreatic cancer who raises the topic of MAID during a routine visit. Her longtime oncologist, who has managed her care for two years, is not willing to serve as attending physician for the process and refers her to the health system's ethics/MAID coordination line. The patient is connected with an attending she has never met. The study's findings suggest this scenario is common rather than exceptional — and that the NP's role in this moment (clear communication about the referral process, thorough transfer of clinical and personal context to the new attending, and continued emotional support through the transition) may matter as much to the patient's experience as the outcome of the MAID request itself.

 

■ A NUANCE WORTH FLAGGING This is a single-institution, retrospective study from one academic medical center in one state, covering a 5-year window that ended in 2021 — findings may not generalize to other health systems, states, or the current regulatory landscape. MAID itself remains a legally and ethically contested area of practice: participation is a matter of individual clinician conscience in every state where it's legal, and the study authors' framing of opt-out as a systemic “challenge” reflects their own analysis rather than a settled clinical consensus. Reasonable clinicians, professional bodies, and patients hold genuinely different views on MAID itself; this research speaks to system-level access and continuity of care, not to the underlying ethical question of whether or how MAID should be practiced.

 

What This Means for Practice

For NPs working in oncology, primary care, palliative care, or any setting where patients may raise end-of-life options, the practical takeaway isn't about whether to personally participate in MAID — that remains an individual and often deeply held decision. It's about making sure your health system has a clear, documented, and reasonably fast pathway for patients whose own clinician opts out, so that a legal right doesn't become an access barrier layered on top of an already difficult diagnosis.

References

Ward-Lev E, Cvitanovic M, Tabor HK. Physician opt-out and patient pathways to medical aid in dying at an academic medical center. JAMA Network Open. Published online August 19, 2026. doi:10.1001/jamanetworkopen.2026.29872

This post is intended for clinical education purposes and does not represent an institutional or editorial position on medical aid in dying. It does not replace individualized clinical or ethical judgment.