Retention Is a Communication Problem, Not a Compliance Problem

Most dropouts don't announce themselves. A participant doesn't call to quit; they stop answering, reschedule twice, then go quiet. By the time the coordinator notices, the relationship has already cooled. Retention work is mostly communication work, and it's almost entirely front-loaded.

Set expectations before the first dose

The highest-leverage retention conversation happens at consent. Not the consent form reading — the practical conversation after it. Walk the participant through what the next six months actually look like in their calendar terms: how many visits, which ones are long, which require fasting, which require a driver, which can flex by a week and which can't.

Sites that do this well hand over a one-page visit map the participant can put on their refrigerator. Dates, durations, prep requirements, and a phone number. When someone knows visit four is the four-hour PK day, they plan around it instead of discovering it the night before and cancelling.

Also name the hard parts out loud. If the diary is daily and tedious, say so. Participants who were told the truth tolerate burden far better than participants who feel ambushed.

Match the channel to the person

Ask at enrollment how each participant prefers to be reached, and record it where anyone covering the study can see it. Then honor it. Calling someone who only texts produces voicemail no one hears; texting someone who doesn't read texts produces silence that looks like disengagement.

A workable default pattern:

• Appointment reminders: text, with a reply-to-confirm option
• Prep instructions: text plus email, sent 48 hours out, repeated the morning before
• Anything sensitive or complicated: phone call, from a named person
• Missed visit follow-up: call same day, text if no answer, second call next day

The pattern matters less than the consistency. Participants learn your rhythm and start expecting contact.

Build in contact that isn't a request

Most study communication asks for something: come in, take this, fill that out. Participants who only hear from you when you need something start treating your number like a bill collector's.

Mix in contact with no ask attached. A check-in call after a difficult visit. A note when a lab value they asked about comes back normal. A card at the holidays. Study-level updates when enrollment closes or when results publish — participants consistently say nobody ever told them what happened with the research they gave a year to.

For long studies, a short between-visit call every six to eight weeks catches problems while they're still fixable. You hear about the side effect they were going to tolerate silently, the new work schedule, the ride that fell through.

Treat the first missed visit as an emergency

In most site datasets, the single missed or rescheduled visit is the clearest early warning you get. Respond that day, not next week. The call should be curious rather than corrective: what got in the way, and what would make the next one easier?

Then fix the actual obstacle. Move to a different time of day. Arrange transport. Consolidate two assessments into one trip if the protocol allows it. Document what you changed so the next coordinator doesn't undo it.

Make it one person's job

Retention degrades fastest when every coordinator assumes someone else called. Assign each participant a named primary contact, introduce that person by name, and have them make the non-visit calls. Participants stay in studies for people, not protocols.