Adherence, compliance, persistence, concordance: What they mean and why they matter

Ask ten clinicians what they mean when they say a patient is “non-compliant” and you may get ten different answers. This article untangles four foundational constructs — compliance, adherence, persistence, and concordance — explains how they differ, and shows why the distinction has real consequences for clinical practice and research.

Compliance, adherence, persistence, and concordance: Four terms, four different meanings

These four terms are often used interchangeably, but they describe fundamentally different things.

Compliance is defined as the extent to which a patient follows the health professional’s advice and takes their treatment. The word implies a passive, obedient patient doing what they are told — a model that is increasingly regarded as obsolete.¹

Adherence is the degree or extent to which a patient’s behaviour — in terms of timing, dosage, and frequency — corresponds with agreed recommendations made by their prescriber.¹ Importantly, the WHO definition of adherence requires that recommendations be agreed upon with the patient, not simply handed down.¹ In practice, adherence and compliance are often used as synonyms in the literature, and major indexing services such as MEDLINE treat them as such.²

Persistence is a separate and distinct construct entirely. It refers to the act of continuing treatment for the prescribed duration — defined operationally as the duration of time from initiation to discontinuation of therapy.² In other words, adherence asks how well a patient takes their medication, while persistence asks how long they keep taking it at all — two different questions that must be defined and measured separately.²

Concordance, unlike the other three, does not describe patient behaviour — it describes the nature of the interaction between clinician and patient. It is defined as an “agreement between patient and healthcare professional, reached after negotiation, that respects the beliefs and wishes of the patient in determining whether, when, and how medicine is taken”.¹ There is therefore no such thing as a “concordant patient” — it is the consultation that is concordant, or not.¹

Understanding what these terms mean in theory is one thing — but in chronic disease, the consequences of confusing them become very concrete.

Stopping vs. skipping: How non-persistence and non-adherence drive different outcomes in chronic disease

In chronic disease, both non-adherence and non-persistence are common — but they represent different failure points requiring different responses.^1,3

The literature has historically conflated these two constructs, with many studies measuring one while calling it the other.² This matters because the determinants — and therefore the solutions — are not the same. A patient who forgets doses may need different support from one who has decided to stop treatment because they feel well or fear side effects.

In chronic disease, non-adherence is strikingly common. ¹ Adherence to long-term therapies averages around 50% in developed countries.¹ Approximately half of this non-adherence is intentional — an active decision not to take medication — while the other half is unintentional, arising from forgetfulness or regimen complexity. ¹ Non-persistence compounds this picture further. A patient may be taking their medication imperfectly — but at least they are still taking it. Once they stop altogether, even that partial benefit is lost.

A review of systematic reviews across 19 disease categories found 771 individual factors associated with non-adherence to long-term treatment, 47 of which were specifically identified as determinants of non-persistence.⁴ This historic imbalance highlights how persistence was overlooked as a distinct research construct — with more attention paid to how patients take their medication than to whether they continue taking it at all.⁴

Recognising that non-adherence and non-persistence are different problems is the first step — the next is understanding how each is defined and measured in research and practice.

Measuring medication-taking behaviour: Definitions for research and routine care

The most widely used conceptual framework for defining adherence in both research and routine care is the ABC taxonomy, developed by the Ascertaining Barriers to Compliance Project. Adherence to medications is composed of three phases:⁵

Initiation (taking the first dose)
Implementation (the extent to which dosing corresponds to the prescribed regimen from initiation until the last dose), and
Discontinuation (when the patient stops, for whatever reason).⁵

Persistence, within this framework, is the length of time between initiation and the last dose preceding discontinuation.⁵ This taxonomy emerged from a European consensus process involving 80 participants from 13 countries, with the stated aim of providing researchers and clinicians with a common language for describing adherence.⁵

In clinical trials, the first dose is typically administered on-site at enrolment, meaning initiation is usually assumed for all participants. The measurement focus therefore falls on implementation and persistence across the observation period.⁵

