Abstract:
|
The Children's Oncology Group (COG) is a consortium of approximately 240 medical research facilities and physicians located primarily in the United States and Canada. However, the geographic range of institutions and physicians extends from Australia to Switzerland. The COG was formed in July 1999 by the merger of the then four NCI-funded children's groups: Childrens Cancer Group, Pediatric Oncology Group, Intergroup Rhabdomyosarcoma Study Group, and the New Wilms Tumor Study Group. That makes the COG the single NCI funded group responsible for protocol design, treatment, compliance monitoring, data acquisition, archival, and distribution for clinical trials for all children with cancer. A brief history of COG's evolution is presented with particular attention to the structure of its Department of Statistics and its Research Data Center. The role of group sequential statistical designs in COG studies is described, particularly for Phase II efficacy studies. Examples of Kepner/Chang designs are given along with theorems supporting the goodness and general applicability of these designs. Comments are made about the conduct of children's cancer clinical trials.
|