Returns and Adherence Data

Part of Trial Pharmacy (PHARM) Training: Dispensing and the Participant.

About 2 minutes. Free, in your language, no account needed.

Why this matters

What comes back is the measurement. Counting returns is how a trial knows what was actually taken, and pharmacy usually performs that count.

A count done casually produces an adherence figure that nobody can rely on.

Count before you accept

Count in front of the participant where practical, or at least before the packaging leaves your control.

Once a bottle is in a bin or a returns box, the count cannot be repeated and any discrepancy becomes unresolvable.

Record the count, not the conclusion

If sixty were dispensed and eleven return when four were expected, record eleven.

Do not adjust for what the participant explains. Record the count and the explanation as two separate facts, and let the discrepancy stand in the record.

A reconciliation that balances because somebody smoothed it is not a reconciliation.

Empty packaging matters

Participants routinely discard empty blisters and bottles, which destroys the count.

Say it at every dispensing, not just the first. Give them a bag. Make it easy, because an instruction given once at randomization is an instruction most people will not remember in week twelve.

What a discrepancy means

It does not automatically mean non-adherence. Product is lost, spilled, left at a relative's house, or taken and the packaging thrown away.

It means the measurement is uncertain, and that uncertainty belongs in the record rather than being resolved by assumption.

In practice

A participant returns a bottle with four more tablets than expected and says they forgot doses on a holiday.

You count, record the number, record their explanation separately, and inform the coordinator so the adherence data and any protocol threshold can be assessed.

You do not reassure them it does not matter. It may.

What this looks like when it goes wrong

A pharmacy records returns as matching expectation when they are close, to keep the log tidy.

At close-out, the sponsor cannot distinguish sites with genuinely high adherence from sites that rounded. The adherence variable for the whole study is weakened by a habit that felt like neatness.

Key takeaways