By Aaron Abajian, MD
Retention is the quiet crisis of clinical research. Studies commonly report dropout rates around 30 percent, and every departure costs more than the empty chair: screening and enrollment spend, statistical power, sometimes an extra participant to recruit under a deadline. Sponsors write retention plans; sites live them.
Here is the useful reframe: most dropout is not a decision, it is drift. A moved apartment, a changed number, a missed visit that made the next one awkward, a growing sense that the study forgot about them. Drift is a logistics problem, and logistics problems respond to unglamorous, consistent communication far better than to motivation campaigns.
The three-message retention layer
- The visit confirmation: 3 to 5 days before every visit, with a one-tap confirm or reschedule. This is retention's front line: every rescued reschedule is a dropout that did not happen.
- The between-visit check-in: for protocols with long gaps, a short monthly touchpoint. "Still with us? Anything we should know?" Participants who feel remembered stay enrolled.
- The contact-info ping: every few months, "is this still the best number and address for you?" A bounced number found in month 4 is an update; found in month 14, it is a lost participant.
None of these messages carry study content. Date, time, location, a question, a link. That keeps the thread free of protected health information and keeps the IRB conversation short.
Silence is the signal
The retention value of SMS is not the sending, it is the answering. When every message offers a one-tap response, the roster becomes a live instrument: confirmed, needs reschedule, or no response, per participant, per visit. The no-response column is your early-warning system. A participant who stops answering reminders in month 7 is telling you something in month 7, not at the month 9 visit they skip.
The operational rule that follows: set a no-answer deadline on every confirmation, and route the flag to a human. A coordinator who gets a notification that Maria has not confirmed Thursday's visit makes one call and usually saves the visit. A coordinator who finds out Thursday at 9:15 writes a deviation note.
What this costs, honestly
A retention layer like this, run through a usage-priced service, is nearly free. An 80-participant study sending 6 to 10 messages per participant per month uses 480 to 800 texts, roughly $20 to $32 a month at credit-pack rates, with email delivery free. The comparison points, research messaging platforms at $99 a month and up, or coordinator hours spent dialing, are both more expensive than the entire text budget.
Running it without adding work
The playbook only works if it runs itself. With Cronote, a coordinator imports the roster from a CSV, writes each message once with merge fields for names, and schedules the sequence for the length of the study. Participants confirm or request rescheduling with one tap in their browser, no app. The roster shows who confirmed and who has gone quiet, and no-answer deadlines notify the coordinator automatically. Messages carry no PHI, and a Business Associate Agreement is available for research sites at cronote@cronote.com.