As a former Director of Biostatistics at the University of Maryland Cancer Center, Dr. Thompson was responsible for developing statistical designs for phase I, phase II, and phase III studies. These designs included studies on leukemia, solid tumors, and infectious diseases coincident with profound, persistent neutropenia. Phase I studies generally included development of maximum tolerated doses using a modified Fibonacci search. Phase IIa studies were designed to indicate clinical activity of new antineoplastics, while larger phase IIb and phase III studies were usually conducted within oncology groups. Dr. Thompson also has input on those studies.