How to calculate a clinical trial visit window.

The basic calculation

You need three things from the protocol: the anchor visit, the number of days from that anchor to the target, and the allowed window.

• Target date = anchor date + days to target
• Earliest date = target date - window
• Latest date = target date + window

A worked example

Say the protocol sets Visit 3 at 28 days after baseline, plus or minus 3 days, and the subject's baseline visit was March 3.

Where teams get tripped up

• Day numbering: if baseline is Day 1, then Day 29 is 28 days later, not 29
• The wrong anchor: check whether windows run from baseline or the previous visit
• Actual vs planned dates: a late visit can move later windows if they anchor on the date it happened
• Other rules: some protocols use windows based on age, or different windows per visit
• Weekends and holidays: check that your office is open before the window closes

If a visit falls out of window

An out-of-window visit is generally a protocol deviation. Document what happened and why, and follow your sponsor's and IRB's reporting requirements. Check the protocol too, because some visits are marked as optional.

Let the system do the math

In PAS OneCore, windows are calculated from the schedule of activities, not typed in. Relative windows re-anchor on the date a visit was performed, and visits outside their window get flagged.