Improving medication adherence at 90 days
Medication adherence at 90 days is the point where a chronic therapy either holds or is quietly abandoned. Adherence rarely fails on day one — it decays as the initial motivation fades, side-effects accumulate and refills lapse. Improving it means detecting the drift while it is still drift, finding out why, and getting the right intervention to the right reason.
The 90-day mark matters because it is late enough for the initial resolve to have worn off and early enough that the patient can still be brought back.
Why does adherence fall off around 90 days?
Day one adherence is usually fine. The patient has just seen a clinician, the instructions are fresh, and the reason for the medication is emotionally vivid — particularly if it followed a hospital stay.
All three of those decay. The instructions blur, the clinician is a memory, and the frightening event recedes. What remains is a daily task with no immediate feedback: an antihypertensive does not feel like anything, so taking it produces no sensation of benefit and skipping it produces no sensation of harm. The behaviour has no reinforcement, and unreinforced behaviours decay.
By 90 days the first repeat has usually come due, the first side-effect has usually been noticed, and the first "I felt fine, so I stopped for a few days and nothing happened" experiment has usually been run. The patient who is still taking their medication at 90 days is likely to keep taking it. The one who is not has usually stopped for a reason nobody has heard.
What are the four reasons patients actually stop?
These matter because they need different fixes, and treating them as one problem — "the patient forgot" — is why reminder-only interventions plateau.
They forgot
The only one a reminder solves. It is real, and it is the reason vendors like: it has a software fix. It is also the minority of the problem, which is why reminder apps show early promise and then flatten.
A side-effect
The patient felt worse taking it than not, and made an entirely rational trade-off. The fix is a clinical conversation — a different drug, a different dose, a different time of day. A reminder here is not neutral, it is irritating: you are pinging someone to do the thing that makes them feel unwell.
Cost
They could not afford the refill. Patients systematically under-report this, and they under-report it more to software than to a person who is kind about it. The fix is not clinical at all — it is a cheaper alternative, a different pack size, or a benefits conversation.
They decided it was unnecessary
They felt fine, so they concluded they did not need it. For an asymptomatic condition that is a reasonable inference from the evidence available to them. The fix is understanding, not nagging, and the patient will not volunteer this reason unless asked without judgement.
What actually moves adherence at 90 days?
Detecting drift rather than absence. One missed dose is noise. A pattern of missed evening doses, a refill collected four days late, and a patient who stopped answering is a signal — and it is visible weeks before the patient would appear on any report as non-adherent.
Asking why, and treating the answer as data. The four reasons above need four different responses, so the reason has to be captured as a structured field rather than as a sentence in a call note. An intervention that cannot distinguish a cost problem from a side-effect will apply the wrong fix to both.
Routing the clinical ones to a clinician. A side-effect report is a pharmacist conversation. An agent that tries to counsel the patient there has exceeded its remit; an agent that logs it and moves on has wasted the most valuable thing it learned all month.
Making the contact cheap enough to be regular. Adherence is a pattern across weeks, so observing it needs contact across weeks. Quarterly contact does not observe a pattern — it samples it, badly.
No adherence-improvement figure appears in this guide. Published adherence numbers vary enormously by condition, population, how adherence was measured, and whether it was self-reported — a lift measured on a statin cohort tells you very little about an oncology one. OneDose does not publish an adherence figure it cannot attribute to a named deployment.
Frequently asked
- Why does medication adherence drop at 90 days?
- Because the things sustaining it early all decay: the instructions blur, the clinician becomes a memory, and the event that motivated the prescription recedes. Meanwhile the medication itself gives no feedback — an antihypertensive does not feel like anything — so the behaviour has no reinforcement. By 90 days the first repeat, the first side-effect and the first "I stopped and felt fine" experiment have all typically happened.
- Why do reminder apps stop working after a few weeks?
- Because forgetting is only one of four reasons patients stop, and it is the minority one. Side-effects, cost and a considered decision that the medication is unnecessary all need different responses. A reminder sent to someone who stopped because of a side-effect is not neutral — it is pinging them to do the thing that makes them feel unwell.
- How do you find out why a patient stopped taking their medication?
- By asking, in a way that makes the honest answer easy, and capturing the reason as structured data rather than as a note. Cost in particular is systematically under-reported, and patients under-report it more to software than to a person who is kind about it — which is a real limitation of automated outreach worth knowing.