Rewriting Pre-Screening Scripts So Good Candidates Stay on the Phone

The call script is where recruitment spend either converts or evaporates

Most sites put real money into generating leads and then hand those leads to a script that was written once, pasted from the protocol's inclusion criteria, and never revisited. The result is predictable: long calls, inconsistent data, and candidates who hang up polite but unconvinced.

A pre-screening call does two jobs at once. It qualifies, and it sells. Scripts that only do the first one leak eligible people.

Lead with why they called, not with your questions

The most common structural mistake is opening with demographics and diagnosis questions. The candidate reached out because something in their life isn't working — pain, fatigue, blood sugar, a kid's symptoms. Before you ask anything, name that.

Something like: "You reached out about the migraine study — can you tell me a little about what your headaches have been like lately?" You'll get eligibility data anyway, in the candidate's own words, and you'll get it while they feel heard rather than processed.

Budget sixty to ninety seconds for this. It pays for itself in completion rates.

Order your questions by exclusion power, not by protocol order

Protocols list criteria in a logical order for regulators, not for phone calls. Rewrite the sequence so the criteria that disqualify the most people come first.

In most site datasets, three or four criteria account for the bulk of pre-screen failures — often a lab value range, a washout requirement, a competing medication, or a hard age/BMI window. Ask those in the first three minutes. A call that ends at minute four is a good outcome; a call that ends at minute eighteen with the same result burned a coordinator's afternoon.

Practical rule: if a criterion knocks out more than roughly a quarter of callers, it belongs in the opening block.

Write the hard questions out loud

Sensitive items — substance use, psychiatric history, sexual activity and contraception, incarceration status — are where coordinators improvise, and improvisation produces inconsistent data and awkward calls. Script exact wording and normalize before asking:

• Normalize: "We ask everyone this, and it doesn't automatically rule anyone out."
• Be specific: "In the past 30 days, about how many days did you have four or more drinks?" beats "Do you drink?"
• Give the out: "If you'd rather discuss this with the study doctor instead of me, that's completely fine."

Build branches, not a flat list

A good script has forks. If someone fails a washout window, the branch isn't "thank you, goodbye" — it's a date calculation and a callback offer. If someone doesn't know a lab value, the branch is a records-release offer. If they're ineligible for this protocol but plausible for another, the branch is a warm handoff, with their permission.

Each dead end in your script is a lead you paid for and discarded.

Close every call the same way

Three things before you hang up, every time:

1. State the next step with a date and time, not "we'll be in touch."
2. Confirm the best contact method and get explicit permission for texts.
3. Set expectations for the visit — how long, whether fasting is required, what to bring, whether travel is reimbursed.

Revise the script on data, not vibes

Pull your pre-screen records monthly. Which question terminates the most calls? Where do people drop off mid-call? Which coordinator has a noticeably higher pass-to-visit rate, and what are they saying that others aren't?

Listen to a few recorded or shadowed calls per coordinator each month. Scripts drift. The ones that stay sharp are the ones someone owns and edits.