Rebuilding a Pre-Screening Script That Doesn't Sound Like a Form

The script problem most sites don't notice

Most pre-screening scripts start life as a copy-paste of the protocol's inclusion and exclusion criteria, reformatted into questions. That's why so many calls feel like an interrogation and why coordinators end up ad-libbing halfway through. The protocol was written to define an evaluable population. A phone script has a different job: figure out quickly whether this person is plausibly eligible, and leave them wanting to come in.

Those two jobs need different structures.

Order questions by cost, not by protocol section

Rank every question by two things: how many people it eliminates, and how uncomfortable it is to ask. Then front-load the high-elimination, low-discomfort questions.

In most site datasets, three or four criteria do nearly all the disqualifying — age band, a diagnosis confirmation, a key medication, and distance from the site. If those live on page two of your script, you're spending five minutes of goodwill on people who were never candidates.

A practical target: know whether the call is likely to continue within the first ninety seconds. Everything else — detailed medication history, prior therapy sequencing, comorbidities — moves later, once the caller has a reason to keep talking.

Ask about experience, not criteria

Candidates cannot reliably answer protocol language. Someone with a confirmed diagnosis may not recognize the term used in the inclusion criteria. Someone taking a prohibited medication may know it only by brand name, or only as "the shot I get every few months."

Translate each criterion into the way patients actually describe their lives:

• Instead of "stable dose for ≥12 weeks," ask "when did your doctor last change the dose?"
• Instead of listing prohibited drugs, ask "can you grab your pill bottles or your pharmacy list?" and read it back
• Instead of "prior biologic exposure," ask "have you ever been on an injection or infusion for this?"

Build a translation column into the script itself, so a newer coordinator isn't inventing phrasing on the fly. Include the common brand names and the common patient euphemisms next to each item.

Design for the maybe

Scripts usually handle clear yes and clear no. They rarely handle the answer that needs a chart. A good script tells the coordinator exactly what to do when the answer is uncertain: which questions to skip, what to request from the candidate, and what to promise about follow-up timing.

Write explicit branches:

1. Likely eligible — move directly to scheduling, while the caller is on the line.
2. Needs verification — collect the specific missing item, set a named callback window, and confirm the best contact method.
3. Not eligible now — check whether the reason is temporary, and ask permission to contact them about future studies.
4. Not eligible — close warmly, and note the disqualifying reason in a structured field.

That last point matters more than it looks. Free-text disqualification notes are unusable. A short picklist of reasons turns every failed call into feedback about the protocol and your outreach targeting.

Keep revising it

Read the recordings or sit in on calls monthly. Listen for the places coordinators consistently go off-script — that's usually the script being wrong, not the coordinator. Track which question first eliminates each candidate; if a question almost never disqualifies anyone, cut it or move it to the visit. Scripts should get shorter over the life of a study, not longer.