Primary non-adherence, secondary non-adherence, and implementation: A practical guide for clinicians

When a patient does not take their medication as prescribed, an equally important question to “why?” is “where in the process has it broken down?” Primary non-adherence — never filling the prescription — and secondary non-adherence — failing to take medication correctly once treatment has begun — are distinct failure points with different causes and different solutions.¹ Understanding which is occurring shapes what conversation to have, what intervention to offer, and whether the support provided will reach the patient who needs it.

Starting before the first dose: Primary and secondary non-adherence as distinct clinical constructs

Primary and secondary non-adherence describe two fundamentally different failure points in the medication-taking process.

Primary non-adherence occurs when a medication is newly prescribed but the prescription is never filled — the patient never starts treatment at all.¹ Secondary non-adherence describes what happens after a prescription has been filled: the medication is not taken as prescribed, whether through missed doses, incorrect timing, or early discontinuation.¹ Primary non-adherence is a failure of initiation; secondary non-adherence is a failure of implementation.²

The distinction matters because the two constructs have different causes and call for different responses. A patient who never fills their prescription may have concerns about starting treatment, doubts about its necessity, or cost barriers — often without disclosing this to their prescriber.² A patient who fills their prescription but takes it inconsistently faces a different set of challenges, whether forgetfulness, side effects, or eroding motivation over time.¹

Primary non-adherence is more common than is often appreciated. A systematic review and meta-analysis of 33 studies across six long-term conditions estimated that approximately 17% of patients did not collect a newly prescribed medication, with rates in that review highest in osteoporosis and hyperlipidaemia.² Many patients who do collect a prescription may still not take the first dose, meaning the true extent of primary non-adherence may be higher than available estimates suggest.²

Recognising which type of non-adherence a patient is experiencing is the first step — the next is understanding what implementation actually means in research and routine care, and how it is captured.

Dose by dose: What “implementation” means in adherence research and routine care

Implementation is defined within the ABC taxonomy, a widely used conceptual framework for defining adherence in both research and routine care developed by the Ascertaining Barriers to Compliance Project, as the extent to which a patient’s actual dosing corresponds to the prescribed regimen, from initiation until the last dose preceding discontinuation.³ It is the “how well” of medication-taking — capturing not just whether doses are taken, but their timing, frequency, and correspondence with what was prescribed.²

Implementation non-adherence rates vary widely across patient groups, conditions, measurement methods, and time periods.² Growing evidence suggests that approximately a fifth of patients do not initiate treatment at all; of those who do, approximately 30%–50% do not implement their treatment as prescribed; and over long follow-up periods in some conditions, 80%–100% of patients discontinue.²

Implementation is also more complex than a simple count of doses taken. It can encompass timing of doses, amount taken, overuse, and adherence to associated instructions such as taking medication with food or fluid.² Two patients with the same overall percentage adherence can have entirely different underlying patterns — one may miss doses erratically throughout, while another may have been fully adherent initially before stopping altogether.³ These different patterns have different clinical implications and call for different responses.³

For researchers, implementation is more precisely captured by electronic medication monitors that record the time and date of dosing.² In routine care, pharmacy dispensing records and self-report measures are more practical, though each has limitations.¹ No single measure is considered a universal gold standard, and a multi-measure approach is often recommended to capture a fuller picture of medication-taking behaviour.¹

A critical limitation of dispensing-based metrics is that they can only be calculated for patients who have already filled at least one prescription — meaning primary non-adherence is systematically invisible to them, and the least adherent patients are excluded from the analysis.⁴

Understanding how non-adherence is measured is one dimension of the picture — understanding why patients do not take their medication as prescribed requires a different lens entirely.

