Automated Debriefing in Simulation-based Training of Paramedics: Comparing Self-guided to Agent Facilitated Debriefing

点击翻译按钮获取中文摘要

The COVID-19 pandemic has highlighted the need for paramedics and other emergency medical service providers, such as emergency medical technicians (EMTs). These first responders can keep a person who is struggling to breathe alive until they arrive at a hospital and receive sustained care, including support from a respiratory ventilator if necessary. This project focuses on paramedics, who have the highest level of training in pre-hospital care. In addition to utilizing airway devices that support breathing, paramedics can revive a heart attack victim, deliver a baby who cannot wait for his/her mother to get to the hospital, administer medications, and perform other life-sustaining procedures. Unfortunately, there is a critical shortage of paramedics in many parts of the US, which is partly due to student failure from paramedic training programs, or inability to pass state and national certification exams. This project will develop a handheld tablet device that will allow small groups of paramedic students to get extra practice in performing pre-hospital care procedures on patient manikins. These supplemental practice sessions will be entirely student-run. Students will download scripted emergency scenarios from their course web site. One student will play the role of lead paramedic, another student will play the role of an EMT who assists the paramedic, while a third student will use the tablet to check off actions (e.g., Checked blood pressure) and to report the results of these actions (e.g., Blood pressure is 80/50). All actions will be recorded in a log that will produce an automated debriefing; that is, feedback on student performance during the scenario. Increased opportunities for students to practice skills during training could raise pass rates from paramedic programs and on certification exams. Better-prepared paramedics will also increase patient safety. Project findings might generalize to training of other healthcare providers such as doctors, nurses, and physician assistants.This project will focus on the debriefing that typically follows a simulation-based training (SBT) exercise, when students reflect on their performance with feedback from a trained facilitator. Although meta-reviews of SBT indicate that debriefing may be its most critical component, there is a dearth of research to guide debriefing practice. This project will help to fill this void. The research team will implement and compare the effectiveness of three approaches to automated debriefing. One version will replicate the current state-of-the-art in automated debriefing, whereby students scroll through an event log with embedded textual feedback on their actions during the scenario. This approach to debriefing does not follow a standardized protocol (script) for debriefing. The second version follows a debriefing protocol whose acronym is DEBRIEF (for Define the debriefing rules; Explain the learning objectives; specify the performance Benchmarks; Review what was supposed to happen; Identify what actually happened; Examine why; and Formalize the take home points). This version of automated debriefing simulates student self-guided debriefing. The system will prompt the student to go through the DEBRIEF protocol steps but will not provide feedback. The third version will simulate human facilitated debriefing. Two automated agents, a facilitator and a peer, will support the student who played the role of lead paramedic during a practice scenario in carrying out a productive debriefing, according to DEBRIEF. The main technological innovation will be development of methods to manage the interaction between these simulated agents (facilitator and peer) and the human paramedic student. The project consists of three main phases. First, to inform implementation of the agent facilitated debriefing system, the research team will observe and video record actual SBT scenarios and debriefings that take place during paramedic training at the University of Pittsburgh, Center for Emergency Medicine. Second, the researchers will iteratively develop, field test, and refine each version of the debriefing system. Third, the researchers will conduct a randomized controlled trial to compare the potential of these approaches to automated debriefing to improve psychomotor skills (e.g., cardiopulmonary resuscitation) and decision-making skills. Study findings might generalize to simulation-based training of other healthcare professionals and to any field in which debriefing plays a major role in training.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.