Protocol-writing support conferences for investigator-initiated clinical trials

Abstract

In investigator-initiated clinical trials, protocols with inappropriate methods might cause bias. However, insufficient data are available to determine which items are important or difficult to discuss in protocol development. We recorded protocol-writing support conferences to determine what items methodologists and investigators discussed. We obtained approval from all applicants to attend our Intelligent Clinical Research and Innovation Center writing support conferences, recorded all the discussions, characterized them, and sorted the items iteratively. In 1 year, we had 18 conferences: nine early protocol conferences and nine rejected protocol conferences. The latter were rejected by the institutional review board, which requested consultation. The most discussed item was outcomes, accounting for ∼20% of the total discussion time. In three trials, the main problem was multiple primary outcomes. The second most discussed item was control. Early protocol conferences had more non-preliminary proposal items than rejected ones (P<0.001). This study showed important items (especially outcomes and control) for investigators to write protocols. Early protocol-writing conferences helped investigators find questionable items.