Without possessing and using a scientific mind, physicians are limited to functioning as technicians. Any case, not only the most difficult, can benefit from a scientific approach.
A patient with controlled diabetes or one receiving anticoagulant therapy, both of which require periodic monitoring and adjustment of their regimen, can be effectively cared for by a physician assistant or nurse practitioner. An unusual set of signs and symptoms that has eluded early success at diagnosis is likely to require an experienced physician who is familiar with the differential diagnosis of those findings and has the judgment necessary to approach the problem in an efficient, cost-effective manner. Some patients have conditions that are sufficiently exotic that they require a super-specialist and perhaps need to become a subject of medical research before the diagnosis and treatment can be determined. There is a point along this continuum at which a scientific approach becomes essential to further progress. Those not trained in the scientific method or not suited to it by temperament will not achieve the best outcomes.
At one end of the spectrum lie technology; applying well defined procedures and policies that are based on the evidence from scientific studies. At the other end is science. Hypotheses are formulated and tested in an iterative fashion until (hopefully) something of use is "proven" to the extent that it can be the basis of a diagnostic test or intervention (treatment). Success and failure both contribute to the formation of theories which, as long as they continue to withstand scrutiny, provide the basis for both inductive and deductive reasoning about unusual or complex situations.
It is the scientific process that is the vital element. Scientific results suggest a degree of certainty and predictability that may be unwarranted and that can stifle innovation and “out of the box” thinking whereas the genuinely scientific mind remains always skeptical - open to the unexpected and the, as yet, unknown. When applied to medicine this means spending more time asking questions and demanding answers and less time following the herd. Questions like: Why are we doing this? What is the evidence that it is safe? Is it the best alternative? Would thinking about and/or exploring the subject in more depth yield a new, different, better result than what we have so far? Consider that the practice of bleeding (venesection) has nothing that we today would consider to be a scientific basis. Modern scientific methods have been understood (at least to some extent) since at least 1850, if not before, but as late 1921 Osler’s Principles and Practice of Medicine states, “To bleed at the very onset [of lobar pneumonia] in robust, healthy individuals in whom the disease sets in with great intensity and high fever is good practice.”
As you can see, a scientific orientation is not a ubiquitous trait among physicians. If it were, a variety of practices having little more rationale than venesection would not still be commonly employed. Without possessing and using a scientific mind, physicians are limited to functioning as technicians. Any case, not only the most difficult, can benefit from a scientific approach. Unfortunately, skill at science can't be acquired overnight once it is recognized as necessary. Indeed, how would one who is untrained recognize the need? It has to be studied and practiced. Continuing medical education should foster this skill, a goal that is perhaps more important than what is commonly the focus of CME - "keeping up" with the latest medical factoid. Striking a balance between science and art is better for the patients and it can make the work more stimulating.