Data does not arrive in one neat package at the end of a trial. It arrives in a steady stream: a lab result from one site on Monday, a reported side effect from another site on Wednesday, a wearable reading every night. Each piece is recorded and stored as soon as it is entered. What does not happen automatically is the reading. In a traditional setup, a monitor reviews the accumulated data at fixed intervals — often every few weeks or once a month — checking entries against the protocol, looking for anything unusual, and resolving questions with the site. Between reviews, the data sits there. If a problem appears the day after a review, nobody sees it until the next one. That gap matters because many trial decisions cannot be made until the data behind them has been reviewed. A safety question, a decision to pause enrollment at one site, or a change to how a measurement is being taken all wait on someone having looked. The trial is not slow because the data is missing; it is slow because the review is periodic while the data is continuous.
A concrete version of the gap
Suppose a site's blood pressure readings start drifting upward for several participants in the same week. The readings are recorded correctly and stored immediately. If monitoring runs on a four-week cycle, those readings sit unexamined for up to four weeks. Only at the next review does anyone notice the pattern, ask the site what changed, and act. The four weeks were not spent collecting data — they were spent waiting for someone to look at data that already existed.