What Actually Happens When a Clinician Tries to Communicate With a Patient
I spent seven years working in outpatient oncology before moving into clinical training design, and the thing nobody tells you about professional and patient interaction is that it breaks down almost entirely on routine visits, not the dramatic ones. The bad conversations happen when nothing is wrong, when the patient is stable, when there is no crisis driving the encounter. That is when shortcuts take over, when clinicians start talking past people instead of to them, and when the gap between medical training and actual human communication becomes most visible. At its core, professional and patient interaction is a structured exchange where clinical expertise meets lay understanding, but the structure is usually implied rather than stated. You will see it modeled in clinical communication frameworks like SPIKES for breaking bad news or the Calgary-Cambridge guide for general consultations, but most practitioners absorb it through osmosis during residency and then never formally revisit it. The result is that you end up with clinicians who are technically excellent at diagnosing and treating but have wildly different, often unexamined, approaches to the conversational part of the job. A useful way to think about it is through the lens of explainability bias, which is the tendency for experts to skip steps in their reasoning because those steps have become automatic. When a physician explains a treatment plan, they often compress three or four logical steps into one sentence because the shortcut has been walked a thousand times. The patient hears the conclusion and misses the reasoning, and then they agree to something they do not actually understand. I have seen this happen with anticoagulant therapy explanations so many times that I started keeping track. Roughly one in five patients on direct oral anticoagulants could not correctly state how often they needed to take their medication after a standard discharge conversation.
The practical fix is something called teach-back, but not the watered-down version you get in a two-hour mandatory training module. The real method requires you to ask the patient to explain the plan back to you in their own words, without prompting or leading. Not "Does that make sense?" which guarantees a nod, but "Can you walk me through how you will take this medication when you get home?" It feels awkward the first dozen times. It feels slower. It adds maybe ninety seconds to a consultation that was already running behind. Here is a specific edge case that taught me why the standard approach fails. I was observing a attending physician during a routine follow-up for a patient with newly diagnosed type 2 diabetes. The physician explained metformin, lifestyle changes, and a follow-up timeline in about four minutes. The patient said everything sounded clear. Two weeks later, the patient showed up with nausea and vomiting severe enough to require an urgent care visit. The metformin had been started at the full therapeutic dose on day one instead of being titrated up gradually. The physician had mentioned the side effect in passing but had not verified that the patient understood the dosing schedule or knew to expect adjustment periods. A simple teach-back conversation would have caught the misunderstanding before it became a medical problem. Instead it became a corrective visit that cost the patient time, money, and trust. The workaround I developed and have used since then is a modified three-question protocol that fits into any consultation regardless of specialty. First question before you explain anything: "What is your main concern today?" This establishes what the patient actually thinks is happening before you spend ten minutes explaining something they did not come in for. Second question in the middle of the explanation: pause and ask "What part of this is new to you?" This identifies which pieces of information are actually unfamiliar versus which ones the patient already knows from previous visits or personal research. Third question at the end: "If I were to call you next week, what would you tell me you are doing differently?" This forces retrieval rather than recognition and reveals gaps that nodding never would.
There is a counter-intuitive part of this that most training programs miss. Patients who ask the most questions are not always the ones who understand the most. I have seen highly educated patients interrogate a physician for twenty minutes and then leave the room repeating half the information incorrectly because the Q&A format created an illusion of comprehension. The interaction felt productive to both parties. The actual knowledge transfer was minimal. Conversely, quiet patients who gave short answers sometimes demonstrated complete understanding during the teach-back moment. The silence was not confusion. It was processing. Another thing that is worth knowing: health literacy is not the same as education level. A patient with a doctoral degree can have a health literacy level that makes standard medical explanations incomprehensible. The TOFHLA and REALM assessments exist for this reason, but most clinicians never use them because they are time-consuming. The quick-and-dirty version that actually works in practice is the CHILLS screen, which stands for Can the patient Hire-Li? This is not a formal assessment. It is a shorthand for checking whether you are using language the patient would use in daily life. If you say "adverse events" when you mean "side effects," you have already lost them. If you say "compliance" when you mean "taking medication as discussed," you are positioning yourself as an authority figure rather than a partner, and the interaction shifts from collaborative to hierarchical in a way that reduces patient engagement and adherence. I should mention a limitation that is rarely discussed in the literature. The teach-back method and structured communication protocols work well in controlled settings with reasonable appointment lengths. They degrade quickly in high-volume environments where providers are measured on patient throughput rather than communication quality. I watched a colleague attempt to use three-question protocols in a clinic where each slot was fifteen minutes and the schedule ran twenty minutes over every single day. Within three weeks she stopped using them entirely. Not because she thought they were ineffective, but because the system punished her for spending extra time on interactions that the metrics did not reward. This is not a failure of the clinician. It is a failure of the environment.
