Diagnostic Case Study Research

Conceptual Overview and Discussion

Diagnostic case studies focus their inquiries at the level of the individual and involve a systematic search for pathology of some sort. The primary goal of diagnostic case studies is classification. Scientific classification requires that a given phenomenon—in this case, a disease—be identifiable in a manner that distinguishes it from other phenomena. By identifying distinguishable disease entities researchers who perform diagnostic case studies contribute to a body of science that allows other researchers and practicing clinicians to communicate more effectively. This enhanced communication, in turn, fosters a deeper understanding of a disorder so that practicing clinicians can more effectively assess and treat patients.

Beyond an intensive focus on the individual, the essential features of diagnostic case studies are (a) systematic assessment, (b) repeated measurement across time, (c) clear operational definitions, and (d) replication. Systematic assessment includes using standardized procedures and well-validated instruments. Repeated measurement is required to reduce the likelihood that extraneous factors are influencing the outcome of the study. Operational definitions precisely specify how a variable is measured. Without clear operational definitions, replication by other researchers is not possible, and all scientific knowledge should be replicable or else its validity is questionable.

The diagnostic process formally begins once all pertinent data have been collected. The data collection process requires an extensive personal [Page 292]interview and, ideally, formal testing. Modern diagnostic endeavors in mental health are based on the principles of parsimony and hierarchy. The principle of parsimony requires that one account for all available data with the fewest possible diagnoses. A single diagnosis that accounts for all observations is always the preferred option. The principle of hierarchy notes that mental illnesses cluster in a hierarchy of syndromes that tend to vary in severity. The most severe disorder that accounts for all the data is the most parsimonious diagnostic option. In practice, diagnostic endeavors follow a decision tree model. A diagnostic decision tree is a flowchart that allows a clinician to identify the most parsimonious diagnosis that fits the available data. Several decision trees are often required before one can arrive at the most parsimonious diagnosis.

In an applied setting (i.e., a doctor's office) the case study subject is the patient, and the data referred to earlier are called signs and symptoms. Symptoms are the complaints reported by a patient (e.g., a headache, hearing voices) and therefore are not directly observable. Signs, on the other hand, are directly observable and measurable (e.g., crying, a rash). Because symptoms cannot be directly observed, they are less reliable and carry less weight when rendering a diagnosis. Until very recently, diagnostic endeavors depended heavily on the skill of the clinicians or researcher and thus were often unreliable. Empirically validated diagnostic tests have now been developed for many conditions, adding considerable objectivity and rigor to the diagnostic process.

...