Abstract 5301: Cancer causation: rethinking the role of analogy

Abstract

Abstract Introduction. Fifty years has passed since the publication of Austin Bradford Hill’s classic paper on causal inference. Much has changed in the biomedical and public health sciences regarding cancer causation during that half century, but Hill’s considerations (often called “criteria”) remain durable. That said, there is a curious phenomenon involving Hill’s seminal work that has escaped the attention of the scientific community. His consideration of analogy—the last on his list just after experimentation—has all but disappeared. The purpose of this paper is to rethink and resurrect analogy as an important consideration in causal inference in environmental epidemiology. Methods. Given that this is a conceptual—non-quantitative—project, the methods used are the informal logic of scientific thinking, methods of systematic reviews, and judgment. Results. In contemporary epidemiology, analogy is either completely ignored (e.g. in many textbooks), equated with biologic plausibility or coherence, or aligned with the scientist’s imagination. None of these examples, however, capture Hill’s description of analogy which emphasizes the contrast of two bodies of evidence. Coupled with developments in the methods of systematic assessments of evidence—including but not limited to meta-analysis—analogy can be restructured as a key component of causal inference. For theoretical support see similar methods in biomedical ethics and artificial intelligence. Practical examples will be discussed. Conclusions. Analogical reasoning—an updated version of analogy—can emerge assuming the following developments: (1) a collection—a library—of known cases of causal inference (i.e. bodies of evidence involving established causal relationships), including, for example, the assessments of the International Agency for Research on Cancer (IARC) and the National Toxicology Program (NTP), (2) a systematic process for describing the key features of a causal relationship creating paradigm cases, and (3) an objective comparison of a “new” body of evidence for a putative causal relationship with a relevant paradigm case. Citation Format: Douglas L. Weed. Cancer causation: rethinking the role of analogy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5301. doi:10.1158/1538-7445.AM2017-5301