An approach to formalisation of clinical reasoning in medicine: a case study in psychiatry

Clinical reasoning involves eliciting clinical information, arriving at diagnoses, making treatment decisions, and assessing treatment progress. Errors in clinical reasoning can be critical, resulting in significant morbidity and mortality. Despite the need for a more reliable and transparent approach in order to minimise errors, clinical reasoning is largely a qualitative reasoning process, which is prone to errors. Even expert clinicians are known to make various cognitive errors, while their reasoning processes are often poorly understood and lacking in transparency due to the implicit nature. This research work was aimed at achieving a paradigm shift in clinical reasoning by introducing a quantitative reasoning framework, which is based on mathematical and computational models. Since the entire medical field is too broad to introduce such a paradigm shift directly, clinical psychiatry was chosen to introduce and evaluate this novel approach. As a result, two diagnostic reasoning algorithms known as the Drill-Locate-Drill (DLD) algorithm and the Select and Test (ST) algorithm have been introduced, and their accuracy was demonstrated using real patient data. Also, a model for making treatment decisions has been introduced and its utility was demonstrated using a complex clinical case. Similarly, a model for monitoring treatment progress and illness recovery has been introduced, and its application was demonstrated using a clinical case. Finally, with the aim of introducing this quantitative reasoning framework to clinicians and engaging them in using this novel approach, the above stated algorithms and models were published as educational resource in the Australasian Psychiatry journal. It is expected that the quantitative reasoning framework introduced in this thesis will improve the accuracy of clinical reasoning and result in better health care delivery. This thesis largely comprises a collection of publications achieved by the author during his research higher degree candidature. In summary, the publications comprise: six book chapters. ; eight journal papers (including one paper, which has been submitted and under review). ; eight conference papers. Every one of these papers was peer-reviewed prior to publication and complies with the Excellence in Research for Australia (ERA) Higher Education Research Collection (HERDC) requirements.