Clinical trials form the cornerstone of the science-based approach to improving patient outcomes. They provide the highest level of evidence about a treatment's safety and efficacy. Pre-clinical and observational data can provide useful hypothesis-generating information, but nothing can replace a pre-specified evaluation of an intervention in a well-defined context. These benefits are achieved only through careful planning.

To advance blood and marrow transplant (BMT) research, clinical trials play an especially important role in determining the best approaches for conducting transplant and optimizing its outcomes in patients with blood diseases. Transplant may be indicated for various malignant and nonmalignant hematologic diseases using either the patient's own blood cells (autologous) or a related or unrelated donor (allogeneic); using bone marrow, peripheral blood, or umbilical cord blood cells; with an assortment of choices of conditioning regimen agents and intensity; and with several possible prophylactic therapies to prevent infections and, for allogeneic transplant, graft-versus-host disease (GVHD) after transplant. With this myriad of options, a rigorous scientific evaluation is necessary to establish the best practices for administering transplants. Clinical trials include the systematic procedures for data collection and patient evaluation required for this evaluation of transplant approaches and can avoid biases that often arise in alternative study designs. The development of a BMT clinical trial requires special considerations, though, including the rare disease populations involved and appropriate endpoints and analysis techniques to evaluate transplant outcomes of interest.

This article focuses upon the scientific and statistical considerations for designing and conducting a BMT trial. We structure our discussion of the principles and practice of trial design by reviewing the sections of a typical protocol, which is the main document describing the design and planned conduct of the clinical trial. A recently completed trial, BMT CTN 1301 [1], will be used to illustrate the practical application of the concepts.