Screening Workload & Cost Calculator
Every "how long does screening take" resource we could find is an article citing one fixed average from one study — useful for context, useless for planning your review, because your record count, your team size and your reviewers' actual speed are never the ones in the article. None of them let you run your own numbers, and none of them separate out the time spent resolving disagreements between two independent reviewers from the time spent screening in the first place — which is real, recurring work that a single per-record average quietly erases.
The defaults below are placeholders, not published estimates
We did not fetch or verify a "typical" screening-speed figure from any study, so none is baked
in as a default the way a published rate would be. Replace every field with a number you
actually have — from a timed pilot batch, from your own team's last review, or from your
protocol. The arithmetic is what this tool contributes; the numbers have to be yours.
What this assumes
- Dual-independent means every record is screened twice, separately. The conflict rate is only the share of records where the two reviewers disagree and need a third opinion or a discussion — not the full record count again.
- Calendar time divides total reviewer-hours by team size × hours per reviewer per day. It assumes work can actually be split across the whole team in parallel; it does not model onboarding, holidays, or reviewers joining partway through.
- Cost is a single flat rate applied to every reviewer-hour. It does not account for reviewers at different pay levels or for time spent on anything other than screening and conflict resolution.
- Nothing here estimates how many records your search will actually return or what fraction will advance to full text — those are inputs you supply from your own protocol, not outputs this tool predicts.