Language Matters

The chart said the patient was non-compliant. She had stopped taking the antihypertensive prescribed at her last visit. The note read like a verdict: a failure of will, a defect of character, a problem to be documented. What the chart did not say, because the visit had been six minutes long and the question had not been asked, was that the medication had given her a cough so persistent that she could not sleep, that she had no money to come back for a different prescription, and that she had stopped because stopping was the only intervention available to her at the time.

Non-compliant. The word travels with the patient through the medical record, into the next clinician's first impression, into the assumption that whatever happens next is going to be hard because the patient is the kind of person who does not do what they are told.

Words like this are not neutral. They name a thing in a way that quietly determines how the thing will be treated. And in healthcare, where a sentence in a chart can shape how a patient is approached for years, the difference between calling someone non-compliant and calling someone non-adherent is the difference between assigning blame and asking a question.

What Compliance Actually Means

Compliance is a word borrowed from a particular kind of relationship. It comes from contexts in which one party issues an instruction and the other party is expected to follow it. The model is hierarchical by definition. The clinician knows. The patient obeys. When the patient does not obey, the failure is the patient's.

This framing has deep roots. For most of the modern history of medicine, the clinical encounter was structured as a transmission of expert knowledge from a credentialed authority to a layperson who was expected to receive the instruction and act on it. Compliance was the natural noun for that transaction. A patient who did not comply was not a patient with a problem to be solved together. They were an obstacle to the treatment plan.

The trouble with this framing is that it does not describe how human beings actually change behavior. People do not change because they have been told to change. People change because something inside them shifts, usually after someone else has helped them think through the situation in a way that respects their reasoning, their context, and their constraints. A frame that begins with obedience cannot get to behavior change. It can only produce documentation of failure.

What Adherence Means Differently

Adherence is a word borrowed from a different kind of relationship. To adhere to a plan is to stick with it, and the implication of the verb is that the plan was, at some point, agreed to. Adherence asks not whether the patient followed orders but whether the plan held up under the conditions of the patient's actual life. When it did not, the question is not what is wrong with the patient. The question is what made the plan unworkable, and what would make a workable plan.

The shift from compliance to adherence was formalized by the World Health Organization in 2003, when the WHO defined adherence as the extent to which a person's behavior corresponds with agreed recommendations from a healthcare provider. The word agreed is doing critical work in that definition. It locates responsibility for the plan inside the relationship between clinician and patient, not inside the patient alone.

Some clinicians have moved further still, toward concordance, which adds the explicit claim that the plan must be co-constructed before adherence to it can be expected. Concordance treats the encounter as a negotiation between two experts: the clinician, who knows the medicine, and the patient, who knows their life. The plan that emerges from that negotiation is the only plan a patient can adhere to, because it is the only plan that fits.

Why the Word Matters Even When the Plan Is the Same

A skeptical reader might ask whether any of this is more than a rhetorical exercise. The medication is the same. The exercise program is the same. Does it really matter what we call the patient's behavior afterward?

It matters for three reasons.

The first is that language shapes attention. A clinician who has been trained to look for non-compliance will find non-compliance, because the word predisposes them to interpret missed doses and skipped sessions as character failures rather than as signals of an unworkable plan. A clinician trained to look for non-adherence will find the same missed doses and ask a different question: what made adherence hard? The data are the same. The investigation that follows is not.

The second is that language shapes the patient's experience of being seen. Patients know when they are being judged. A patient who senses that their clinician views them as the kind of person who does not follow instructions will withhold information that might confirm that judgment. They will not say that they could not afford the medication. They will not say that the side effects scared them. They will not say that the program does not fit into their week. The information that would have changed the plan is the information that the framing has driven into hiding.

The third is that language shapes what the clinician feels permitted to do. A clinician who calls a patient non-compliant has a small set of available next moves: warn, document, discharge. A clinician who calls a patient non-adherent has a larger set: ask, collaborate, adjust, simplify, refer, problem-solve. The language is the door. Some doors open onto more rooms than others.

