Adherence to medical advice

The size of the problem

Roughly half of patients with long-term conditions do not take medication as prescribed. A treatment that works in a trial does nothing for a patient who stops taking it, so adherence is not a side issue — it is often the difference between a treatment working and not working.

Two kinds of non-adherence

Unintentional: forgetting, misunderstanding instructions, a complicated regimen, physical difficulty opening a container. Intentional: the patient decides not to, usually because of side effects, because they feel better, or because they do not believe the treatment is necessary. The distinction matters because the two need completely different solutions — a reminder app is useless for a patient who has decided the drug is not worth it.

Why people decide not to

The health belief model says behaviour depends on perceived susceptibility, perceived severity, perceived benefits and perceived barriers, triggered by a cue to action. Applied to adherence, the practical reading is: patients adhere when the perceived benefit outweighs the perceived barrier, and side effects are the barrier that matters most.

Measuring adherence

Self-report is cheap but subject to social desirability bias. Pill counts are objective but a patient can discard tablets. Biological measures (blood or urine tests) are the most valid but expensive and invasive. Electronic pill containers record each opening — accurate for timing, though opening is not proof of swallowing.

Improving adherence

Simplify the regimen; explain rather than instruct; use reminders and behavioural cues; involve the patient in the decision; and follow up. Marks come from matching the strategy to the type of non-adherence in the case.