Selecting representative case vignettes for clinical judgement studies: Examples from two heart failure studies

摘要

In the Clinical Judgement Analysis (CJA) tradition there are strong arguments in favour of using representative case vignettes. Due to methodological and practical constraints, however, it is often necessary to make compromises regarding representativity. In this article, we discuss our experiences in working with these problems in connection with two studies about general practitioners heart failure diagnoses. CJA is a method for investigating and describing decision-making in which both judges and judgement situations are taken into consideration. This means that when we want to study physicians diagnoses, both the participating physicians as well as the case vignettes should be sampled to be representative for the diagnostic situation. If representativity is interpreted in a strictly statistical sense, the case vignettes should reflect a well identified population of patients regarding values, distribution and intercorrelations of the relevant patient characteristics (variables). Creating a set of representative case vignettes meeting all theoretical requirements will often be difficult. Selecting a relevant population and relevant variables, deciding about the appropriate number of variables and case vignettes, and deciding whether authentic patients or constructed case vignettes should be used are some of the problems that must be dealt with.