Where Compassion Enters the Picture

The connection between compassionate care and patient adherence is one of the most consistently replicated findings in the contemporary healthcare literature. Patients who experience their clinicians as compassionate are significantly more likely to take their medications as prescribed, follow through on therapeutic recommendations, return for follow-up appointments, and disclose information that affects diagnostic accuracy. Stephen Trzeciak and Anthony Mazzarelli's synthesis of more than a thousand studies in Compassionomics and Wonder Drug documents this pattern across nearly every clinical context that has been examined.

The mechanism is not mysterious. Compassion lowers the threshold for honest disclosure. When patients trust their clinician, they tell the truth about what is actually happening at home, what they have actually been doing with the medication, what they are actually afraid of. Honest disclosure produces better plans. Better plans are more workable plans. Workable plans get followed.

In rehabilitation, the same finding appears under the language of therapeutic alliance. A 2021 observational study by Alodaibi and colleagues, published in Physical Therapy, demonstrated that higher patient ratings of therapeutic alliance predicted better functional outcomes during episodes of physical therapy care for low back pain, with effect sizes that exceeded those of many discrete therapeutic techniques. Ambady and colleagues, working two decades earlier, had already shown that brief observations of physical therapists' nonverbal communication predicted geriatric patients' subsequent health outcomes. The patients did not need to articulate what they had received. Their bodies registered it, and the trajectory of recovery shifted accordingly.

What unifies these findings is that adherence is a relational outcome, not a personality trait. It does not live inside the patient. It lives in the space between the patient and the clinician, and it tracks the quality of that space.

Compassion as a Behavior Change Tool

The behavior change literature reaches the same conclusion from a different direction. Decades of work on motivational interviewing, the transtheoretical model, and self-determination theory converge on a finding that should be unsurprising by now: people are more likely to sustain a behavior change when the change reflects their own values, fits their own context, and was arrived at through a process they experienced as collaborative. They are less likely to sustain a change when it was prescribed at them in a way that ignored what they brought to the encounter.

Compassion is the relational substrate that makes collaborative planning possible. A clinician who is genuinely interested in the patient as a person can ask the questions that surface the patient's actual values and constraints. A clinician who is performing interest, or who is too rushed to register the patient as a person at all, cannot. The same clinical recommendations, delivered through these two different relational frames, will produce different outcomes, because the patient is not a passive substrate on which the recommendations act. The patient is a person who is deciding, every day, whether to do what was suggested.

What This Means in Practice

Three small adjustments follow from all of this.

The first is to retire the word compliance from the chart and from the team's everyday vocabulary. When a patient has stopped taking a medication, missed appointments, or not followed the home program, the note should describe what happened and what the patient said about why. It should not assign a label that closes the inquiry before the inquiry has begun.

The second is to ask the adherence question differently. Not "are you taking your medication?" but "how is the medication going for you?" Not "are you doing your exercises?" but "what has gotten in the way of the exercises this week?" The first set of questions sets up a yes-or-no answer that the patient often answers in the way they think the clinician wants. The second set of questions invites the truth.

The third is to recognize that the relational frame the clinician brings into the encounter is doing more work than the recommendation itself. The forty-second moment of compassion that Lois Fogarty and her colleagues documented at Johns Hopkins is not a courtesy. It is the variable that determines whether the recommendation that follows it will be followed. Compassion is not the soft part of the visit that surrounds the real work. Compassion is what makes the real work possible.

A Note on the Word Choice

I have used the word adherence throughout this piece, but I want to acknowledge that some of my colleagues prefer concordance, and that there is principled disagreement about which word does the most work. What both words share is a refusal to treat the patient as a passive recipient of instructions. Either is better than compliance. The most important shift is the shift away from a vocabulary in which the patient's behavior is graded against a standard the patient never agreed to.

Care differently, not less.