Dual Process Interventions for Diagnostic Reasoning Among Medical Students and Junior Doctors
Citation:
LAMBE, KATHRYN ANN, Dual Process Interventions for Diagnostic Reasoning Among Medical Students and Junior Doctors, Trinity College Dublin.School of Psychology.PSYCHOLOGY, 2018Download Item:
Lambe_thesis_corrected_clean_margins_pagenos.pdf (PDF) 13.59Mb
Abstract:
The aim of this thesis was to examine cognitive interventions to enhance diagnostic accuracy based on the dual process model of analytical and non-analytical reasoning. The work sought to use the model as an organising principle to examine existing interventions in the field, to consider the extent to which interventions are grounded in theory and reflect the processes described in the model, and to use the perspective of psychological science to illuminate the psychological diagnostic processes upon which these interventions act. The thesis comprises a systematic review of interventions to enhance analytical and non-analytical diagnostic reasoning and two empirical studies of a reflective intervention, which examine the mandated and voluntary use of the intervention by medical students and junior doctors using both quantitative and qualitative methodologies.
Study 1 was a systematic review, aiming to appraise studies that have empirically investigated interventions to enhance analytical and non-analytical reasoning among medical students and doctors and to critically consider the use of dual process theory to provide clear conceptual rationale for the interventions. Twenty-eight studies were included under six categories: educational interventions, checklists, cognitive forcing strategies, guided reflection, instructions at test and other interventions. While most studies offer some theoretical background and rationale for the interventions they investigate, the discussion of this is generally quite shallow, and interventions overwhelmingly target analytical monitoring processes. Guided reflection interventions emerged as the most consistently successful across five studies. Therefore, the thesis evolved in the next two studies into an examination of guided reflection interventions specifically.
Empirical studies of guided reflection interventions have largely comprised quantitative experiments, with little consideration of how doctors may voluntarily use the guided reflection process and the specific ways in which it changes or enhances their diagnostic process. Additionally, the study of diagnostic reasoning and error within psychiatry has been largely neglected in the broader literature. Study 2 therefore utilised a think-aloud protocol to observe how junior doctors (n=14) made voluntary use of a guided reflection tool in diagnosing three complex psychiatric clinical scenarios. A majority of participants chose to use the tool and responded positively to it. Thematic analysis highlighted the impact of reflection on the process of diagnosis and on diagnostic confidence, and participants broadly indicated that the tool had potential for use in real-world diagnosis, particularly for ambiguous or complex cases.
The potential for unintended negative consequences (e.g., excessive time requirements, over-utilisation of resources) of such interventions has gone largely unexplored in the literature; concerns around compounding of biases and overcorrection are important to consider. Study 3 therefore utilised a between-groups experiment with 186 senior medical students to compare the impact of a short and long version of a guided reflection task. The shorter version of the process elicited accurate diagnostic hypotheses at a similar rate as the longer version, but neither led to significant gains in the accuracy of the final diagnosis over diagnosis based on first impressions alone. No evidence for inflated diagnostic confidence through an induced availability bias was observed, and final diagnoses seldom differed from initial hypotheses. The study found no evidence to support the use of the guided reflection method as a diagnostic aid for novice diagnosticians, and underscored the importance of the first diagnostic hypothesis and the need to develop non-analytical reasoning skills, which play a vital role in diagnostic accuracy.
The findings offer support for the validity of the dual process model?s application to diagnostic reasoning, and underscore the value of theoretical perspectives in the design and interpretation of intervention studies. The guided reflection process shows promise as a tool for education in clinical reasoning and as an aid for diagnosis in the clinical workplace, though questions remain as to the most appropriate audience and contexts for its implementation in these settings. Further research is now needed, with recruitment of more diverse populations, exploration of emerging and qualitative methods, and continued examination of guided reflection interventions specifically.
Sponsor
Grant Number
Irish Research Council (IRC)
Author's Homepage:
http://people.tcd.ie/lambekaDescription:
APPROVED
Author: LAMBE, KATHRYN ANN
Advisor:
Hevey, DavidPublisher:
Trinity College Dublin. School of Psychology. Discipline of PsychologyType of material:
ThesisAvailability:
Full text availableLicences: