Exploring the impact of virtual patient design : medical students' small group learning around medical error

Abstract: Background: The demands on medical and healthcare practitioners are continuously changing, with new technologies, treatments and regulatory guidelines emerging each year. One such example is increased focus on the impact of medical error, which although difficult to measure is generally acknowledged to be responsible for significant numbers of patient harms each year. As a consequence, the provision of education and training must adapt to reflect this, providing learners with an updated range of skills that can meet the needs of their profession. An increase in the use of technology has been one way that educators have sought to achieve this, along with developing pedagogies and approaches such as problem-based learning which better reflect the challenges faced by medicine and healthcare professionals. Virtual patients are interactive simulations of clinical scenarios that have been shown to be well-suited to the development of clinical reasoning skills. They have been widely used in medicine and healthcare training, although they have yet to be fully adopted across the sector. Virtual patients can vary significantly in their design and their use within activities. To date there is a lack of knowledge about the ways different design features of virtual patients can be optimally applied to benefit student learning. Aim: The aim of this thesis is to increase knowledge of how virtual patient design impacts upon undergraduate medical student learning, particularly when targeting medical error education in small-group teaching. Methods: The four studies in this thesis explore design features of virtual patients and how they impact upon different aspects of the student learning experience. Study I investigated the impact of including video elements within a virtual patient used to support a problem-based learning tutorial. Mixed methods were used to capture aspects of both student and tutor experiences, and a thematic analysis was undertaken to identify themes in the unstructured responses. Studies II-IV each explored different outcomes around the use of decision-making elements in virtual patients designed to develop awareness of medical errors. A series of six virtual patients were delivered to undergraduate medical students undertaking paediatric placements across six institutions as part of small-group teaching. Students were allocated to one of two virtual patient designs: a branched design that allowed students to make decisions, or a linear design which followed a pre-determined structure without scope for student decision-making. In study II, following the completion of all virtual patients, students completed a survey instrument designed to measure aspects of their motivation and learning strategies. Comparisons between the branched and linear groups were made to establish the impact of the virtual patient design. Study III used log-linear analysis to explore learner performance in a single-best answer assessment and included an additional group of students who received traditional lecture-based teaching. In the assessment, questions were categorised in one of 3 groups; directly related to decision points in the virtual patient cases, in the same area of management but relating to different decisions and options, and in the same broad area of medicine but different areas of patient management. Study IV compared the self-efficacy and other related factors between students who received the branched and linear virtual patients, and asked participants to complete a survey instrument directly after each virtual patient case. A regression analysis was performed to explore how different factors impacted upon the self-efficacy of students. Results: In study I we identified eight distinct themes relating to the use of video elements. One theme related to levels of engagement, with some participants finding the use of video to be beneficial and others preferring text. Some participants identified that the use of video slowed the pace of the tutorial but was well suited to providing information about procedures. In study II we demonstrated that the use of linear or branched structures for virtual patients had no consistent impact upon the measures of learner motivation and learning strategies. Similarly, in study IV our findings showed that the use of a linear or branched design did not impact upon learner self-efficacy. However, in study III we showed that both virtual patient design and institution had an impact on student learning, and these were retained in our final model. A branching design for virtual patients improved student learning around the decision points in the virtual patient cases. Students performed equally well in the questions regarding the same management approaches but different decisions regardless of whether they received the linear or branched interventions, and scored higher than those that received traditional teaching. There was no difference between any of the groups in the questions that related to different areas of patient management. Conclusions: Overall, our findings suggest that the use of a branched virtual patient design is able to improve student learning around medical error when used in small-group teaching, and is not associated with any impact on learner motivation or self-efficacy. Our studies have not provided any evidence that this learning transfers to other areas of medicine. These findings have been broadly repeated at six institutions, demonstrating that despite evidence of the strong impact of institutional culture on our results, the findings can be generalised to multiple settings. We conclude that educators should seek to design virtual patients which allow learners to rehearse key patient management decisions, supported by video in areas where this can be most beneficial, such as demonstrating procedures. These virtual patients should be embedded in broader learning activities that encourage learners to identify deeper features within the learning, with a view to transferring that learning to other areas of patient management.