Persistence, Continuation, and Long-Term Medication Use

This topic concerns persistence — how long a person continues a medication from initiation until discontinuation — as distinct from day-to-day implementation. Long-term continuation is a recurring challenge in chronic disease, where many patients stop therapy entirely within the first year, and it is measured and analyzed differently from adherence over a fixed window.

Definition

Persistence is the duration of time from initiation to discontinuation of therapy — that is, how long a person keeps taking a medicine — operationalized as the time until a gap in supply exceeds a predefined permissible interval; discontinuation marks the end of persistence.

Scope

The entry defines persistence and discontinuation, contrasts them with implementation, and describes how continuation is quantified (for example, time to discontinuation and the permissible gap that defines a stop). It is reference material and does not advise on continuing or stopping any individual's medication.

Core questions

• How is persistence defined and distinguished from implementation and overall adherence?
• How is continuation measured, and how does the permissible gap affect the estimate?
• How common is early discontinuation in long-term therapy, and why does it occur?
• What is the consequence of treating persistence and implementation as a single measure?

Key concepts

• Persistence versus implementation
• Discontinuation and time to discontinuation
• Permissible gap / refill gap
• Primary non-persistence
• Survival analysis of continuation
• Switching versus stopping
• Long-term therapy in chronic disease

Key theories

ABC taxonomy: discontinuation phase

Within the initiation-implementation-discontinuation framework, persistence is the time to discontinuation; separating it from implementation clarifies that a person can implement a regimen well yet still stop early, or continue while implementing poorly.

Mechanisms

Persistence is measured over time rather than as a single ratio: a patient is considered persistent until a gap between supplies exceeds a permissible interval, after which they are classified as having discontinued. Because the permissible gap and the observation window are analytic choices, persistence estimates depend on them, and survival-analysis methods (time to discontinuation) are commonly used. Conceptually, persistence answers 'for how long?' while implementation answers 'how well during?'; conflating the two — for example, by reporting a single adherence percentage — obscures whether a low value reflects poor day-to-day dosing or early stopping, which have different drivers and implications.

Clinical relevance

Distinguishing persistence from implementation matters for interpreting adherence data and for designing pharmacy services that target continuation, such as refill follow-up. This entry explains the concept and its measurement as knowledge; it is not guidance on whether any patient should continue or stop a medicine.

Epidemiology

In many chronic conditions a substantial share of patients discontinue therapy within the first year, and continuation typically declines steadily over time; the World Health Organization's review of long-term therapies highlights this as a core problem of chronic-disease management. Reported rates vary with the permissible-gap definition used.

Evidence & guidelines

Consensus terminology work (for example, the ISPOR definitions and the ABC taxonomy) recommends reporting persistence separately from compliance/implementation and specifying the permissible gap and observation period so that continuation estimates are interpretable and comparable across studies.

History

As pharmacy claims data became widely available, researchers recognized that a single adherence percentage conflated two distinct phenomena. The ISPOR terminology paper (Cramer and colleagues, 2008) and the 2012 ABC taxonomy formalized the separation of persistence from compliance/implementation, establishing persistence as a measure of continuation over time.

Debates

How should the permissible gap defining discontinuation be set?

Whether a patient has 'discontinued' depends on the allowed gap between refills; different choices yield different persistence estimates, and there is no universal standard, so transparent reporting of the threshold is essential.

Key figures

• Bernard Vrijens
• Joyce Cramer
• John Urquhart
• Sabina De Geest
• Eduardo Sabaté

Seminal works

• vrijens-2012
• cramer-2008
• sabate-2003

Frequently asked questions

How is persistence different from adherence?

Persistence measures how long a person stays on a medicine before stopping (the duration of therapy), whereas adherence in the implementation sense measures how closely they follow the regimen while they are taking it. A patient can stop early yet have taken doses correctly until then, or continue for a long time while frequently missing doses.

What is the permissible gap in persistence measurement?

It is the maximum allowable interval without medication supply before a patient is counted as having discontinued. Because persistence estimates change with this threshold, studies are expected to state the gap they used.