The Two Halves of a Health Care Directive

Most people who sit down to make one of these spend their time on the instructions and very little on the appointment. The proportions should be reversed, because it is the appointment that will actually be used.

The appointment

It names a decision-maker. Somebody authorized to make healthcare decisions when the maker cannot make or communicate them themselves.

It covers everything. Including situations nobody anticipated, which is why it is more useful than any list of instructions could be.

It should name alternates. Since a first choice may be unavailable, unwell or the person in the ambulance rather than the one taking the call.

It should be discussed first. With the person appointed, who is entitled to know they have been named and what the maker would want.

And it takes effect only on incapacity. So the maker continues to decide for themselves for as long as they are able to.

The instructions

They record preferences. About treatment in defined circumstances, most commonly at the end of life or in states of permanent unconsciousness.

They are useful as evidence of values. Even where they do not fit the situation exactly, since they tell an agent what the maker cared about.

They rarely fit precisely. Because clinical reality is more varied than any drafted scenario, which is the central limitation of this half.

They can be specific or general. And a statement of values is often more useful to an agent than a checklist of interventions.

And they should address what matters. Feeding, resuscitation, ventilation, hospitalization from a facility, and antibiotics in a late-stage illness.

How an agent should decide

As the maker would have decided. Which is the standard, and is why the agent needs to know the maker's values rather than simply their own view.

Following clear instructions where they apply. Since a directive that squarely covers the situation should be followed rather than second-guessed.

In the maker's best interests otherwise. Where nothing indicates what they would have wanted, which is the fallback rather than the primary standard.

With the clinical information in front of them. Which means asking what the realistic outcomes are rather than deciding on an abstract preference.

And without deferring to the loudest relative. Since the authority is the agent's, on the duties described in what an agent owes.

Making it work in practice

Give copies to the right people. The agent, the alternates, the physician and the hospital, since a document in a drawer helps nobody.

Have the conversation. Which is worth more than the document, because an agent who has discussed this can decide confidently under pressure.

Tell the wider family. So that the appointment is not discovered during a crisis by relatives who disagree with it.

Review it periodically. Since views change, agents move away, and a document from twenty years ago may name somebody unsuitable.

And pair it with a financial document. Because the two cover different things, on the split described in what a power of attorney does.

What it is not

Not a medical order. Which is a different instrument signed by a clinician, examined in a directive and a medical order.

Not effective while the maker can decide. Since capacity to make a particular decision displaces the document entirely.

Not a refusal of care. A directive can ask for treatment as easily as decline it, and many people use it to insist on intervention.

Not binding on the impossible. Clinicians are not obliged to provide treatment that would not work, whatever a document requests.

And not permanent. Since it may be revoked or replaced at any time while capacity lasts.

The instinct to spend the effort on instructions is understandable and misplaced. Instructions address a small number of foreseeable situations; the appointment addresses everything else, which is where the decisions actually arise.

Choosing the right agent is therefore the central decision. It should be somebody who can be reached, who can hold a position under pressure from other relatives, and who knows what the maker would want rather than what they themselves would prefer.

The conversation with that person is the highest-value part of the whole exercise. It costs an hour, it is uncomfortable, and it is what allows somebody to decide well at three in the morning.

Telling the wider family who has been appointed prevents a category of conflict that otherwise plays out at a bedside. It is easier said once, calmly, than explained during a crisis.

Copies matter more than people expect. A directive nobody can find is a directive that does not exist, and hospitals act on what is in front of them.

And the document should be revisited every few years. Agents move, relationships change, and views about intervention shift considerably between sixty and eighty-five.

It is worth saying something about the discomfort of this subject, because it is the reason so many people never complete these documents at all. Nobody enjoys deciding who should make decisions for them, or thinking about the circumstances in which somebody would need to.

What usually helps is reframing what is being decided. The question is not whether these situations will arise but whether, when they do, the family will be making decisions with guidance or without it. Everybody who has been on the wrong side of that says the same thing afterward, which is that they would have given a great deal for an hour's conversation beforehand.

For adult children raising it with a parent, the useful framing is the same. This is not about anticipating decline; it is about making sure that if something happens, the person deciding is the one the parent would have chosen, deciding what the parent would have wanted. Put that way, the conversation is generally shorter and easier than anybody expects. Most parents have thought about it already, and a good many are relieved that somebody has finally raised it out loud.