|
|
|
|
|
|
|
2.
The Data Cleaning Cycle |
|
|
|
|
|
|
|
|
We may not only lose data because of protocol violations. Quality of work varies widely between centers, so what contingency should we include to allow for query resolution and consequent delays to database lock? With most companies maintaining records of investigator performance, this is critical information to include. If queries are being tracked electronically, it is relatively simple to generate statistics on data query incidence and turn-around. However, these are traditionally used mainly to feed performance data back to centers during the study, rather than for planning new projects. Detailed information on study center quality performance is a powerful planning tool [2]. The author's company has excluded some high-recruiting centers from new studies because protocol compliance and data quality are so poor that many patients recruited are invalid. Interestingly, there is no evidence that so-called centers of excellence, the high-profile teaching hospital units, are any better in this regard, and their quality performance is usually inferior to a well-trained general practice center. |
|
|
|
|
|
|
|
|
E.
Clinical Trial Risk ManagementA Summary |
|
|
|
|
|
|
|
|
Clearly, any detailed examination of clinical trial risks could fill a whole chapter, and here we have discussed just some of those which have been prominent in the author's experience. No doubt, this section has raised more questions than answers, and this, indeed, is the essence of the message. Unless the project manager asks those questions, based on What happens if ?, the most elegant plans will be vulnerable to sudden and unexpected change or will be destroyed altogether. Perhaps this is why 90 percent of project management software purchasers just use it to do initial planning and never update their plansit would be too disappointing if they did! |
|
|
|
|
|
|
|
|
F.
Getting More from LessMultiple Projects, Priorities, Workload, Progress Control |
|
|
|
|
|
|
|
|
All of the complexity described so far would apply even if each person were involved with only one clinical trial. The reality is that most people are doing all of this for several trials, multiplying the problems and introducing new ones. This is substantially different from where project management grew up, in construction and engineering. The generally accepted view there is of a manager responsible for one project, although this may be anything from the local apartment block to the Channel Tunnel. Thus, clinical re- |
|
|
|
|
|