Rewriting Pre-Screening Scripts So Candidates Stay on the Line

Most pre-screening scripts start life as a protocol summary someone pasted into a Word doc. They get the eligibility questions right and everything else wrong. The caller sounds like they're reading a form, the candidate answers in monosyllables, and a person who might have enrolled hangs up feeling interrogated.

A good script does two jobs at once: it filters accurately, and it keeps interested people interested. Here's how to rebuild yours.

Open with why you're calling, not who you are

The most common opening failure is a long identification preamble — site name, study number, sponsor, IRB reference — before the candidate knows why the phone rang. People decide whether to stay on the line in the first eight seconds.

Better structure:

1. Name and reason. "Hi, this is Marcus from Riverside Clinical Research — you filled out a form about a migraine study last Tuesday."
2. Time ask. "Do you have about eight minutes? I have some questions to see if you'd be a fit."
3. What happens next. "If it looks good, the next step would be an in-person visit, and I can tell you more about what that involves."

That third piece matters more than sites expect. Candidates commonly disengage mid-call because they can't picture where this is going.

Order questions by likelihood of exclusion, not by protocol order

Protocols list inclusion criteria in clinical logic order. Your script should run in knockout order — the criteria that disqualify the most people go first.

If your washout requirement or an excluded concomitant medication eliminates half your callers, ask about it in the first two minutes. Don't spend six minutes on demographics and symptom history before hitting the wall. This respects the candidate's time and dramatically reduces your average call length across the funnel.

Build the order from your own data. After fifty pre-screens, look at which question produced the most stops and move it up.

Write the questions the way people actually answer them

Protocol language does not survive contact with a live caller. "Do you experience prodromal symptoms?" gets you a confused pause. "Do you ever get warning signs before a migraine starts — like seeing spots, or feeling off an hour beforehand?" gets you a usable answer.

A few rules that hold up:

• Avoid leading questions on exclusionary items. "You're not taking any blood thinners, right?" invites a no. Ask open: "What medications are you taking right now, including over-the-counter?"
• Use anchors for frequency and duration. "How many days in the last month?" beats "Do you have them often?"
• Never ask a yes/no question where the honest answer is "sort of." Give ranges or examples.

Script the exits, too

Most scripts end where eligibility ends. That leaves coordinators improvising the hardest sentence of the call. Write out three closings in advance:

• Qualified: schedule on the call, confirm the address, state what to bring, and set the reminder cadence out loud.
• Not qualified for this study: name the reason plainly, ask permission to keep them on file, and mention what other conditions you study.
• Unsure or needs to think: agree on a specific follow-up time rather than "I'll reach out."

Test it on the phone, not on paper

Read the script aloud before it goes live. Anything you stumble over, rewrite. Then have two coordinators run the same script for a week and compare where calls die. Scripts should be revised monthly in the first quarter of a study, not written once and filed.