Why patients take — or don’t take — their medication: Behavioural science frameworks

In behavioural science terms, medication-taking behaviour is understood as a complex, dynamic process shaped by a patient’s beliefs, practical circumstances, and motivation — not a fixed trait or a simple matter of knowledge.^2,5 A foundational distinction in the field is between intentional and unintentional non-adherence. Unintentional non-adherence occurs when a patient wants to take their medication but is prevented by factors beyond their control — forgetting, difficulties with the regimen, or lack of access.⁵ Intentional non-adherence occurs when a patient makes an active decision not to follow the prescribed recommendations, driven by beliefs and preferences about the treatment.⁵

The Necessity-Concerns Framework is a well-evidenced model for understanding intentional non-adherence. It proposes that adherence is shaped by patients’ implicit judgements of personal need for the treatment — their necessity beliefs — balanced against concerns about potential adverse consequences.⁵ A meta-analysis of 94 studies involving 25,072 patients across a wide range of long-term conditions found that higher adherence was consistently associated with stronger necessity beliefs and fewer concerns, with these relationships remaining significant across different countries, study designs, and adherence measures.⁵ For every standard deviation increase in necessity beliefs, the odds of adherence increased by a factor of 1.7; for every standard deviation increase in concerns, the odds of adherence decreased by a factor of 2.0.⁵

The Perceptions and Practicalities Approach draws on this evidence base to propose that adherence support should address both the perceptual dimension — beliefs and concerns — and the practical dimension — capacity and resources — tailored to the needs of the individual.²

A complementary framework applied to medication adherence is COM-B (Behavior model which considers three components: capability, opportunity and motivation), which proposes that a patient will not take their medication reliably without the physical and psychological capability to do so — such as memory and the ability to swallow — combined with the social and physical opportunity, such as support from family and access to medication, and both reflective and automatic motivation, including beliefs, impulses, and habits.² COM-B is useful for mapping the specific barriers a patient faces and informing the development of targeted intervention strategies.²

Non-adherence is recognised as a dynamic behaviour rather than a stable patient characteristic — it changes within the same individual over time, across treatments, and in response to life circumstances.² A single assessment of adherence at one point in time therefore provides an incomplete picture, and patients who are adherent now may not remain so as treatment continues.

What you can do

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

Check if non-adherence is primary or secondary. At follow-up, check whether the patient is actually taking the medication — not just whether they collected the prescription. Understanding whether non-adherence is primary or secondary will help tailor an appropriate response.¹
Explore necessity beliefs and concerns at the point of prescribing. Patients who doubt the necessity of their treatment or harbour concerns about adverse effects are less likely to adhere.⁵ A non-judgmental opening — “Many patients have questions or concerns about starting a new medication — is there anything you would like to ask?” — can help surface these early, before they become a barrier to taking the medication.
Distinguish between practical and perceptual barriers before intervening. A patient who forgets doses needs different support from one whose non-adherence is driven by concerns about side effects or doubts about whether the medication is necessary.⁵ Using a COM-B framework — asking about capability, opportunity, and motivation — may help identify where the barrier lies.²
Recognise that adherence changes over time. A patient who is adherent at the first follow-up may not remain so. Because non-adherence is a dynamic behaviour rather than a stable trait, a patient’s adherence at one point in time may not reflect their adherence over the longer term.²
When reviewing adherence data, know what the metric measures. Pharmacy dispensing records estimate whether medication was collected — not whether it was taken. Primary non-adherence, where the prescription is never filled, is invisible to these measures.⁴

Conclusion

Understanding whether a patient has never started their medication, is taking it inconsistently, or has stopped altogether points towards fundamentally different conversations and interventions. The behavioural science frameworks explored in this article — from the Necessity-Concerns Framework to COM-B — offer clinicians a structured way to move beyond the question of whether a patient is adherent, towards understanding the beliefs, circumstances, and motivations that shape their medication-taking behaviour at each stage.

