Author: Jayshil J. Patel, MD1
Gwynne’s Grammar1 is my favorite writing guide. In it, he advocates for precision in the use of language, which is the machine for thinking and necessary for clear judgment and good decisions. Gwynne creates a nine-step argument for the importance of words, even suggesting: “If enough societies in the world crumble as a result of bad decisions taken because of bad thinking, yes, the whole of world civilisation faces collapse, with consequences for each individual that are literally incalculable.”
The field of medicine is also a complex society with specialized divisions of labor and a unique system of communication. If Gwynne is correct that words reflect our thought process and without them we cannot reason things out, which can lead to poor decisions, then medicine (as a civilization) faces risk of collapse, merely as a result of the words we choose.
Therefore, the words we use matter. They matter for our decision making and thus our patients’ well-being. Words also matter for new students learning to use language as a tool in medical practice, which requires semantic precision.
Herein, I outline four linguistic pitfalls, two I used during patient and family encounters (the leading question and “I understand”), two I hear as an academic critical care physician during bedside presentations (“poor historian” and “unremarkable” or “normal”), and outline their impact on clinical practice.
Patient and Family Encounters
The Leading Question
“You don’t have [insert favorite symptom here], do you?”
In residency, I cared for a man admitted for chest pain, or so I thought. By the time I had finished evaluating him, I had already sifted through electronic notes, many of which consistently stated, “chest pain.” When I finally obtained a history from the patient, I kept asking, “You don’t have [insert symptom here], do you?” Before I left the patient’s room, he said, “What about my toe pain?”
“Toe pain?” I asked.
“Yes, my right toe has been hurting for a couple of days and that is why I came to the hospital,” he replied.
“What about your chest pain?” I asked.
“I don’t have chest pain right now,” he responded.
“You don’t?” I asked with astonishment
‘Incompetent’ must have been written all over my face.
“No, someone downstairs kept asking me to describe the chest pain that I had two weeks ago and so I did, but that was one time and I haven’t had it since.”
That patient encounter reminded me that interviewing is a two-way street. The patient was evaluating me just as much as I was ostensibly evaluating him. Consequently, he lost confidence in me, as did I in myself that day.
Today, by the time house staff actually see and examine a patient, many of them have already biopsied the electronic health record. A 14-point review of systems is read to be “unremarkable” (see below). Consequently, the patient is led to a conclusion the physician wants to hear. This style of questioning (“leading the witness”) may promote cognitive errors such as anchoring, confirmation bias, and diagnostic momentum, which drive poor clinical decision making. Worse, patients can lose confidence in an uninformed and assumptive clinician. Creating patient confidence is the first step in shared decision making and therefore relies on gathering accurate information with unbiased open-ended questions.
A middle-aged man presented with abdominal pain and his wife was explaining, in great detail, how he (and she) felt when he slumped over at home. I recalled being in a hurry because I had to roll another patient to radiology. As she spoke, I furiously nodded and tapped my foot to (probably subconsciously) let them know I needed to move on.
I must have repeated “I understand” a few times while she expressed anxiety over her husband’s illness.
“No, you do not understand!” she responded.
“Have you ever had someone you love slump over in front of you?”
She was right. I certainly did not “understand.” How could I? The patient’s wife had lost faith in my ability to simply pause and listen.
As a fellow, I had to break bad news to the family of a young patient who had experienced a cardiac arrest and severe anoxic brain injury. The family expressed their shock, disbelief, and ultimately grief. During this encounter, I had more time to actually listen. Despite ample time, the phrase, “I understand” carelessly slipped out of my mouth.
My carelessness was met with contempt. The family’s grief turned to anger and the meeting ended shortly thereafter.
In both instances, hasty assertion that “I understood” their feelings in such an emotionally charged situation had the opposite effect from what I had desired.
Each individual has a unique capacity for emotions or conscious experiences. There is no stereotypical way to truly understand another person’s loss, despair, sorrow, suffering, disapproval, anger, sadness, fear, apprehension, or the infinite combinations of experiences. Rather, we can pause to imagine what it must be like, for example, to suffer. “I understand your [insert emotion here]” should probably be replaced with “I can try to imagine your [insert emotion here].” Pausing to imagine gives us an opportunity to strengthen our emotional intelligence, a prerequisite for establishing patient and family trust, particularly when confronted with sensitive decision making.
