SCROLL TO BOTTOM FOR THE INTERACTIVE SIMULATOR
My current approach, which replaces other work now removed from this site, is motivated by a substantial shift in the clinical-reasoning literature. The familiar interpretation of dual-process theory, in which rapid intuitive reasoning is presumed to be error-prone and slower analytical reasoning is presumed to correct it, is no longer an adequate foundation for a clinical-reasoning.
Norman and colleagues argue that both rapid and analytical reasoning can be correct or incorrect, and that diagnostic performance depends critically on whether relevant knowledge is available and retrievable rather than on the mere choice of a processing mode. More recent work similarly calls for models that are ecologically valid, knowledge-centred, context-sensitive and able to accommodate metacognition, tacit knowledge and the social and material setting of practice.
The clinical knowledge framework [CKF] I am now developing makes five commitments:
- Expert reasoning is primarily knowledge-dependent rather than algorithm-dependent.
- Rapid recognition is not intrinsically inferior to explicit analysis; it can be a hallmark of expertise when supported by well-organised experience and knowledge.
- Similarity is an important feature of knowledge retrieval and recognition, but it is not the driving logic of the framework.
- Metacognition is not generic “thinking about thinking”. It is the knowledge-dependent monitoring and regulation of whether the current clinical representation and intended action remain adequate.
- Clinical reasoning is situated. It occurs through interaction with patients, colleagues, artefacts, technologies, organisational routines and constraints, not only inside the clinician’s head.
The framework is intended to support diagnostic, managerial and anticipatory reasoning. This is essential because many of the clinical prediction models with which the CKF must interact do not diagnose disease. They estimate a future state or change in vulnerability, and the relevant clinical question is whether that information should alter the current representation or management plan.
This version currently incorporates the following cognitive plugins:
- Stopping Rule
- Counterfactual Reasoning
- Boundary Case Clarification Predictor
- Representation Trajectory
- Burnout/Brownout context
You can review the latest version of the CKF below the references. As a simulator and also for clinical use, the model provides a running commentary down the side of what cognitive plugins are being deployed.
Key references:
- Ng IKS, Long V, Lee ARYB, Lahiri M, Ko CJ, Morgan DJ, Lim TK. Re-imagining clinical reasoning: an updated cognitive framework for real-world medical practice. Journal of Evaluation in Clinical Practice. 2026;32(5):e70560. doi:10.1111/jep.70560.
- Norman G, Pelaccia T, Wyer P, Sherbino J. Dual process models of clinical reasoning: the central role of knowledge in diagnostic expertise. Journal of Evaluation in Clinical Practice. 2024;30:788–796. doi:10.1111/jep.13998.
- Parsons AS, Wijesekera TP, Olson APJ, Torre D, Durning SJ, Daniel M. Beyond thinking fast and slow: implications of a transtheoretical model of clinical reasoning and error on teaching, assessment, and research. Medical Teacher. 2025;47(4):665–676. doi:10.1080/0142159X.2024.2359963.
- Pelaccia T, Wyer P, Sherbino J, Norman G. Self, others and context: reframing clinical reasoning as a socio-cognitive practice. Journal of Evaluation in Clinical Practice. 2026;32(6):e70619. doi:10.1111/jep.70619.
Clinical Knowledge Framework.