In routine care and real-world database studies, adherence is most commonly estimated from pharmacy dispensing records, prescription claims data, or electronic health records. Two constructs are central: primary adherence — whether the patient ever fills the initial prescription — and secondary adherence — whether they continue to refill it as prescribed.⁶ The most commonly used metrics are the Medication Possession Ratio (MPR) and the Proportion of Days Covered (PDC), both of which estimate the proportion of days a patient has medication available during a defined observation period.⁶

A critical limitation of most real-world metrics is that they cannot be calculated for patients who never fill their first prescription — systematically excluding the least adherent patients from the analysis.⁶ Terminology across studies also remains inconsistent, with the same constructs frequently referred to by different names and used interchangeably to refer to different things — a source of confusion that limits comparability across studies.⁶

How discontinuation specifically is defined — and how that definition changes depending on the setting — deserves closer attention.

Treatment discontinuation: How it is captured in studies and real-world practice

How treatment discontinuation is defined depends significantly on whether you are reading a clinical trial or a real-world database study, and the difference matters when interpreting the evidence.

In clinical trials, discontinuation is typically protocol-defined: a prespecified gap in dosing, a formal withdrawal, or a missed visit triggers a discontinuation event, and reasons are usually categorised and carefully recorded.⁷ This level of precision is valuable, but the controlled conditions of trials — selected populations, close monitoring, and protocol-driven follow-up — do not always reflect the complex realities of real-world care.⁸

In real-world database studies, discontinuation must be inferred from gaps in dispensing records.⁶ Raebel et al. propose defining it as the failure to have a medication dispensed within a defined number of days after exhaustion of the previous days’ supply — commonly 180 days — though this threshold varies across studies and should be specified explicitly.⁶ In addition, database studies cannot usually distinguish between patient-initiated and prescriber-initiated discontinuation, nor can they capture clinical context.⁶

Adverse effects, perceived lack of efficacy, feeling well, regimen complexity, and cost are among the most consistently identified factors across disease areas.⁴ These are modifiable — which is precisely why distinguishing discontinuation from poor day-to-day adherence matters. An intervention targeting forgetfulness is unlikely to help a patient who has made a reasoned decision to stop.

What you can do?

The following steps reflect evidence-informed practice and should be adapted to your clinical context, local guidelines, and patient needs:

Use precise language. Distinguish between adherence (day-to-day dosing behaviour), persistence (continuing treatment over time), and concordance (the quality of your shared decision-making conversation). Conflating these makes it harder to identify where the problem lies and what to do about it.⁵
Ask about non-adherence directly and without judgment. Research shows patients rarely volunteer this information, and doctors rarely ask. A simple, non-judgmental opening — “Many patients have difficulty or concerns about taking their medication every day — how are you finding it?” —may open the door.¹
Explore the reason before you intervene. A patient who misses doses due to forgetfulness needs different support from one who has stopped due to side effects or a belief that the medication is unnecessary.Understanding the determinant should precede the intervention.⁴
Address persistence explicitly, not just adherence. At follow-up, specifically ask whether the patient is still taking the medication at all — not just how well they are taking it.²
Aim for concordance in your consultations. Explore the patient’s understanding, concerns, and expectations about their medication before and during prescribing decisions. A patient who has genuinely participated in the decision is more likely to be committed to it.¹
When reviewing adherence data, know what the metric measures. MPR and PDC measure whether medication was dispensed — not whether it was taken. Primary non-adherence (never filling the prescription) is invisible to these metrics.⁶

Conclusion

Compliance, adherence, persistence, and concordance describe distinct constructs with different clinical implications and different measurement requirements — and getting the language right matters for understanding where the problem lies, designing appropriate interventions, and interpreting the evidence. For the HCP in daily practice, a useful shift may be from asking ‘is this patient compliant?’ to asking ‘what kind of non-adherence is this, and what does this patient need?

This article was written with the assistance of generative AI technology and reviewed for accuracy.

