Avoid procedure bloat, which occurs when a clinical trial protocol is prepared by taking the previous one and adding some additional tests. This is fairly common as a drug passes through the development path.
As soon as patient enrollment decreases below projections by a fixed number determined in advance of the trial, seek to learn the cause. Determine if this is evident at all or only some sites (i.e., is it site-specific or protocol-specific). Either adjust the protocol or change the recruitment strategy. Do not simply add sites until you determine that that is a reasonable action and is likely to address the problem.
Monitors must understand the intensity and latitude they are allowed in their monitoring activities. Sometimes it may only be necessary to monitor the high enrollers, or a random group of sites in large trials. In mega-trials it is impossible and not regulatoraly required to monitor these trials as extensively as pivotal trials.