Freud S, James W.

Croskerry P, Evans JS, Kahneman D, Mamede S, Sloman SA, Stanovich KE, Wason PC

However, faculty members have also noted that Jason struggles when patients present atypically, or when they present with multiple complaints.  In those cases, he tends to “shot-gun a bunch of tests” without giving much thought about his differential diagnosis.  The residency leadership is also concerned because he scored well below his classmates on this year’s in-training exam.

The residency leadership want to intervene to bring Jason up to the right standard, but are struggling to identify the flaw in his reasoning abilities, particularly because he seems to excel so well in some areas and lag behind in others.  What might be Jason’s problem?

Dual Process Reasoning is a theory which has evolved over the past century, dating back to Sigmund Freud, who suggested that reasoning involved two separate systems for information processing: one unconscious and associative, the other rational and conscious.  The current framework for this system, described by Stanovich and West, identifies two systems of thinking, which individuals alternate between depending on their own pre-existing framework and characteristics.¹ System 1 is automatic and unconscious and relies heavily on contextual clues. System 2, on the other hand, is conscious and analytical and is used independent of context.^2-4  It is proposed that an individual’s cognitive abilities are dependent on the degree to which they employ each of the two systems.  This results in two types of intelligence: analytic intelligence (System 2) and interactional intelligence (System 1). It was proposed that individuals with higher cognitive abilities employ more System 2 thinking (i.e. they are able to rely more on their analytic abilities and less on heuristics available in System 1).¹ Some believe that each of these two systems evolved to serve a different evolutionary purpose.  System 1 evolved to serve the process of reproduction and gene propagation, while System 2 evolved to serve “the interests of the whole person.”

The distinction between intuitive and analytical thinking was first described by the Greek philosophers. The ancients believed that the intuitive thought process was “actually a superior state of mind.”¹ However, in the early twentieth century, French philosophers described the intuitive thought process as unconscious, highly subject to bias and sometimes irrational.¹ The dual process theory is one of many approaches towards decision making.  However, the dual process theory is highly regarded and two Nobel prizes have been awarded for the scientific work regarding human decision-making (Herbert Simon in 1978 and Daniel Kahneman 2002). In 2009, Pat Croskerry published a manuscript in Advances in Health Science Education entitled  “Clinical cognition and diagnostic error: applications of a dual process model of reasoning.”³ This manuscript describes the application of the dual process model in clinical practice.

Dual Process Theory has informed models of clinical reasoning that are generally accepted in the medical education literature.  There is general consensus that both Type 1 reasoning (which can be thought of as pattern recognition) and Type 2 reasoning (or analytic reasoning) occur in clinical decision making.  However, there are differing opinions regarding the association between reasoning type and cognitive errors, as well as how applications of dual process learning may be able to reduce or prevent errors.

Croskerry posits that most medical errors are a failure of cognitive reasoning and over-reliance on System 1 processes and cognitive biases that result.  He suggests that clinicians should focus on reducing medical errors and cognitive biases and that physicians should make every effort to access system 2 to force analytical thinking in a hectic clinical environment.³ However, system 2 is a slower and often inefficient system.  How does one increase the efficiency of system 2 and decrease the error rate of System 1?  Croskerry believes that clinicians need to improve their ability to recognize patterns and be keenly aware when a scenario does not fit an established pattern.³

Daniel Kahneman, in “Thinking Fast and Slow,” argues that “memory also holds the vast repertory of skills we have acquired in a lifetime of practice, which automatically produce adequate solutions to challenges as they arise, from walking around a large stone on the path to averting the incipient outburst of a customer. The acquisition of skills requires a regular environment, an adequate opportunity to practice and rapid and unequivocal feedback about the correctness of thoughts and actions. When these conditions are fulfilled, skill eventually develops, and the intuitive judgments and choices that quickly come to mind will mostly be accurate.”⁴

In accordance with Dual Process Theory, many of today’s undergraduate and graduate medical education programs focus on early patient encounters, analysis of clinical cases and simulation.

In a recent Perspectives piece, the literature describing the relationship between Type 1 and Type 2 processing and reasoning errors is examined, and two theories regarding errors are explored: 1) All errors originate from heuristics employed in Type 1 reasoning and not corrected by the application of Type 2 reasoning; and 2) That errors arise from both processes and errors may be reduced by increasing knowledge.⁵ The authors conclude the errors can arise from both types of processing, and they do not find evidence to support that errors can be reduced by training physicians to recognize biases or warning them to “slow down” or try to force Type 2 reasoning.  However, there is some evidence that reorganizing knowledge and addressing knowledge gaps may decrease diagnostic error.

Outside of the clinical setting, dual process theory can be applied to learner performance on simulated cases in written or computer exams, oral presentations and low- or high- fidelity scenarios.

