When I arrived to supervise care in the intensive care unit for the night several years ago, Fran, a third-year medical resident from Oregon, explained to me that “Room 506 is Do Not Intubate/Chest compressions okay.”

She noted that fact with a scowl. “What a stupid code status,” she complained, shaking her head in disbelief. “We can’t intubate her, but if she arrests we can shock her and break her ribs and then leave her to die without an endotracheal tube? Stupid.”

Fran was right: The code status didn’t make much sense. But the stupidity belonged to the medical system, not the patient in 506, a woman in her mid-60s named Belle. The system had set up both Fran and Belle to fail. That apparently incoherent code status made painfully clear how much we need integrated approaches to medical care when life may be threatened, approaches that are true to us as people and to the arc of our lives.

Fran and I went together to speak with Belle. She had struggled a bit with her diabetes but had been mostly healthy before her current episode of pneumonia. She had claustropho­bia, though, and hated the thought of the breathing tube, which she remembered gagging her when she came out of anesthesia after gallbladder surgery. She said that waking up after surgery was “a total nightmare, choking to death on that tube and I couldn’t say a word.” As we explored her wishes together, Belle made it clear that she also never wanted to end up in a permanent vegetative state, “like poor Terri Schiavo.”

But, Belle said as we discussed her current medical situation, as long as she had a decent shot at recovery she would want aggressive medical care, including intubation and CPR with chest compressions. She just wanted help with her claustrophobia if it came to that. These worries are what gave rise to the code status that Fran had reported to me. Somehow, communication had failed, but why?

Most of us want to have control over events when our health worsens or life seems short. As a society, we worry about the high cost of treatments with little chance of success in patients near the end of life, treatments disproportionately provided in the ICU. Many have hoped that living wills would provide such control, but those documents bounce between too much detail and too little, depend on unreliable predictions, and poorly represent our actual values and aspirations. Traditional living wills list a few specific medical procedures to refuse or accept in specific, rare circumstances, like permanent vegetative state. But this is a bit like the auto mechanic asking a car owner which wrench to use if the fuel injector should ever happen to crack on a cold winter morning in Florida.

My group at the Center for Humanizing Critical Care has developed a more targeted approach to move beyond the confusion and frustration of traditional living wills and code status. We call it the Five Approaches and they run from “all heroic treatments” to “just make me comfortable,” with three levels between those extremes that better match the ways people actually experience life-threatening illness.

1: Do Everything

For decades “do everything” has been the best-known response to life- threatening illness. It is still the strategy most consistent with many Americans’ sensibilities about terminal illness. For oth­erwise healthy people, this may be appropriate.

Unfortunately, in many cases this default setting may not be the right response. It has probably been overused in individuals with advanced terminal diseases and at the extremes of age. Honest, open discussion is crucial, even when something like Approach 1 is chosen.

2: Be Aggressive Only if I Have Reasonable Chances

If we look at the more common serious medical conditions requiring ICU admission (septic shock and severe lung failure, for instance) and consider them together, about two-thirds of patients will survive the hospitalization and about half of those will experience some disability after.

That means that on average a bit over a third of patients admitted to ICUs with serious critical illness will ultimately have a strongly positive recovery and another third will have new disabilities, within which they will be able to thrive, albeit on new terms. Recovery will cost them a great deal in terms of their pain and discomfort, but by about six to twelve months after hospitalization, these survi­vors will often have a reasonable to excellent quality of life.

Where Approach 1 is the extreme package, Approach 2 would be something like a standard package: Use aggressive medical treatments as long as the chances of recovery are reasonable. It acknowledges that in some circumstances it might be better to allow a natural death rather than continue with medical treatments that are unlikely to succeed.

3: Only Admit Me to the Intensive Care Unit if I Have Excellent Chances

This would tend to apply when a life is close to the end but sud­den death would not be welcome. Aggressive medical treatment — with its attendant pain, expense, and discomfort — would make sense only if the patient were very likely to enjoy a strong recovery.

I find an increasing number of patients with chronic medical conditions favoring something like this when they come under my care. An ICU stay for a bleeding ulcer would make sense, even if it involved temporary intubation, because most people do well after such an episode. For severe pneumonia, on the other hand, an older patient who adopted Approach 3 would probably not undergo mechanical ventilation because the odds of recovery are not favorable. Pneumonia is, after all, the “old man’s friend.”

4: Don’t Admit Me to the Intensive Care Unit

Approach 4 is adopted when the ICU is a marker of a disease process that will cause significant suffering and discomfort, with only modest chance of recovery. For many individuals, particularly those late in life or near the end of a long illness, this approach will be appealing. If they had a mild kidney infection, one that could be tackled with antibiotics and fluids on a normal hospital ward, Approach 4 says they should be treated because doing so won’t deform the process of dying. But a disease severe enough that they would likely die without ICU admission would be the sign that a natural death was underway and should not be interfered with.

5: Don’t Admit Me to a Hospital; Focus Only on My Comfort

This is the approach when medicine can only delay or deform the experience of dying. In general, hospice provides this care today. This approach is traditionally used for patients with advanced cancer, but more diseases than cancer should qualify for this approach. The final stages of liver cirrhosis when transplant isn’t available are exhausting, debilitating, and associated with terrible delirium, while patients with very late heart failure may slowly and predictably suffocate. People with those conditions may much prefer a quiet death at home to a conveyor belt leading from nursing home to hospital and back.

Approach 5 acknowledges that there is a difference between the actual primary burden or discomfort associated with a treatment and the burden that results from postponing death. Antibiotics are a classic example. By and large, antibiotics aren’t painful. But over the course of human history, worn-out bodies have tended to die from urinary tract infection or pneumonia, so treating those conditions can prevent a peaceful passage from life.

While many hospitals aren’t quite ready for this kind of guidance, it’s good for all healthcare providers to start the conversation about how best to care for us when we fall sick. Instead of the confused “code status” conversations of prior decades, these approaches could help both patients and doctors to understand better how to tailor treatments during a serious illness. Even if they only prompt better communication, these approaches will have served an important purpose.

Samuel M. Brown, M.D., is the Director of the Center for Humanizing Critical Care at Intermountain Medical Center and author of Through the Valley of Shadows: Living Wills, Intensive Care, and Making Medicine Human (Oxford, 2016).