How to track medication adherence without relying on memory

Self-reported med adherence is famously off. People say "yes, mostly" when the truth is "four out of seven last week." This is not a moral failing. It is what memory does. Here is how to set up tracking that gives you and your clinician something real.

Why "mostly" is not enough

The question your clinician is going to ask at the next visit is whether the medication is working. They cannot answer that without knowing whether you are taking it. NIMH lists medication adherence as one of the central elements of long-term bipolar care2024. The same is true across psychiatric medications more broadly.

Generic answers — "most days," "I think so" — leave the clinician guessing. Honest data, even imperfect, is the better input.

Three tactics that work

1. Log at the moment, not the end of the day. Ambulatory assessment in psychiatry has consistently shown that real-time logging captures things retrospective recall does not2013. The five seconds it takes to tap "taken" right when you swallow the pill beats trying to reconstruct the day at bedtime.

2. Log misses honestly. A "not taken" tap is worth more than skipping the entry entirely. The pattern of misses — Saturdays, mornings after late nights, the day after travel — is the signal your clinician needs.

3. Log against mood and sleep. Smartphone-based self-monitoring in bipolar disorder works partly because mood, sleep, and meds are all in one place2015. A missed dose followed by a poor night and a low day three days later is a complete data point. Three separate apps would have lost it.

What to log, specifically

The minimum useful set is the simplest one:

• Whether you took your meds today. A daily yes/no is enough to surface the pattern of misses.
• A short note when something changed (a dose adjustment, a new med, a side effect day, a switch). The note field is where the context goes.

You do not need to track time-of-day to the minute. The pattern of taken-vs-not days is the dataset. Anything more granular usually does not change clinical decisions.

What patterns are worth flagging

• Adherence dropping below your usual. A run of misses is more informative than the average.
• Adherence dropping alongside sleep changes. Sleep and mood are tightly linked2008; a missed-meds stretch that precedes a hard week is the kind of pattern that drives a treatment review.
• Notes that line up with dose changes or side effects. The notes field is where to capture the context, and the chart is where to read it back.

Adherence is not a moral test

People miss doses. The job is not to be perfect; the job is to give your clinician an accurate picture so they can adjust the plan to how you take it in real life. A 70% adherence rate is something they can work with. An idealized "yeah, I take it" is not.

What MoodSync does with this

MoodSync has a daily "meds taken" toggle on every entry, alongside mood scales and sleep — and a notes field for the context (dose changes, side effects, switches). The sleep-and-meds feature page walks through what the chart looks like.

A serious note on discontinuation

Stopping a psychiatric medication abruptly is rarely safe. Many psychiatric medications — across the mood-stabilizer, antidepressant, antipsychotic, and benzodiazepine classes — can cause withdrawal symptoms, discontinuation effects, or relapse of the underlying condition when stopped without a planned taper. The specific risk profile depends on the drug and your history. Adherence data is for talking to your prescriber, not for self-managing a stop or a switch.

What this is not a substitute for

This is not pharmacology advice. Do not change doses or stop a medication based on adherence patterns alone; that decision belongs with your prescriber, who can plan a taper and watch for the things tracking alone cannot catch.