From Clues to Algorithms in Clinical Reasoning

Medicine has undergone a major shift in its dominant mode of clinical reasoning, a Kuhnian paradigm change centred on diagnostic method rather than disease theory. Until the mid-20^th century, clinical practice followed an experience-based model in which clinicians acted as detectives, assembling symptoms and signs as 'clues' to infer underlying pathology. This Holmesian style of reasoning relied on tacit judgment acquired through apprenticeship. In the mid-20^th century, this paradigm entered a crisis of confidence as empirical studies revealed extensive variability and error in clinical observations. These findings undermined faith in intuition and professional mystique. In response, medicine increasingly adopted numerical indexes, decision trees, flow charts, protocols, and guidelines. These tools replaced the spatial assemblage of clues with formalized, conditional 'if-then' logic that linked measurement directly to outcomes and action. The clinical algorithm has become the key exemplar of a new algorithmic paradigm, one that underpins contemporary evidence-based medicine and risk-factor management.