Closing out a large clinical trial: lessons from the prospective study of pravastatin in the elderly at risk (PROSPER).

Relatively little has been written about the practicalities of the closeout of large, multi-centre clinical trials, but this aspect of trial conduct and design is both important and requires careful planning in order to be accomplished in a timely and orderly fashion. Here, we document our recent experiences of closing down the Prospective Study of Pravastatin in the Elderly at Risk (PROSPER). This five-year study with over 5800 subjects was closed down and published in 2002. We describe the methods used to ensure the speedy progression from the start of the closeout period through data lock to publication and presentation of the results. We discuss the strategic planning of all aspects of the closeout process, the training of staff for final visits, the methods used to follow-up all study participants including those used in dealing with "difficult to contact" and defaulted patients, and the strategies employed to ensure that study participants were left with positive feelings about the study. We also detail the methods employed to expedite the throughput of study paperwork and endpoints and the cleaning of data in preparation for data lock and subsequent publication and presentation of the results. Based on our experiences we summarize the most important aspects of closeout design and make recommendations for future studies, the most important of which is that a well-planned and well-managed closeout is a key feature of any large scale clinical trial and a coherent and practicable closeout strategy should be an integral part of the design.