During rounds in the intensive care unit, I often hear, “The patient is a poor historian.” I began thinking about the impact of this term after hearing it being used to describe consecutive patients during morning rounds one day. First, for the patients labeled “poor historians,” I made less effort to gather additional patient-derived history compared with those patients who were not labeled “poor historians.” In effect, I had “framed” the patient. Next, the term poor historian was typed throughout electronic health record, ready to bias the next clinician. Lastly, I identified semantic imprecision. The term historian was incorrectly applied.
A historian, by definition, is one who tabulates history.2 My team and I were using the term to label those patients who were unable to provide an accurate history. In actuality, I was the poor historian.
Later, when I ignored the term and began to pay attention, I concluded encephalopathy, dementia, depression, frustration, anxiety, and anger were reasons why patients did not provide an accurate history. In fact, there was always a reason why the patient wouldn’t or couldn’t provide a history. I learned that a patient’s inability to provide a history should not limit my ability to correctly obtain one.
Unremarkable or Normal Findings
Many academic medical centers are highly specialized, technology-laden environments gripped by work-hour restrictions. This increase in technology has created a discordant juxtaposition. On one end, technology has sped up communication, eased acquisition of diagnoses through less invasive means, and enhanced patient care support systems. On the other end, this increase in medical technologies has created more uncertainty.
Protocol-driven medicine coupled with an electronic health record can lead to nondiscriminant ordering of tests. In addition, work-hour restrictions have increased patient handoffs. As a consequence, not infrequently, an overnight admission note is literally read aloud verbatim during morning rounds, with less emphasis placed on the presenter’s physical examination, and greater emphasis placed on laboratory and imaging data.
Laboratory and imaging results verbalized as “normal” or “unremarkable” without context of the history and examination creates an irrational primacy effect because of which, information delivered earlier in a presentation may be perceived as being more important. In effect, the early and often verbalization of “normal” may pave the way for premature closure. A “normal” serum procalcitonin stands in the way of a clinical diagnosis of sepsis. An “unremarkable” chest radiograph challenges a clinical diagnosis of pneumonia. A “normal” echocardiogram refutes a clinical diagnosis of heart failure.
Arbitration and interpretation of the merits and perils of technology-derived data in the context of pretest probability and their sensitivity and specificity for the condition in question are necessary to promote clinical reasoning; otherwise, technology, despite its triumphs, will only continue to muddy clinical acumen.3
I have described just four linguistic pitfalls that impacted my clinical practice: leading questions undermined history taking and diminished patient confidence, carelessly stating “I understand” created distance between myself and a patient’s family, mislabeling patients as “poor historians” inadvertently overshadowed my observations as to why my patient could not provide a history, and noncontextual use of “unremarkable” or “normal” may promote irrational primary effect. Certainly many more linguistic pitfalls exist that potentially drive poor decision making.
If clinicians are to think correctly and apply our thoughts usefully, words need to be used precisely. Usage of words reflects the clarity (or clumsiness) of our thinking and decision making, no matter how simple or complex, and ultimately impacts the health of people who are affected by our decision making. It is essential to “think about the way we think,” starting with the words we choose in everyday clinical care. Doing so may guide us to recalibrate our vocabulary toward semantic precision. Otherwise, if Gwynne is correct, we risk diminished effectiveness of our decisions, merely as a result of the things we say.
Author: Jayshil J. Patel, MD, Division of Pulmonary and Critical Care Medicine, Medical College of Wisconsin, 9200 W Wisconsin Ave, Ste E5200, Milwaukee, WI 53226
1. Gwynne NM. Gwynne’s Grammar: The Ultimate Introduction to Grammar and the Writing of Good English. New York, NY: Knopf; 2014:5-8.
2. Tiemstra J. The poor historian. Acad Med. 2009;84(6):723.
3. Mukherjee S. The Laws of Medicine. New York, NY: Simon & Schuster; 2015.