The Perception Approach: Level of Observer Variability Rating and Fusion Diagnoses.

ABSTRACT

Diagnostic uncertainty is a problem faced daily in clinical medicine. This uncertainty has many guises, such as vague wording in reports or pseudo-exact statements. High rates of intra- and interobserver variability and a corresponding low level of reproducibility are, unfortunately, issues that have to be taken into consideration in many medical fields. We hypothesise that the level of observer variability can be predicted in specific situations during the on-going assessment/diagnostic process and that the application of a communication concept based on Fusion Diagnoses might lead to improved patient treatment. A Fusion Diagnosis consists of the diagnosis itself combined with the estimated Level of Observer Variability (LOV) of the diagnostic process which is evaluated by the individual decision-maker during the on-going assessment process stated in parenthesis. Currently, the "Perceived Difficulty" of the investigator in a three-tiered grading system might be the most promising approach for the operationalisation of the LOV in the diagnostic process (D1 = easy to assess, D2 = average, D3 = difficult to assess). An example of a diagnosis according to this concept could be Dysplastic Nevus (D3). Based on this concept, standardized clinical decision pathways could be investigated and developed. Studies would have to provide data about rational decisions in relation to the estimated LOV. For example, data might reveal that "ST-elevation myocardial infarction (D3)" as well as "no ECG abnormalities (D3)" are necessary indications for an immediate consultation of a colleague in the setting of suspected myocardial ischaemia. The concept is hypothetical and critical aspects, like over- and underestimation and the calibration of the physician, are likely limiting the benefit. Nevertheless, we assume that the suggested pragmatic concept might lead to superior patient treatment in specific fields.