Who decides when the patient can’t

There is an order to it: the patient’s own voice first, then someone speaking as the patient would, and only last a judgement of what is simply best.

The idea

When a patient can no longer make a decision, authority does not jump to “the family” or “the doctor.” It follows a cascade. First, the patient’s own known wishes — a valid advance directive is the patient still speaking. Next, a surrogate (a named proxy, or the default hierarchy) using substituted judgment: deciding as the patient would, not as the surrogate prefers. Only when the patient’s wishes are truly unknown do you fall to the best-interests standard.

Two things trip people up: a proxy only activates once the patient lacks capacity, and when a proxy’s wish clashes with a clear directive, the patient’s own documented voice is the anchor — not the loudest person in the room.

How it works

Walk the cascade top to bottom and stop at the first voice that applies:

does the patient have capacity now?
  YES -> the patient decides.  (directives + proxies stay dormant)
  NO  -> keep going

is there a valid directive that fits this decision?
  YES -> follow it.  it is the patient's own voice
  NO  -> keep going

is there a named proxy / agent?     -> proxy, by substituted judgment
  else a default surrogate?          -> surrogate, by substituted judgment
  else no one who knows their wishes -> best interests

The two standards are not interchangeable. Substituted judgment asks “what would this patient have chosen?” and needs evidence of their values. Best interests asks “what would benefit a reasonable patient here?” and is the fallback only when the first cannot be answered.

When a directive and a proxy clash: a clear, applicable directive is the patient speaking directly, so it anchors the decision. If a proxy pushes the other way, pause — check whether the directive really fits the situation, and involve an ethics consult or legal review rather than simply deferring to the agent.

When to use it

Watch out for

• Skipping the capacity question. Reaching for a proxy while the patient can still decide. The proxy only activates once capacity is lost — and capacity can fluctuate.
• Substituting the surrogate’s values for the patient’s. “I couldn’t let my mother go” is not substituted judgment. The question is always what the patient would have wanted.
• Letting a proxy override a clear directive by force of insistence. A valid, applicable directive is the patient’s own voice — treat a contradiction as a reason to pause and seek review, not to defer to whoever is present.
• Treating “the family” as one voice. When equal-rank surrogates disagree there is no automatic winner; return to what the patient would want and seek consensus, an ethics consult, or, if needed, a court-appointed guardian.
• Jumping to best interests too early. It is the last resort, used only when the patient’s wishes truly cannot be known — not a shortcut past a hard conversation.

Worked example

An interviewer asks: “An unconscious patient has a living will declining a ventilator in this exact situation. Their son, who is the named health-care agent, is begging you to intubate. What do you do?” A strong answer walks the cascade aloud: the patient lacks capacity, so we don’t defer to a bedside preference by default; there is a valid directive that clearly addresses this situation, and that directive is the patient’s own voice. The son’s wish, however understandable, is not substituted judgment if it contradicts what the patient documented. So I’d pause, confirm the directive genuinely applies, sit with the son to explain that honoring his mother’s wishes is respecting her, and bring in the ethics service — rather than simply intubating because the agent asked. Naming the standard, and the anchor, is the point being tested.