Norbert Donner-Banzhoff, University of Marburg, Germany

Abstract: When things seem to go wrong: Diagnostic error in primary care

In the area of therapeutics and management, reflections on past errors usually focus on conditions and causes preceding an event and possible improvements to avoid future incidents. Regarding diagnostic decision making, the main difficulty is to decide, whether an error has occurred at all. In the case of wrong-side surgery or prescription of drugs despite contraindications, wrongdoing is usually obvious. In these cases a behavioural clinical standard can be clearly described.

In diagnosis, however, the clinical standard there is difficult to define. This is even more pronounced in generalist settings such as hospital emergency departments or in primary care. In patients with identical presenting complaints, additional characteristics or symptoms may lead to widely differing disease likelihoods. As a result, appropriate next diagnostic steps differ considerably from patient to patient. Under these circumstances, the comparison standard would consist of a potentially unlimited algorithm.

Despite these difficulties, clinicians confronted with an undesirable outcome need support for their critical reflection on a case and their own diagnostic decision making. To this end, I have developed a heuristic to distinguish true diagnostic error from so called pseudo error – the Diagnostic Error Reflection Heuristic (DER-Heuristic).

I will present results from a qualitative survey of 30 primary care reporting cases of possible diagnostic error. Moreover, in an ongoing project we are evaluating the usefulness of the DER-Heuristic. The distinction between true diagnostic and pseudo error is important, because in the latter case learning, i.e. change of future clinical behaviour, is not indicated, it may even lead to over-diagnosis, over-treatment and waste of resources.