Historically, the CEME methodology evolved from the fundamental tenet whereby a quasi-chaotic relationship exists between experimental data and real world clinical outcomes. That is, small differences between physical properties of drugs result in significant differences in real world clinical outcomes.

For a deterministic system that is chaotic, final outcomes are not predictable and are sensitive to starting values. However, in a quasi-chaotic context-dependent system, by comparison with behaviour in a chaotic system small differences in starting values likewise result in large differences in final outcomes, but unlike a true chaotic system, the relationship between the initial conditions and the final outcome is fixed. In a quasi-chaotic system CEME displays exquisite sensitivity in differentiating drugs that share the same approved indication. For a specific drug, given that experimental data uniquely code for the drug’s real world clinical profile and patterns then within an approved indication value-differentiation is determined by some previously-unknown relationship between the drug and data that characterizes the approved indication. It is the clinical correlate of this previously-unknown relationship that shows with a high probability of success where value-differentiation (as real world evidence) resides within big data.