Can anxiety in undergraduate students in a high-fidelity clinical simulation be predicted? A randomized, sham-controlled, blinded trial

High-fidelity clinical simulation has implied a revolution in health science training. Despite its benefits, some drawbacks could hinder the learning process, especially the anxiety produced during such scenarios. The aim of the present work is to develop a predictive model capable of determining which students will present high levels of anxiety. We performed a randomized, sham-controlled, blinded trial in which students were randomly assigned to four scenarios and played one of two possible roles. Before and after the simulation we assessed the anxiety level along with physiological and analytical parameters. The main analyzed outcome was an increase of ≥25% in anxiety compared with baseline. The type of scenario or the role played had no effect on anxiety. The predictive model presented an Area Under the Receiver Operating Characteristics of 0.798 (95% CI: 0.69–0.90; p