Running Recruitment From a Dashboard You Actually Use

Most sites already have recruitment data. What they lack is a small set of numbers they look at on the same day every week and actually act on. A dashboard isn't a reporting exercise for the sponsor — it's the instrument panel you use to decide where coordinator hours go next week.

Start with four numbers, not forty

The fastest way to kill a dashboard is to build it comprehensively. Pick the metrics that change a decision:

• Referrals in, by source. Raw volume per week per channel.
• Contact rate. Of referrals received, what share did you reach a live human on?
• Pre-screen pass rate. Of those reached, what share cleared the phone screen?
• Screening visit show rate. Of scheduled visits, what share happened?

Those four, plus consented and randomized counts, tell you nearly everything. If enrollment is behind, the problem lives in one of those stages, and the numbers point at it before anyone has to argue about it in a meeting.

Build the funnel as a conversion chain

A table of totals hides the diagnosis. Display each stage as a percentage of the one before it. A study with strong referral volume and a weak contact rate has an operational problem — call timing, staffing, or a phone number patients don't recognize. A study with excellent contact rates and a collapsing pre-screen pass rate has a targeting problem, or a protocol whose eligibility is narrower than the ad implies.

Sites commonly find that the stage with the worst conversion is not the one leadership assumed. Advertising spend gets increased when the real leak was a two-day lag between referral arrival and first call.

Add time, not just volume

Two time metrics earn their place on the panel:

1. Median hours from referral to first contact attempt. This is the single most controllable number in recruitment, and it moves show rates.
2. Median days from pre-screen pass to screening visit. Long gaps mean more cancellations and more time for a patient to lose interest.

Track medians, not averages — a handful of stale referrals will distort an average and hide a process that's mostly fine.

Make it per-study and per-coordinator

Aggregate site numbers are for the sponsor. Internally, you need the same funnel split by study, so you can see which protocol is consuming disproportionate effort per enrollment, and split by coordinator, so you can coach rather than guess. Use coordinator-level views carefully: differences in caseload and protocol difficulty matter, and a metric used punitively will get gamed within a month.

Set thresholds before you look

Decide in advance what triggers action. For example: contact rate below a set floor for two consecutive weeks triggers a review of call windows; referral volume from any channel dropping by half triggers checking that the form and phone line still work. Without thresholds, dashboards become wallpaper.

Run a fifteen-minute weekly review

Same day, same agenda: which studies are off pace, which funnel stage is responsible, one action per problem with an owner. Keep a short log of the actions and revisit them the following week. The log is what turns a dashboard from a view into a management system — and it's also the evidence base for your next feasibility questionnaire, since you'll have real conversion history instead of recollection.

One more habit: audit your data entry monthly. Dashboards built on referral records nobody updated after the first call will confidently point you in the wrong direction.