A Distributed Data Processing System for a Multicenter Clinical Trial

Abstract

A distributed data processing system was developed and implemented in the Systolic Hypertension in the Elderly Program (SHEP) pilot study, a multicenter randomized double-blind clinical trial. The SHEP system was based on minicomputer hardware and database management software that was extensively customized for our application. The system provided on-site data entry, editing, and retrieval at each of five clinical sites, and permitted local randomization of the study participants upon confirming eligibility. The system included protocol guidance and monitoring functions to assist clinicians in administering the complex, stepped-care treatment protocol. Finally, the system provided a communication network for messages and documents, as well as data, which has many advantages in a study involving multiple centers in various locations. Dispersing much of the data processing to the clinics in SHEP increased the access to and responsibility for the data at the clinic level, while simultaneously facilitating coordinating center monitoring and control of protocol administration. Distributed systems pose some special technical challenges, but have many advantages for use in collaborative studies.