To our knowledge, this is the first interprofessional study in which the evolution of multidimensional mental models of leadership is explored by incorporating metacognitive strategies in the development of this non-technical skill.
Leading clinical teams effectively requires the development of complex and multi-dimensional mental models. For mental models to result in enabling behaviors that are transferable to professional practice it is first necessary to understand scientific concepts. Then it is also necessary to establish relationships between those concepts and the ontological, cognitive-linguistic, and motivational dimensions related to clinical team leadership. It also requires leadership mental models to be related to models of other non-technical skills. This enabling correlation is a progressive longitudinal process that requires deliberate practice combined with guided reflection and self-awareness. This took place after 8 simulations in the fourth week of interprofessional training.
In this study, undergraduate medicine, and nursing students in their last year of training had a low baseline performance in leadership in simulated emergency cases. This is consistent with a prior study by Cha et al which looked at medical students management of emergencies that focused only on medical students (39). Barnes et al, in a recent systematic review, concluded that medical students' self-perception is low regarding their leadership preparation in their undergraduate education (12).
In the qualitative aspect, the participants initially showed a relationship in their mental models with the epistemological dimension in the great majority, and of these, the majority was in a basic development, the focus was related to the definition of the concept. Ex.: “Leadership is the ability to influence the will of others”.
The improvement in the conceptual domain was rapid, after the first week the vast majority described scientific concepts, nonetheless, this was not related to improvement in performance. We can interpret this from the taxonomy of knowledge, which can be classified as either declarative, procedural, and conditional. The declarative aspect is representational, it refers in the case of skills to the aspect that can be narrated, described, or enumerated, on the other hand, the procedural aspect is neither true nor false, it is more related to performance and demonstration (40), thus, having declarative capacity for concepts does not equal immediate translation into actions.
In the second week, performance showed a slight improvement, meanwhile, the mental models presented an important change, mainly in aspects not related to knowledge, although most of them were at the level of scientific concepts (59.6%). The ontological dimension was related to most leadership models, we attribute this finding to the trust generated through the deliberate construction of a safe learning and reflection environment with the research group and among the team (41, 42).
The performance in the third week improved with respect to the two previous weeks, it can be evaluated as acceptable, in the semantic networks there is an emergence of new codes associated with other non-technical skills (communication, situational awareness), despite this, the mental models were still disabling.
The participants in their reflections considered that the difficulty of leading in a crisis is high, the noise generated by the way of communicating and the pressure of the context do not allow them to reason calmly. This is consistent with studies in which the cognitive load of undergraduate and postgraduate teams in crisis situations is very high and does not allow professionals to be efficient when performing a difficult task (43, 44). In specific reference to leadership, the management of cognitive load in the emergency room can be managed from shared leadership between doctors and nurses, which could be associated with better performance (45).
Mental models changed from disabling to enabling by week four of the study and this was strongly associated with improved performance, where leadership execution appeared to be automatic, given the speed and accuracy of leadership in task execution, which is consistent with Fitts and Posner's theory of three stages in the development of skills (cognitive phase, associative phase, and autonomous phase) (46), that has been validated in subsequent studies in other fields of knowledge (47).
We explain the notable improvement observed in the fourth week based on two main reasons: firstly, the reflective level of the participants was at a high level, demonstrating metacognitive activity, a concept that was present in their oral and written speech, which denotes consistency; it is possible that deliberately working on metacognitive strategies has allowed its development (48). Second, the mental models of leadership were interrelated with models of other skills (communication, situational awareness, teamwork) and with several dimensions of the model, which possibly allowed the participants to have more resources at the time of carrying out the task and generating shared mental models.
Based on our findings, we recommend that transition training programs using immersive patient simulation in combination with metacognitive strategies should be implemented to optimize the development of non-technical skills through the evolution of mental models in any postgraduate training program where the development of non-technical skills is particularly relevant, for example in specialties such as emergency medicine, intensive care, surgery, and anesthesiology. Effective leadership is a complex skill that is not acquired through one isolated training activity. Instead, it requires combining multiple opportunities for deliberate practice with structured reflection over time. Additionally, we recommend that interprofessional immersive simulation be conducted in the continuing education of health system professionals.
Limitations
We believe that this study has some limitations that make it difficult for its generalization, such as the lack of a control group, the follow-up of the mental models was done only during the study, lack of knowledge regarding decline in performance as a function of prolonged period; finally, this research was conducted in Spanish, in Colombia, and results could vary in other languages and cultural contexts, which was not investigated in this work. These limitations could be addressed in future studies.