Loading today’s stories
Science1 distinct publisher2 min readPublished
The state's new aid-in-dying law makes patients record their own request. A STAT essay argues the burden it cannot document starts earlier, when a doctor first names death as an option.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
The tape proves one thing well: that the words came from the patient [2][3]. It says nothing about who introduced the subject, and the state has written no rule about that [8]. The essay's argument sits in exactly that gap. Its claim, borrowed from the philosopher David Velleman, is that an option can leave someone worse off even if he declines it, and even if he takes it and benefits [4]. If that holds, the injury occurs when death is first named as something a physician can arrange [10], and the required recording is made after the fact, by a patient who has already crossed the line the argument cares about.
That structure is awkward for anyone trying to evaluate the program. A recording is evidence that consent was not extracted, which is what proponents want it for [3]. Velleman's harm is compatible with consent that is entirely genuine and with a patient who is glad he asked [4]. Uptake counts and satisfaction surveys would both come back clean on a case where, by the essay's own account, the patient would have been better off never having faced the question [11]. There is no field on the form for that.
So initiation policy is the only real lever, and New York has left it to institutions [8]. Some hospitals tell patients outright that they must start the conversation themselves [9]. That does shield the patient who does not want the question put to him. It also means the option reaches, in practice, only people who already know the law exists and feel entitled to raise it with a doctor [2]. Burden and access run in opposite directions here, and no single institutional rule sets both: a hospital that raises aid in dying routinely spreads the information evenly and imposes the choice on everyone, while a hospital that waits imposes nothing and spreads the information unevenly.
Worth being plain about the evidence. This is a philosophical argument in an opinion column, not a measurement [4][11]. Nobody has counted how many terminally ill New Yorkers now experience staying alive as a decision that requires reasons, and the essay does not pretend to have done so. Velleman's formulation of the question is still the sharpest thing in it, the prospect of having to answer why you have once again declined to kill yourself [7]. New York's reply, so far, is a recorded request plus whatever each hospital decides about who speaks first [2][9].
Ranked by verification strength, evidence, and original report placement.
The philosopher David Velleman argued decades ago that simply having an option can be harmful, and that the harm can occur even if the option is never exercised, and even if it is exercised to the person's benefit.
Velleman's illustration: a teacher offers free tutoring to a quiet student who is in fact doing well; the offer makes him doubt himself, and he may end up needing the help, benefiting from accepting it while being worse off than if the offer had never been made.
Some hospitals explicitly tell patients that they must initiate the conversation about medical aid in dying, which avoids imposing the choice on every terminally ill patient.
Earlier in the month before publication, New York made medical aid in dying available to terminally ill New Yorkers with a prognosis of less than six months to live.
Unlike most states, New York requires that patients record their oral request on audio or video in order to receive life-ending medication.
Proponents of medical aid in dying argue that, absent undue influence, giving terminally ill people the choice empowers them, and that the recording simply helps prove the choice was really the patient's.
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Single signed essay; statutory facts uncorroborated, harm argument conceptual
All material rests on one opinion piece by a named bioethics academic. Its statutory claims (availability under a six-month prognosis, the recording requirement, the absence of a uniform rule on raising the option) are specific and internally consistent but carry no statutory citation and no second source. The central harm argument is philosophical, resting on an uncited Velleman thesis and a tutoring analogy rather than any patient-level evidence.
Statute live statewide; uptake and institutional coverage undocumented
There is real-world availability: the law is in force in New York and at least some hospitals have written patient-initiation policies. But the cluster supplies no request or prescription counts, no named institutions, and no share of health systems with any policy at all, so practical uptake is essentially unmeasured beyond legal availability.
Modestly overstated: asserted patient burden outruns any measurement
The essay is explicitly hedged — it disclaims being an argument against aid in dying and frames the problem as a design tradeoff — which keeps the gap small. But it generalizes from a tutoring analogy to the lived experience of terminally ill New Yorkers, and asserts that some patients may be harmed by being offered the choice, with no patient data, clinician account, or study in the cluster. Directionally the framing runs ahead of the evidence base by a modest margin.
Signed academic opinion; disciplinary stake, no disclosed financial interest
The piece is labeled opinion and bylined by Monika Piotrowska, an associate professor of philosophy at the University at Albany, SUNY working on bioethics — a professional stake in advancing a philosophical framing of the debate, and a publisher incentive to run provocative commentary on a live state law. No commercial, provider, or advocacy funding relationship is disclosed or evident, and no product or organization stands to benefit from the argument, so distortion pressure is moderate rather than high.
Coherent but uncorroborated single-source read
Confidence is limited by cluster structure: one publisher, one source, no independent confirmation of the statutory or institutional facts and no counter-voices. The factual and derived claims are consistent and the author is credentialed in the relevant field, which supports moderate confidence in the descriptive claims; confidence in the asserted patient harm is low because nothing measures it.
science
A vaccine communicator's testable claim: hesitancy answers to burden, not to better facts1 distinct publisher
science
Four years of Dobbs data: one prediction in four survived contact with the numbers1 distinct publisher
science
Stewardship is aimed at the wrong half: rich countries use the resistance drugs, LMICs need them1 distinct publisher
science
A cancer charity sunsets its liquor gala; the courts are less consistent1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 26, 2026