Compliance Essentials for Texting Trial Participants

Texting is the highest-response channel most sites have. It is also the channel where a sloppy setup creates the most exposure — regulatory, contractual, and reputational. Here is what actually needs to be in place before you send the first message.

Consent is two separate permissions

Sites routinely conflate two things that are legally distinct:

1. Permission to be contacted at all, usually collected on a pre-screening form or during a phone call.
2. Permission to be contacted by automated or bulk text, which in the U.S. falls under telemarketing and consumer protection rules and generally requires explicit, documented, written-form opt-in for anything sent through an automated platform.

An interested caller who verbally says "sure, text me" satisfies the first, not necessarily the second. Build a checkbox into your web form with plain unbundled language: what kinds of messages they will get, roughly how often, that message and data rates apply, and that they can reply STOP at any time. Do not pre-check it. Do not make it a condition of pre-screening.

Log the consent so you can produce it

If someone complains, you will need to show when and how permission was given. Your record should capture the timestamp, the phone number, the exact consent language shown at that moment, and the source (form URL, coordinator entry, call log). Screenshots of your current form are not evidence of what a person saw eighteen months ago. Versioning your consent text matters.

PHI belongs out of the message body

Texting is not a secure channel by default. Keep condition names, diagnoses, medication names, and test results out of outbound texts to anyone who has not been consented into the study. Even after consent, be conservative.

Safe patterns:

• "Hi Maria, this is Dana from Northside Research about the study you asked about. Do you have 5 minutes today?"
• "Reminder: your visit is Thursday at 9am. Reply C to confirm."

Risky patterns include naming the indication in a first-touch message to a household phone, or referencing lab values. Assume someone else may read the screen.

The IRB has to see it

Any message that recruits, describes the study, or influences a participant's decision is recruitment material. That means IRB review — including your text templates, the character-limited versions, and typically the consent language on the form that captures the number. Submit templates as a set with placeholders rather than one-off messages, so you are not filing an amendment every time you rephrase a reminder.

Retention-phase texts (appointment reminders, visit-window nudges) are usually reviewable too, and IRBs vary. Ask rather than assume.

Operational guardrails that prevent most problems

• Honor STOP instantly and permanently. Opt-out must propagate across every study and every list, not just the one that triggered it.
• Set quiet hours. No sends before 8am or after 9pm in the recipient's local time zone, and be careful with numbers whose area code no longer matches where the person lives.
• Identify yourself in every first message. Site name and a human name. Unattributed texts get reported as spam, which damages your sending reputation for every study.
• Cap frequency per person. Two or three touches over a week, then stop.
• Never text minors or legally authorized representatives without confirming who owns the number.

Train coordinators on the improvised message

Most violations do not come from campaign sends. They come from a coordinator typing a helpful, unapproved sentence from a personal phone at 7pm. Give staff a small library of approved snippets, a site-owned number, and a clear rule: if it is not in the library, call instead.