Abstract
O’Sullivan and Schofield gave an excellent summary of the problem of cognitive bias in the diagnostic process.1 Metacognition and use of Bayesian theory were highlighted as means to mitigate the effect of bias. We suggest that the routine use of the term ‘working diagnosis’ would draw attention to these strategies by emphasising the uncertainty that prevails. Even when all relevant tests have been made, some uncertainty may remain. The task of the clinician is, therefore, not necessarily to attain certainty, but rather to reduce the level of diagnostic uncertainty enough to make optimal therapeutic decisions.2 It is accepted clinical practice to compile a list of differential diagnoses – these each have varying degrees of plausibility. The ‘working diagnosis’ is simply the differential with the greater probability. By using the term ‘working diagnosis’, the clinician is challenged to re-evaluate the diagnosis in light of new evidence. In this way, anchoring bias, confirmation bias and premature closure bias may be avoided. Clinicians should embrace uncertainty whilst avoiding decision paralysis in the face of uncertainty.