Abstract

Mario Bunge’s medical philosophy emphasizes the importance of mechanismic models in guiding the design, analysis, and practical application of clinical research. By contrast, the Evidence-Based Medicine movement regards mechanismic hypotheses as “evidence” dissociable from, and of secondary importance to, the findings of experimental research. In agreement with Bunge, it is argued here that mechanismic models and mechanismic thinking play essential roles in both clinical research and practice. Mechanismic models in medicine view health and disease as emergent processes occurring in complex biological systems and draw upon established scientific knowledge from multiple disciplines to help identify and control parameters that have decisive effects on clinical outcomes. Models play an essential role in designing efficient and reliable population-based studies, and in detecting and correcting for random error and systematic bias in clinical research. They are important both for extrapolating the results of clinical research to novel contexts and for tailoring interventions to the specific circumstances of an individual case. Contrary to the subordinate status they are accorded by EBM, empirically-validated mechanismic models should constitute the foundation of a scientific approach to medicine.