In the clinical setting, educators should be aware that learners likely use both types of processing, and exploring diagnostic reasoning strategies may give insight to knowledge gaps.  Addressing these gaps may reduce the possibility for error.  Establishing evidence-based clinical pathways may help reduce errors by encouraging adherence to guidelines and decreasing the risk of error due to knowledge gaps.⁴

Kahneman stated, “The way to block errors that originate in System 1 is simple in principle: recognize the signs that you are in a cognitive minefield, slow down, and ask reinforcement from System 2.”⁴ Although the simplicity of this principle is appealing, recognizing that one is in a minefield, and slowing down in the chaotic clinical environment of the ED is likely to prove challenging.  Further, it is not clear that forced analytic reasoning reduces errors or improves performance.⁵  There may be other benefits to teaching learners how to incorporate checks and balances into their clinical routine, particularly in the form of checklists and evidence-based guidelines.  As our understanding of how Dual Process Reasoning Theory can be applied to teaching and assessing clinical reasoning, improving performance and error reduction, further educational applications may emerge.

Stanovich KE, Toplak ME, West RF. The development of rational thought: A taxonomy of heuristics and biases.  Advances in child development and behavior. 2008;36:251-285.

This comprehensive chapter on rational thought from a psychology perspective reviews theory regarding rational thought, or “adopting appropriate goals, taking the appropriate action given one’s goals and beliefs and holding beliefs that are commensurate with available evidence.”  Rational thought is defined, dual-process theory is described and the authors provide a taxonomy of rational thinking errors, classify heuristics and biases, and provide exemplary studies of categories of errors.

Croskerry P. Clinical cognition and diagnostic error: applications of a dual process model of reasoning. Advances in Health Sciences Education. 2009;14(1):27-35.

In this paper, Croskerry describes Dual Process Theory as it applies to medical decision making, presents a schema for how approaches to decision making fit within the intuitive-analytical continuum.  He compares and contrasts System 1 and System 2 thinking and proposes applications of the model, including: a template for teaching decision theory; a platform for future research, and a proposal for how errors may occur in diagnostic processes.

Kahneman D. Thinking, fast and slow. Macmillan; 2011.

This book, written for the general public, written by psychologist and economist Kahneman, describes Dual Process Theory in terms of “fast and slow thinking,” referring to subconscious versus conscious thought. The author describes his understanding of judgement and decision making, using anecdotes and examples the reader can relate to.  The book walks the reader through a description of System 1 and System 2 thinking, describes the literature on heuristics, explores our tendency to have excessive confidence in what we know, applies economic principles to decision making and differentiates between the “experiencing self and the remembering self.”  A final chapter explores the implications of the distinctions drawn in the book.

Ilgen JS, Humbert AJ, Kuhn G, et al.  Assessing Diagnostic Reasoning: A Consensus Statement Summarizing Theory, Practice, and Future Needs.  Acad Emerg Med 2012; 19:1454-1461.

This proceedings paper from the 2012 Academic Emergency Medicine Consensus Conference on Education Research outlines existing theories of diagnostic reasoning as they apply to Emergency Medicine and describes strategies for assessing clinical reasoning.  The authors then propose gaps in the reasoning literature and consensus-based priorities for future research on clinical reasoning.

Norman GR, Monteiro SD, Sherbino J, et al.  The Causes of Errors in Clinical Reasoning: Cognitive Biases, Knowledge Deficits, and Dual Process Thinking.  Academic Medicine 2016.

This literature-based opinion piece by experienced clinical reasoning researchers reviews dual-process models, describes their application to cognitive decision making and describes the literature that explores the relationship of reasoning processes to cognitive errors.  Readers will find the overview helpful, the arguments relevant and the references of interest.

Dual Process Reasoning can be thought of as a general framework by which one may approach various problems of cognition, either in medical education or in other areas that require problem-solving and the complexities of human thought. While the theory can be incorporated the way that we analyze physician performance, how medical trainees integrate new information, and why errors occur, it should not be considered the only available lens.  While much insight can be gained by considering the dual process system, other theories may provide additional clarity in certain scenarios.  Educational theories allow us to approach issues from a variety of angles, and should be thought of as synergistic rather than exclusive of each other.  The true depths of human cognition will never be amenable to incorporation into a single, simplified theory.

Evans and Stanovich summarize multiple limitations of Dual Process Reasoning into themes. They review existing (and at times conflicting) definitions of Dual Process Reasoning and describe how researchers’ attempts to create clusters of attributes associated with each system have not provided consensus consistent across all accompanying theories. Although single theory explanations of Dual Process Reasoning have been proposed,^2,6  there may be limitations in dichotomizing a process that may represent more of a continuum. Further, the applications of Dual Process Reasoning are controversial, particularly when considering the complex relationship of processing strategy to cognitive errors.^5-8

Since cognitive ease may improve analytical thinking, Jason may benefit from a low stress and supportive environment.  Simulation would be ideal for this.  Jason was placed on a remediation program that involved dedicated simulation time. Jason meets weekly with the simulation division whose faculty members expose him to small tweaks of a clinical case to assess if he notices the subtle differences in each variation of the case.

Jason has his shifts paired with a well-seasoned, efficient senior resident where he is not relied upon to “move the meat” and is able to really digest each case.  He also meets with a faculty member to discuss a variety of clinical decision rules/illness scripts. Finally, recognizing that knowledge gaps may be responsible for cognitive errors regardless of the processing system applied, the residency leadership works with Jason to develop a plan for identifying and remediating knowledge gaps and ongoing self-directed learning to improve his foundational understanding of what differentiates, and how to approach, common EM presentations.