Ms. Jones was an 87-year-old woman whose heart stopped while she was at home watching TV.

I was working the night shift in the emergency department when she was rolled into the trauma bay with a man kneeling atop her gurney, pushing forcefully on her chest, pumping her heart for her. Immediately, she was stripped naked, a central line, multiple peripheral IV’s, a chest tube, and a breathing tube were inserted.

I stood in the corner, wide-eyed and bewildered by the organized chaos unfolding before me as everyone in the room worked swiftly to save her life. In just a few minutes, Ms. Jones’ heart began beating on its own again. A miracle, I thought. I was a third-year medical student at the time, and it was my first time witnessing someone come back to life.

Our patient remained ventilated while we tried to contact her family for recommendations about what to do next. In her electronic medical record, she did not have a code status documented, meaning we did not know what she would have wanted us to do for her. Everything? Nothing? Her family was not reachable, so we had to default to everything.

At the time, I remember thinking that this was best for her anyway. Our interventions are working. Why would we stop now?

Thirty minutes later, I was in the middle of examining another patient when an overhead page sent me running back to the trauma bay. It was Ms. Jones – crashing again, pulseless. “Start chest compressions,” the chief resident commanded me, even though I had only ever done CPR on rubber thoraxes. I began pushing on Ms. Jones’ chest like I had the mannequins so many times before except this time I felt ribs crack beneath my hands. “It’s okay. That means you’re pushing with adequate force,” the chief assured me.

In a few minutes, a nurse took over my role. I stood beside Ms. Jones’ bed, wondering if I would ever forget the way it felt to crack another person’s sternum. I stood in the line of nurses waiting to rotate as the compressor, and I held Ms. Jones’ hand, unsure of what to do besides pray that she couldn’t feel what we were doing to her. Several minutes later, her heart began beating on its own again and the crowd of medical personnel dispersed. I lingered in her room, watching in silence as bruises began forming on her bare chest.

Our patient was stable for about an hour until I was paged back to her room as she died again, and the same sequence of events began. Compressions, epinephrine, compressions, epinephrine. Bruises expanding. Hopes of a meaningful recovery dwindling.

Yet, we continued.

We weren’t allowed to stop until she was pulseless and breathless after three rounds of CPR, and she continued to revive on round two. For about six hours, her heart stopped and started again cyclically until it eventually just didn’t start again, and she was pronounced dead.

What I originally thought was a magical pursuit was really just the start of a night of torture for our patient before her ultimate passing.

Ms. Jones arrived at the emergency department dead, and she left the same. Except she arrived in her pajamas, looking clean, intact, and now, she lies in the trauma bay with four broken ribs, a punctured lung, and a sunken left chest full of blood. Needles and tubes emerged from every possible orifice. Pajamas in the corner, cut to shreds. I stood at the foot of her bed with my gloved hand resting atop her rigid toes, and I felt the crease of my N95 moisten with tears.

What have we done?

As doctors-in-training, we are taught how to push pressors, place chest tubes, insert central lines. We learn how to cut the body open, remove organs, stop bleeds, and sew the layers back together like nothing ever happened. We can pump patients’ hearts, force air into their lungs, push nutrients through their veins. Over time, we grow comfortable administering aggressive interventions. We become immune to the barbaric nature of modern medicine.

The intensity of care we provide to our patients is not our decision. It is a patient’s duty to designate their end of life wishes along a spectrum of aggressive life-saving measures to comfort-care in documents called advanced directives. At the start of medical school, the hardest deaths for me to cope with were the patients who decided to refuse medical interventions. I saw them as lost opportunities, people we could have saved. Yet as I progressed through my training and witnessed several deaths similar to Mrs. Jones’, I began to hold the opposite sentiment. Now, I often feel relieved when certain patients downgrade their code status. My advice for family members who ask me about code status has shifted. Even thoughts about my own code status for the future have changed. This is a shared feeling amongst many providers. In a survey of over 1,000 doctors about their end-of-life wishes, 88% stated they would elect to have a do-not-resuscitate status.

Doctors are forced to pursue aggressive medical interventions to keep patients like Ms. Jones alive, even when they do not see benefit to doing so. Patients with out-of-hospital cardiac arrest have less than a 6% chance of survival. Yet, physicians are forced to complete three rounds of CPR and ventilation in the hospital before they can pronounce death in full-code patients or patients without a documented code status. Such aggressive interventions in futile situations come with physical harm to patients in their final moments and emotional harm to physicians. In a survey of 42 internal medicine resident physicians, 21% experienced at least one symptom of post-traumatic stress disorder (PTSD) after participating in a patient resuscitation.

When I think about the events that ensued at the end of Mrs. Jones’ life, I often imagine a reality where she was spared the bruises, fractures, needles, and pain. An alternate world in which she was allowed to experience nothing but peace in her final moments. Her story, while unfortunate, illuminates the importance of discussions amongst patients, families, and primary care providers regarding code status, a difficult topic that will one day be relevant for all of us.