Reflections About Evidence-Based Pathology

Abstract

expert. In anatomical pathology, as at the bedside, the trajectory of diagnostic inference is guided by clinical instinct, constantly reinforced by experience. The more times we have solved the problem, the better we become at solving the next one. But if we are to incorporate evidence-based practice into diagnostic anatomical pathology, the choice among alternative diagnostic methods and workup algorithms must be an informed choice, informed not by anecdote (not even repeated anecdote) but by objective evidence distilled from well-designed studies. The link between evidence and diagnostic reasoning must be transparent (so as to be easily apprehended by those in training) and continuously refined, and the evidence used to support the conclusions should be periodically updated. In the clinic, much of the objective evidence used in EBM is derived from meta-analysis of carefully designed and conducted trials. Anatomical pathologists are in a unique position because deciphering the gross and histopathological appearance of lesions can be challenging and far from objective, although the image is considered objective relative to the clinical narrative. With some frequency, placing a lesion in a defined category involves weighing the morphological elements that compose the lesion differentially and integrating these data in the context of the information provided by other approaches (anamnesis, imaging, laboratory values, and outcomes). Choosing what morphological elements of a lesion to use in reaching a diagnosis and selecting which features will guide therapy comprise an act that can be based on experience or on EBP. Mixing the two in the right proportion is an exercise that every practicing anatomical pathologist does with a frequency dependent on the complexity of the practice environment. A vast body of scholarly work in pathology provides data on the diagnostic significance of an entire universe of morphological features for many diseases and serves as the basis for EBP. When only personal experience is used in making a diagnosis, with little reference to Asignificant part of a recent meeting of directors of anatomical pathology was devoted to discussing evidence-based pathology (EBP), a subject that was comprehensively treated in an issue of Seminars in Diagnostic Pathology. This may in part reflect a sense that anatomical pathology needs to catch up with other medical disciplines in terms of evidence-based medicine (EBM) (ie, “the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients.”) Evidence-based medicine has emerged as a favorite teaching tool to instruct medical students and housestaff in the use of objective and rational data as the basis for deciding between alternative diagnostic, prognostic, or therapeutic choices. In EBM, clinical instinct and intuition are not so much replaced as aided by the critical use of evidence (mostly derived from randomized trials), an evaluation of its strength, and patient-based inference that is carefully reasoned along the principles of logic. The act of clinical diagnosis has traditionally rested on the capacity of the master clinician to identify, among the noise, the few cardinal elements that constitute a pattern (the disease entity) and to prove that he or she is correct judging by the efficacy of the therapy or the accuracy of the prognosis. The more noise, the more difficult it is to formulate the correct diagnostic hypothesis, and in complicated cases a final conclusion is only reached after considering several hypotheses and choosing the one most likely to be corroborated by objective data. The correct final diagnosis usually explains all the noise, and the cloud of data that may have confused the neophyte disappears under the light shed by the