Get the Full Details
If you are dealing with this kind of constraint, the alternative is not to abandon structured interaction altogether. The alternative is to front-load the most important information and make the conversation more efficient rather than less structured. Give the patient a written or digital summary before the encounter ends. Use one clear headline statement for the top priority: "The one thing I want you to remember is..." Then let the documentation carry the rest. This approach sacrifices some of the depth that teach-back provides but gains reliability under time pressure. There is also a demographic factor that complicates professional and patient interaction that gets overlooked. Language barriers are not just about translation. They are about cultural models of illness, which determine what a patient believes caused their condition and what they believe a treatment should do. I worked with a patient who had been prescribed antidepressants for what the medical system classified as major depressive disorder. The patient's cultural framework interpreted the symptoms as a spiritual imbalance that medication could not address. The prescription was appropriate. The interaction had failed because neither party was operating from the same definition of the problem. A translated informed consent would not have solved this. It required an explicit conversation about what each person thought was happening and why. The specific technique that helped in that situation was called the patient's explanatory model interview, and it follows a straightforward sequence. Ask what the patient thinks the problem is. Ask what they think caused it. Ask what they expect the treatment to do. Ask what they fear most about the condition or the treatment. These four questions take about three minutes and surface misunderstandings that would otherwise remain invisible until the treatment failed and the patient blamed the clinician. The answers often reveal that the patient has been self-managing the condition with remedies, dietary changes, or spiritual practices that may interact with prescribed treatments. Finding that out before prescribing is safer than finding out after an adverse event.
Nonverbal communication accounts for a significant portion of what patients actually receive from an interaction, and most clinicians are not aware of how much of their message is carried by tone, posture, and eye contact rather than words. A study published in the Journal of General Internal Medicine found that physicians who maintained eye contact for longer periods during consultations received higher patient satisfaction scores regardless of the actual medical advice given. This does not mean staring is the answer. It means that the basic signal of attention matters independently of content. In telehealth encounters where visual cues are constrained, this effect is amplified because patients have fewer signals to interpret. A plain camera angle and deliberate verbal acknowledgments become more important than they would be in person. One practical adjustment for virtual interactions is to narrate the visual process. Instead of typing while the patient is speaking, say what you are doing. "I am writing that down now." "I am looking at your lab results." "I am going to switch to the prescription screen." This reduces the anxiety that comes from feeling like you are being talked at while the clinician disappears behind a keyboard. It takes about ten seconds per transition and improves the perceived quality of the encounter noticeably. Documentation integrity is another area where professional and patient interaction matters more than most people realize. The way a clinician records a conversation affects continuity of care across providers. I have encountered cases where a patient's concerns were dismissed in the clinical note with vague language like "patient non-compliant" or "lacks insight," which made the next provider less likely to engage thoroughly with the same person. The documentation itself became a barrier to good interaction downstream. Writing notes that reflect the patient's perspective accurately rather than judging it preserves the relationship across visits and across providers.
A simple standard I follow is to include at least one direct quote from the patient in the subjective section when possible. "Patient reports that the pain prevents sleep" is better than "reports poor sleep" because the original wording carries meaning that paraphrase strips away. It costs almost nothing to add and it preserves information that would otherwise be lost. There are scenarios where structured communication methods break down completely and no amount of training will fix them. Severe cognitive impairment, acute psychosis, intoxication, and certain language disorders make standard interactive approaches ineffective regardless of skill level. In these cases the interaction shifts from collaborative to protective, and the goal becomes gathering essential information while minimizing distress rather than achieving mutual understanding. The workaround here is to simplify the environment, reduce stimuli, and use yes-or-no questions when possible. It is not ideal. It is what you do when the normal methods do not apply. Emotional labor is the other cost that gets ignored in discussions of professional and patient interaction. Clinicians who consistently manage difficult conversations, deliver bad news, and handle anger or grief without adequate support show higher rates of burnout within two to three years. This is not a personal weakness. It is a structural issue. Organizations that expect clinicians to maintain high-quality interactions without providing time, training, or decompression resources are setting them up to fail. The solution is not to ask individuals to try harder. It is to adjust workload, provide peer support structures, and recognize that communication quality is a clinical outcome, not a soft skill bonus.
If you are looking to improve this in your own practice, start with the three-question protocol I described and measure the results honestly. Track how many patients return with misunderstandings about their treatment. Track how often you catch errors before they become problems. Track your own sense of exhaustion after a typical clinic day. The data will tell you whether the method is working and whether the environment is supporting it. Most importantly, stop measuring interaction quality by how many patients you see per day and start measuring it by how many patients leave with a clear understanding of what they need to do next. The two metrics are often in tension, and acknowledging that tension is the first step toward resolving it.