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

FAQ

Primary non-adherence can take several forms, not all of which are immediately visible to the prescribing clinician. The most straightforward example is a prescription that is written but never presented at a pharmacy — the patient leaves the consultation with no intention of filling it, or changes their mind before doing so.² Less obviously, a patient may collect their prescription but never take the first dose, with studies of medication waste reporting returns of unopened packets to community pharmacies, suggesting that not all dispensed medication is started.² Cost and availability are also documented barriers, particularly in lower-resource settings, where patients may not collect a prescription simply because the medication is unaffordable or unavailable.²

Secondary non-adherence encompasses a wide range of medication-taking behaviours that fall short of the prescribed regimen.¹ The most common examples include missing doses entirely, taking doses at the wrong time, taking too little or too much of the prescribed amount, and failing to follow associated instructions such as taking medication with food or fluid.² Secondary non-adherence is not always intentional; forgetfulness, regimen complexity, and practical difficulties accessing medication are among the most frequently reported unintentional causes. ^2,5

Primary non-adherence is more prevalent than is often recognised in routine practice. Studies across six long-term conditions estimated that approximately 17% of patients did not collect a newly prescribed medication.² Rates vary with patient age, treatment type, and healthcare system factors including cost and availability.² Importantly, these figures are likely to underestimate the true extent of primary non-adherence, as patients who collect a prescription but never take the first dose are not captured by dispensing-based measures — meaning the problem may be larger than standard data sources suggest.²

Secondary non-adherence is highly prevalent across chronic conditions, though estimates vary considerably depending on how adherence is defined and measured, the condition studied, and the time period assessed.² Evidence suggests that 30%–50% of patients who initiate treatment do not implement their medication as prescribed.² Rates can be considerably higher for treatments that are difficult to take or access. In one review of intravitreal therapy for macular degeneration, implementation non-adherence rates as high as 95.6% were reported.² Over long follow-up periods, 80%–100% of patients in some conditions discontinue treatment altogether.² Together, these figures suggest that secondary non-adherence is a cause for concern, particularly when persistence is taken into account over longer follow-up periods.²

Primary non-adherence can be measured by linking prescribing data with pharmacy dispensing records to identify prescriptions that were written but never filled within a defined time period.² This approach requires a centralised system in which both prescribing and dispensing data are captured and can be linked — a capability that varies considerably across healthcare systems.¹ Where such linkage is not possible, primary non-adherence is largely invisible to standard measurement, as most commonly used metrics can only be calculated for patients who have already filled at least one prescription.⁴

Two widely used metrics are the Medication Possession Ratio (MPR) and Proportion of Days Covered (PDC), both of which estimate the proportion of days a patient has medication available during a defined observation period.⁴ Self-report questionnaires are practical in routine care and can help identify reasons for non-adherence, though patients may report better adherence than they actually achieve.¹ Electronic medication monitors offer a granular picture, recording the time and date of each dose, but are expensive.² No single measure is considered a universal gold standard, and combining methods is generally recommended.¹

Necessity beliefs refer to a patient’s implicit judgement of how much they personally need their prescribed treatment — whether they believe it is essential for maintaining or improving their health.⁵ These beliefs are not always aligned with clinical evidence; a patient may be prescribed a clinically indicated medication yet perceive it as unnecessary, particularly if their condition causes no symptoms or they feel well.⁵

Concerns refer to a patient’s worries about the potential adverse consequences of taking their medication, including side effects, dependence, and long-term effects.⁵ Concerns are common even when medication is well tolerated, and are often related to more general beliefs about pharmaceuticals as a class of treatment.⁵

COM-B proposes that behaviour only occurs when three conditions are met: Capability, Opportunity, and Motivation.² Applied to adherence, a patient needs the capability to take their medication — such as memory and the ability to swallow — the opportunity to do so — such as access to the medication and family support — and the motivation, including their beliefs, habits, and impulses.² The value of COM-B in clinical practice is that it provides a structured way to identify where the barrier lies for an individual patient, rather than assuming all non-adherence has the same cause.²

Both carry clinical risk, but evidence suggests that the poorest outcomes are associated with patients who never initiate treatment or who discontinue early, rather than those who take their medication imperfectly.² A patient who misses occasional doses may still achieve meaningful therapeutic benefit, whereas one who never starts — or stops altogether — receives none at all. Rather than ranking one above the other, the more useful clinical question may be which type of non-adherence is occurring for a specific patient — and what is driving it.