Interpreting outcome data in hematological malignancies: a paradigm for clinical studies

Results of clinical studies are often contradictory in real time, and in other instances therapies may be adopted due to information from clinical studies where the data may be premature or resulting from small studies. Much of the data may have inherent selection biases, and their interpretation may be confusing and difficult. The hematological literature is full of such examples, and this review will describe some such instances in the hope of introducing both a cautionary note and encouraging more precise description of study conditions as well as an appreciation of the importance of allowing data from clinical studies to mature. Several examples will be drawn from clinical studies in lymphomas, leukemia, and bone marrow transplantation.