Causal Inference for Definitive Clinical End Points in a Randomized Clinical Trial With Intervening Nonrandomized Treatments

trials in oncology in the United States.Thirteen years ago, the Armitage report 9 highlighted the need to reduce the "Tower of Babel" in clinical cancer research, which, unfortunately, the most recent report from the Institute of Medicine also highlights.Again, the history of the US automobile industry shows an eerily similar situation.In the 1980s, the US auto industry received the first of many early warnings of a major challenge from overseas.10 In the early 1990s, the industry received additional warnings that were only partially heeded.11 Fewer than two decades later, two of the Big Three were bankrupt.Let us hope that oncology clinical trials in the United States do not follow the same path.Finally, it will take more than evidence, recommendations, and hope to solve the existing problematic situation with clinical cancer research in the United States.The early warning signs are there; the canary is not looking at all well.Other seemingly dying industries have overcome major international competition to come back even stronger, semiconductor chip manufacturers being one example.Oncology clinical research in the United States can choose to follow this path instead of that of the US automakers.But it will take commitment from all major stakeholders to set aside differences and to concentrate on finding practical solutions that work for all.