Using big data for cancer prevention
TL;DR
This research demonstrates the feasibility of using big data to conduct case-control analyses to identify associations between possible risk factors and cancers and demonstrates the limitations of the traditional epidemiological approach by requiring less effort, time, and cost.
Abstract
Background/Context: Cancer, a very prevalent and deadly disease, is expected to increase significantly to approximately 35 million new cancer cases and 17 million deaths by 2050. There has been progress in cancer research; however, this progress applies to relatively few cancer types, and current approaches to epidemiological research are limiting. Problem Statement/Objective: Ninety to ninety-five percent of cancers are due to external exposures, but only 30–50 percent of all cancer cases can be prevented by reducing exposure to risk factors. Therefore, there is a lot to learn about the causal risk factors of cancers. Methods: Using secondary EHR data, we identified twelve cancer groupings that we consider poorly understood (5,950 cases, and 17,850 controls) and conducted case-control analyses to identify associations with selected clinical and environmental exposures. Results: Radiation exposure, hormone therapy, breast implants, orthopedic implants, and formaldehyde exposure were associated with increased odds in the incidence of poorly understood cancers. Conclusion/Implications: This research demonstrates the feasibility of using big data to conduct case-control analyses to identify associations between possible risk factors and cancers. This methodology is the first step to identify causal risk factors between cancers and exposures. It can quickly provide guidance on which factors seem to have an association, and which do not. This can increase the rate at which the causal risk factors of cancers are identified and lead to an equitable reduction in the number of expected cancer cases and deaths. This methodology can be replicated with varying cancer types and exposures and enables research on rare conditions that was not possible due to small numbers. This methodology also overcomes the limitations of the traditional epidemiological approach by requiring less effort, time, and cost. The use of secondary data for research can produce results at a faster rate while incurring less cost and effort.