page 30 The need for relatively large numbers of patients in order to gauge the effectiveness of treatment is readily appreciated when you consider the following hypothetical example. Suppose a clinic treats 400 patients a year, that there are five principal types of cancer, and each may be classified in four stages. in each specific”group. On the average, there would be 20 patients If you wished to change the treatment, you copuld have 10 in the new group and 10 in the standard one. If you expect’“am survival to be improved from 5 owtit of 10 (50%) to 7 of 10 (70%),it would be difficult to establish. Obviously, much larger numbers of patients and good statistical analysis would be required. Berge: Next question was, you said that you had been brought to Harvard by Shields Warren, and I was wongdering if you could talk a little bit about him. Kohn: Shields Warren was chief Pathologist at the Deaconess and Professor ed of Pathology at the Harvard Medical School. He became interest, in ; % pathology A y mfimradiation mmmk/in the late thirties owing to his interest in cancer. In 1939 he bacme an officer in the Navy's medical department, and circa 1942, with Dunlap, Gates and Friedman, wrote a series of papers summarizing what was known about radiation pathology.He was in the first team to visgit Nagasaki and Hiroshima after the bombing, and I believe he was the primary instigator of what THE Atomic Bomb Casualty Commission. later tunred into the MMI MVMEMNMMMMNOMM. When the AEC was established in 1947, he became the first director qg@i=the—Divhwmes of the Division of Biology & ~“N Medigfce, a post which heheld until they could find a permanent appointee. uv Dr. Warren had establish, is own Cancer Research Institute at the New England Deaconess Hospital, and looking back upon it now, I suppose he wanted