Knowing When To Stop
“When do we stop?” my patient’s son asked me.
“That’s really hard to know,” I answered.
We were discussing when to stop medical interventions in hopes of trying to save his father’s life. He’d been diagnosed with a severe gastrointestinal bleed from a stomach ulcer. Though most bleeding of this kind is identified in time and stopped, my patient had lost nearly half his blood volume and passed out before anyone realized anything was wrong. By the time his family had brought him to the emergency room, he was hypotensive, unconscious, and in acute kidney and liver failure. He was admitted directly to the ICU, transfused eight units of blood, and put on a ventilator. Forty-eight hours later, his kidney function had worsened enough that we were considering beginning him on dialysis, and he’d developed an aspiration pneumonia.
“It’s just one thing after another,” his son lamented when I told his family this news.
I nodded. We see this commonly, I told him. When one organ system starts to fail for any reason, others frequently follow, like lemmings diving off a cliff, either as a direct result of the failure of the first organ system or as a result of the same insult that caused the first organ system to fail (as in his father’s case). Additionally, if there’s one adage in medicine that’s both tragically true and unappreciated by laypeople it’s that interventions (like ventilators) to reverse or manage complications (like brainstem injury) frequently lead to more complications (like aspiration pneumonia).
Brainstem testing in my patient on the third day, however, revealed improvement, suggesting the initial damage he’d sustained as a result of prolonged low blood pressure might be at least partially reversible. And though on that particular day his kidney function had taken a slight turn for the worse and we’d decided to begin dialysis (hoping to use it as a bridge until his kidney function recovered completely) the day before it had actually improved. Which direction was it really going? We didn’t know. His liver, on the other hand, was definitely coming back.
So many difficult questions confronted us, questions we needed to ask not just once but every day: which systems were improving? Which were worsening? Which could improve and which couldn’t? The answers changed every day as the enormous volume of data we collected on every aspect of his body’s function changed. Individually, each organ system that had taken a hit retained the potential to heal—no insult, as far as we could tell, was definitively irreversible. But we also knew, statistically speaking, that with each additional organ system affected in a critically ill patient, the likelihood that the patient will recover diminishes significantly. Every cog in the human body retains influence on every other cog, it seems.
The most challenging problem in inpatient medicine—especially in ICU medicine—turns out not to be figuring out how to fix what’s broken, but rather figuring out when not to try. We’re confronted by this problem in other contexts in medicine as well. With patients who suffer from dementia, that cruelest of diseases (which always progresses), we often have to ask ourselves when to stop intervening. Dementia, many people don’t realize, is uniformly fatal. Given enough time, all demented patients will become non-verbal and stop eating. At what point, then, do we withhold antibiotics for pneumonia, for example? Or insert a feeding tube (via the skin overlying the stomach) to augment declining nutrition? For a doctor lacking a personal connection to the human being their severely demented patient once was, these choices are difficult at best. For family members who retain a vivid picture of their loved one in full possession of all their faculties, they’re agonizing. For this reason, laying the burden of making life-and-death decisions on family members has always struck me as particularly cruel. And yet a balance must be struck between assuming paternal authority to make such decisions unilaterally in order to spare families from feeling responsible for ending their loved one’s life and allowing family members to act as surrogate decision makers who must often guess at what their loved one would have wanted (as often no discussion of end-of-life wishes ever took place). And the reality is that even when patients themselves are awake, alert, and competent, they often have difficulty themselves navigating all the potential risks and benefits of the interventions we in the medical profession are able to offer.
So we simply all do our best. We try to balance a reaction to today’s data and the interventions they suggest we make with the broader picture and the likely long-term outcome. Keeping a sharp eye on both, knowing which to believe more on any given day, is extremely challenging. But this is where the art of medicine lives and wisdom and compassion are most vitally required.
In the end, my patient survived. His kidneys recovered and he came off dialysis. But the injury to his brain proved extensive: he was no longer able communicate in any way other than through grunts and could no longer bathe, dress, or feed himself. Prior to his hospitalization, he’d been a CPA heading up a busy, privately-owned business. Afterward, he required twenty-four hour care. I wanted to ask him if he was glad he’d survived, if we’d made the right choices for him. But it wasn’t clear to me at that point if he would have even understood the question.