How to migrate from Castor EDC to Capture

A self-serve EDC from the mid-market is already faster than a services-led enterprise build, so teams that move are usually not escaping slowness. The questions we hear are about structure: how many lines make up the invoice, whether sites and users are priced individually, whether ePRO, eConsent and randomization are one platform or several, and whether a protocol amendment is something the team does or something it requests.

If those are your questions, put them to your current vendor first. Contracts vary, and the answer may be that nothing needs to change. If they are not satisfactory, our Castor EDC alternatives page is the neutral comparison and Capture vs Castor is the head-to-head. Use hidden EDC fees as a checklist for the quote you already hold.

Look at your agreement for the term, the notice period and any fee for exporting data at the end. These set the earliest practical date to move, regardless of what the technology allows.

Treat the Castor export as the record of everything collected before the cut-over. Export it as CSV with the audit trail, store it with your trial master file, and write a one-page variable map that pairs each old variable with its Capture field. When the study ends, the two datasets are combined at analysis, and the map is the only document the statistician needs.

Capture's exports are wide-format CSV or Excel, one row per submission, with an automatic data dictionary or codebook, raw and decoded files, and date-range and by-site filtering. That makes the merge a straightforward join on participant identifier. If your analysis plan needs consistent coding, agree the choice lists before you build the forms, not after.

Be careful with identifiers. Keep participant numbers the same across the two systems where the study continues, so one person is never counted twice, and record the rule in the data management plan.

The best moment to move is before a study starts, or between studies. The second best is a defined break inside one: a new cohort, an open-label extension or the next protocol. The worst is the middle of recruitment, when sites are comfortable with the old system and every change has a cost in queries and training. If you cannot avoid it, read how to switch EDC vendors mid-study first.

Run the sandbox rebuild in parallel with the current study. You will learn how long your own forms take, discover which logic needs simplifying and give the team confidence before the contract question comes up. That is a free, low-risk way to decide, and it costs nothing if the answer is to stay.