Hello UCSD IM family!
I just read this perspective piece in this week’s NEJM.
As you set out to care for patients, a thought provoking reminder of how important it is for each of us to listen to our patients!
Raphael Rush, M.D.
July 4, 2019
N Engl J Med 2019; 381:9-11
“Not so hard!” howled the man with abdominal pain, his body tense against his stretcher.
His skin looked sallow under the fluorescents illuminating the crowded emergency department (ED) hallway. My fingers had barely brushed his belly.
Although we’d just met, his records told a familiar story of months shuttling between EDs with various pains, no cause ever found. Old prescriptions in his chart had raised my suspicions. The normal preliminary ultrasound report seemed to confirm them. Still, I resigned myself to referring him to the surgeon on call. I wrote my note carefully, thinking about what to say.
Writing a history is as much a process of editing as composition. It takes a long time to learn what matters. As my training progressed, each note formed a rough draft for the next, another chance to focus on the bare facts of each case. My presentations became shorter, as online templates and the glazed eyes of senior residents taught me which questions to ask and which to ignore. Others’ notes identified the relevant diagnoses and framed my perspective on each patient. Just as patients all changed into the same blue gowns, we fit everyone’s life into the small set of stories we prewrote in our minds.
Brevity wasn’t the only reason details were cut. One attending physician with the frown of a grammar teacher marked residents’ notes in red pencil. One she returned to me bore a bold circle around a woman’s social history.
“Take out her marijuana usage,” she told me, as we sat side by side in her cramped examination room, reviewing my notes. “It’s not relevant, and I don’t want that staying in her chart.”
The patient’s marijuana use seemed unimportant to me, but I understood what concerned my attending. The words in a chart stay there forever, passed on to everyone a patient meets. If I wasn’t careful, all kinds of conclusions might be drawn from my notes. Even then, during fellowship, I still wasn’t sure what went into the record and what stayed out.
The details that are recorded can significantly affect how a story is perceived. Susan Sontag, writing about the stories conveyed by war photography, questioned the belief that photographs are meant “not to evoke but to show” — to state facts rather than imply conclusions. Sontag pointed out that whether intentionally or not, photographers choose which details to show. Viewers, removed from the context in which pictures were taken, use those details to impose the biases of their experiences in developing their own interpretations of the underlying story.
All I really knew about the patient with abdominal pain was what he’d told me and the results of my exam. But my perceptions were colored by his old notes and by my experiences with other patients. As I wrote my note, I remembered what he claimed about his alcohol and smoking history; I thought about the tension in his abdomen and his red-rimmed eyes.
Then I recalled the list of narcotic prescriptions on his chart. Other, more experienced doctors had seen him before. None of them had found anything wrong.
I was used to making quick decisions in the ED using all the information I had available and felt certain about my initial conclusion. But as I came to the end of my note, something made me hesitate.
The psychologists Daniel Kahneman and Amos Tversky named the tendency to lend excessive weight to a compelling story the “representativeness heuristic.”2 It’s a heuristic that physicians often apply. A group of researchers presented family physicians with the case of a 43-year-old woman with a sudden-onset headache and stiff neck, classic signs of meningitis. Physicians were 14% less likely to order further investigations if they were also told that the woman had a recent history of stress.
In the end, I didn’t record my patient’s apparent lack of reaction when he seemed unaware that I was feeling his abdomen. Nor did I indicate my own conclusion as to why he was there: that his abdominal pain was a ruse to obtain narcotics. Slipping his chart almost stealthily into its cubbyhole, I consulted general surgery and called in the next patient from the packed waiting area.
I find it a constant struggle to determine which details matter most. The challenge is partly prospective, when I’m writing notes that others will read. It’s tempting to say that the solution is to write even less in a chart, avoiding any detail that might bias a reader.
Certainly some things are not worth recording. When I returned to the hospital for another shift, I opened the chart of the man with abdominal pain. The surgeon who saw him had rethought the case. Unconvinced by the ultrasound, she ordered a repeat study, which showed cholecystitis. His pain resolved once his gallbladder came out. I tightened my lips, ashamed. I was relieved only that I had left my suspicions out of my note: my omission meant that another physician could assess him with an unjaundiced eye.
But details that are irrelevant at one point may prove crucial at another time. In another context, the patient’s old medication list could have saved his life. And it is often impossible to predict how a note will later be read. Sontag suggested that any detail can reinforce existing biases or introduce new ones. From the moment they see a photograph, viewers start reinterpreting its story to fit what they think they already know.
So our larger problem is retrospective: how to interpret what is already written about the patients we see. The surgeon who cared for my patient saw the same chart I had seen. But she was willing to assess him without preconceptions. She didn’t interpret his symptoms in light of his chart — she recast the old notes in light of the patient in front of her.
My past mistakes live on within me. I thought of that patient recently, when a woman came to my clinic with pain “NYD” — not yet diagnosed. It wasn’t for lack of trying: several other physicians had seen her already.
Sometimes I’m frustrated by requests for second opinions. The woman’s general practitioner had sent me the assessments of the other specialists she’d seen. It seemed unlikely that I would be the one who diagnosed her problem. For a moment, I was tempted to spend the appointment reading her chart, copying and pasting and inserting my own brief additions while she spoke. She seemed to sense this inclination, sitting down without removing her coat, as though she didn’t expect to stay long.
Shutting the door behind her, I noticed that my small waiting room was empty. From the window behind us came the gray light of late afternoon. In the corner of my screen, I had an eye on my schedule. Clinics always seem to run late. But after her name on my calendar came a large block of unscheduled time. There would be nobody else for a while. My time was my own, to use as I chose.
I turned my chair away from the computer and angled toward her. My stethoscope weighed on my neck and I removed it, along with my smudged glasses, which forced me to lean in. I picked up a pen and some paper, ready to transcribe whatever she said.
After a moment, she took off her jacket and settled into her chair, resting her coffee cup down on my desk. She seemed to relax. I confirmed her name and date of birth, and then we began.
“I have a lot of records from your other doctors,” I said. “We’ll review those together in a bit. But I want to hear your story again, in your own words, if that’s OK. From the beginning.”