Sociological Perspectives on Improving Medical Diagnosis Emphasizing CAD

The Process Of Diagnosis

Diagnosis is sociologically a process of bounded rationality, consisting of evidence gathering punctuated by a series of conjectures about a patient’s ills before arriving at the best conjecture and treatment plan. This view of diagnostics originates from Zeldenrust's (1990) version of the garbage can model of organizational behavior that sees diagnosis as a linkage between a demand (patient with an illness) and a solution (physician guesses nature of underlying problem and offers a plan for treatment). The solution may include medicine, behavioral modification such as bed rest, change in diet etc., and perhaps tests to rule out certain competing diagnostic hypotheses consistent with applicable constraints. The physician builds on the information learned not only from the patient reports but also on test results and on other objective data obtained. The physician treats symptoms uncovered until the patient either is cured and a final diagnosis is achieved or at least until all competing hypotheses about the condition are ruled out.

Improving the diagnostic process hinges on rapidly expanding the diagnostic information available to the physician. This diagnostic information is not just any data but the sort that expands the number of hypotheses of possible diagnoses, i.e. pivotal data. The physician can then examine the maximum number of plausible hypotheses. The idea is that the more hypotheses the physician needs to rule out at the outset the more likely the physician will find the disease(s) behind the patient’s complaint and provide the proper treatment (including referral for specialized services). The process also covers multiple possible pathologies that may exist at any moment, although the clinician may place ordering priorities on these possible multiple pathologies.