FAQ

Poor provider-patient relationships, weak communication, and lack of trust in healthcare have all been identified as factors that negatively affect adherence, while the quality, duration, and frequency of interaction between patient and doctor, along with the doctor’s ability to demonstrate empathy and elicit patient concerns, have a positive effect.⁴ In other words, how you talk to your patient about their medication is not separate from whether they take it — it is part of the same equation.

Non-adherence is driven by multiple patient-related factors, and no single characteristic reliably predicts it — meaning clinicians cannot assume a patient will or won’t adhere based on demographics alone.⁴ Factors with a consistent negative effect include forgetfulness, health beliefs that conflict with the treatment rationale, denial of diagnosis, substance abuse, and psychiatric comorbidities such as depression and anxiety.⁴ On the positive side, higher self-efficacy, knowledge of the disease, perceived necessity of treatment, and strong social support are associated with better adherence.^1,4,9

It matters more than it might seem. Over ten different terms have been used in the literature to describe medication-taking behaviour, often with different meanings.⁵ If a patient is describe as “non-compliant” without distinguishing whether they are missing doses, taking them erratically, or have stopped altogether, it becomes harder to target the response effectively.

Not necessarily — and this is exactly why aggregate adherence figures can be misleading. Three patients all taking 75% of prescribed doses over the same period can have completely different underlying patterns: one may have initiated late but taken doses consistently thereafter, another may have initiated on time but missed doses erratically throughout, and a third may have been fully adherent initially but stopped altogether before the end of the period.⁵ The same overall percentage conceals fundamentally different clinical situations — which is why the ABC taxonomy distinguishes between initiation, implementation, and discontinuation as three separate, quantifiable phases rather than collapsing them into a single adherence score.⁵

Not necessarily. Collecting a prescription and taking medication correctly are two different things. Persistence and compliance are distinct constructs that need to be measured and reported separately to fully characterise a patient’s medication-taking behaviour.² A patient can be persistent without being compliant, and vice versa. Knowing which issue you are dealing with changes what you do about it.

The distinction has both clinical and economic consequences.²
Poor day-to-day compliance affects the degree of therapeutic benefit achieved, while non-persistence — stopping treatment before the prescribed duration has elapsed — means the patient is no longer receiving any treatment at all. Clinical outcomes are affected not only by how well patients take their medications but also by how long they take them.² Treating both problems the same way is unlikely to work for either.

Concordance is not a description of patient behaviour— it describes the nature of the consultation itself. It is defined as an agreement between patient and healthcare professional, reached after negotiation, that respects the beliefs and wishes of the patient in determining whether, when, and how medicine is taken.¹ This means there is no such thing as a “concordant patient” — it is the consultation that is concordant, or not.¹ Concordance invites clinicians to move beyond simply conveying a treatment plan, towards helping patients with decisions they make for themselves and supporting them in seeing these through.¹

This is a recognised pattern in clinical practice, and understanding why it happens can help address it. Research shows that patients frequently hold reservations they do not voice, and may actively decide not to take prescribed medication without disclosing this.¹ At the same time, studies indicate that patients are generally passive in consultations and that their ideas, concerns, and expectations about medication are not always fully explored.¹ The apparent agreement seen in the consultation may reflect a patient’s natural tendency to defer rather than genuine commitment to the treatment plan.¹

Start by naming the medication and explaining what it does — patients are better equipped to make an informed decision when they know what they are taking and why.¹ Be open about both benefits and common side effects rather than focusing on benefits alone, and invite the patient to share any concerns they have about taking it.¹ Share your own clinical opinion — patients want to hear what their doctor thinks.¹ Finally, rather than assuming adherence, ask about it directly and without judgment — a simple question about how the patient is getting on with their medication is more likely to surface real barriers than waiting for them to volunteer the information.¹

It depends on how discontinuation was defined in that study.⁶ In real-world database studies, discontinuation is usually inferred from a gap in prescription refills — specifically, when a patient fails to collect their next supply within a defined number of days of running out, commonly set at 180 days.⁶ Database studies also cannot distinguish between a patient who chose to stop and a prescriber who discontinued the medication,⁶ meaning that discontinuation figures across studies may not always be directly comparable.