More on “Fast” and “Slow” Thinking in Diagnostic Reasoning

To the Editor: I was interested in Norman and colleagues’1 recent study on “fast” versus “slow” thinking and the related Commentary by Croskerry and colleagues.2 I agree with Norman et al that we need to move beyond the “vague distinction between System 1 and System 2 towards more precise models of diagnostic decision making.”1 It is clear that such a two-system approach to cognitive processing is a gross approximation, in many ways repackaging the age-old concepts of the conscious and unconscious mind, while adding little more. I disagree, however, with key points in Croskerry and colleagues’ critiques of this latest work. First, Croskerry and colleagues state that the external validity of Norman and colleagues’ study is so low as to make their conclusions “extremely tenuous,” yet in the absence of further comparative research, the actual external validity of Norman and colleagues’ experiment is unknown and does not necessarily rely on the experiment’s two study environments being equally realistic as Croskerry and colleagues suggest. Second, our modes of thinking are largely out of our voluntary control. High-level conscious attention is a limited resource, and evolution has given us a mind that appropriately uses this effortful system sparingly. Slowing down will not switch off “fast” System 1 processing or the conclusions that result from it. These are triggered automatically in any expert when he or she encounters recognized patterns. Without this, experts would be largely equivalent to novices, which seems entirely disadvantageous. Although self-awareness of the limitations of System 1 thinking can prompt reevaluation of the evidence (Croskerry and colleagues’ Table 1 is a useful summary), asking every clinician to reevaluate every System 1 conclusion sounds a lot like the exhortations to be more careful and to stop making errors that have been a demonstrable failure throughout health care for decades.3 Albeit highly trained, doctors are human, and normal human psychology therefore applies. The future of the work of improving clinical decision making lies in finding new practical ways to formally support and enhance doctors’ decision making, and this needs to be more substantive than telling doctors to try harder or to slow down. Craig Webster, PhD, MSc Senior lecturer, Centre for Medical and Health Sciences Education, University of Auckland, Auckland, New Zealand; [email protected]