"This is the end of my life.” These were my patient’s only words — an economy of phrasing made necessary by an all-consuming air hunger. She had just arrived in the emergency room, Code 3 critical, after a lights-and-sirens ambulance transport from her nursing home. Awake, alert and intensely focused, every effort of her frail, 90-year-old body was concentrated on the simple act of breathing. Her weak heart and failed kidneys had caused her lungs to fill with fluid, every breath becoming a mixture of water and air. The analogy to drowning is inevitable. As her physician, I was going to have to make some big decisions quickly, including this one: How much should I do to save her life?
Typically, when patients gasping for air and with very low oxygen levels show up in my emergency room, it's obvious what to do first: Intubate them. You put a tube in their mouth down into their lungs and connect them to a respirator. It's an aggressive treatment, and anyone awake for it will tell you it's unpleasant. But without enough oxygen in the bloodstream, bad, often irreversible things happen to the body very quickly.
But this patient wasn't typical. She was old, she had no next of kin, and it was clear that a series of recent medical interventions had failed to stabilize her. She'd been in our intensive care unit three times for over a week in the last four months alone. Her most recent stay had involved a seven-day stint on a respirator. Her eventual discharge to the nursing home was a medical high-wire act; an all-too-familiar pattern of readmission to discharge to readmission was playing out.
Perhaps with this in mind, my patient had months earlier completed an advance directive. I was holding it in my hand. We were allowed to make her comfortable: pain medication, oxygen. Explicitly proscribed were intubation, vasopressor medications to support blood pressure, invasive procedures and chest compressions. Her chart noted that she had been intubated the last time she was at the hospital because, as is often the case, paramedics didn't have the advance directive. The chart also noted she was upset about this.
Here's the thing about advance directives: They are guidelines. They do not have the force of law. This is a good thing. Despite what many of us think, we can't truly anticipate how we will feel when we are about to die. Advance directives are especially valuable when someone has a terminal disease or a stroke, and they want to tell doctors what to do when they can no longer decide for themselves. But with a patient such as mine, who was alert despite her distress, it's much more complicated. The human will to live is powerful. Most of the time when patients truly face death, they insist we do what we can to keep them alive.
And so three minutes after her arrival, my patient was surrounded by doctors, nurses and respiratory techs getting ready to do what we do, getting ready to save her life. An oxygen mask was already placed over her nose and mouth, intravenous lines started, cardiac monitor leads attached. A doctor waited at the head of the bed, sedation medications in hand, poised to intubate.
As the senior doctor on duty, what happened next was my call. I knew her medical history. I knew what her advance directive said. And most importantly, I knew what she had said to me minutes ago, when I laid eyes on her: “This … is the end … of my life.” There had been no inflection in her voice. It was not a question.
Before I explain what I did next, I want you to stop and consider what you would have done. I ask because society has been debating for decades the role physicians should play in how and when we die. In June, California will join four other states in allowing doctors to prescribe lethal medication to terminally ill but mentally competent patients.
While I support the new California law, I also worry about its implementation and the gantlet of checks and balances patients and caregivers will have to navigate. Who will be in charge of enforcement? Will there be an appeals process for denials? Can any law codify something as complex, emotionally charged and controversial as end-of-life decisions. In truth, the options aren't black and white but intensely gray, defined by nuance and competing goals, and in the moment, doctor and patient may struggle to find common ground.
Although the preservation of life is paramount in medicine, the inevitability of death necessitates a pivot at some point by patient and physician. When and how this happens will be different every time. The patient with terminal cancer choosing to forgo chemotherapy is familiar to most of us. But what of the bedridden stroke patient with a failing heart who now requires dialysis, or someone with severe Alzheimer’s who needs open heart surgery, or the alcoholic who will die without a new liver but continues to drink?
As we age or when we become ill, the many decisions we make with our doctor about our care, even small ones, link to create a system of values we will draw upon when our time comes. Put another way, choices about how to live one’s life ultimately and simultaneously are choices about one’s death. A doctor can inform us, advise us, but in the end — and right up to the end — the choice is each of ours to make.
The Hippocratic oath is both vague and specific about how to manage death. It cautions against “playing God” while acknowledging “it may be within our power to take a life; an awesome responsibility that must be faced with great humility and awareness of one’s frailty.” What this means is that to first do no harm, you should think about how you would like to be treated in the same situation.
At the patient's bedside, I chose to do less rather than more: "Supportive care only," I said. It was an uncomfortable, conflicted moment for everyone in the room. Our job is to heal the sick and save lives. Then and there, though we could have righted this sinking ship, we did not.
What was left was to bring a measure of comfort to what would inevitably be her death. I gave her a small dose of narcotic in her intravenous line to quiet her air hunger. The effect was almost immediate. Her breathing slowed. Her face relaxed. Her eyes remained open, but she was asleep.
It took about an hour for her to die. First the alarm on the cardiac monitor started going off, unaccustomed to the kinds of numbers it was registering. Then her breathing became erratic — fast and deep, punctuated by long pauses. Her heart rate slowed to 20 beats per minute, then accelerated, then dropped back into the 20s. We disconnected her from the monitor, and the beeping went silent.
I waited and watched for a full 10 minutes more, half expecting her to resume breathing. She didn't. The drugs hadn't killed her. It was the lack of oxygen in her blood. The drugs just made her more comfortable while that process took place — which is what she had wanted.
I have thought a lot about this patient in the months since that night. I know we did the right thing, and I know many doctors who would have done exactly what I did. But I also know many who would have intubated her in the absence of a signed affidavit reaffirming her advance directive.
I can't help wondering if what makes end-of-life decisions so complicated is our definition of medicine itself. The idea of doctoring as a pitched battle against disease is compelling. But it belies the uncomfortable fact that physicians are really more stewards than soldiers. Our patients, and indeed each of us, always die in the end. It's remarkable and good that medicine allows us to live 25 years longer on average than we did 100 years ago. But sometimes my profession forgets that the end point of medical innovation and intervention isn't immortality.
Back in the ER, I completed my patient’s medical chart, laboring to put into words the complex choreography of the last hour — a life reduced in death to time stamps, vital signs, monitor readings and the like. Lost in my recounting was the drama of what happened, the nobility of a woman utterly self-aware and in tune with the narrative of her own mortality. She had rejected the false hope of a science that still struggles to understand or communicate its own limitations.
At the end of my shift, the charge nurse reminded me to call the coroner. After giving the coroner my patient’s basic demographic information, I was asked the inevitable question — “Doctor, what’s the cause of death?” I hesitated — knowing he’d be looking for "cardiopulmonary arrest," "heart failure" or some other expression of convenience. He needed to check a box on a form and yet, I couldn't bring myself to say my lines.
The patient died with a weak heart and failing kidneys, not because of them. I wanted to acknowledge this woman’s choice — her role in her death — and ours.
After what must have seemed like a very long time, I spoke. “Natural causes,” I said. “She died of natural causes."
Dr. Eric Snoey is vice chair of the department of emergency medicine at Highland Hospital in Oakland.