Rapid Emergency Skills for Prehospital Optimization, Notification, and Delivery (RESPOND-Guatemala): A comprehensive mHealth program for emergency prehospital skills and triage notification

ABSTRACT Over 6 million individuals die from traumatic injuries annually, with 40 million permanently injured. Trauma care innovations developed in resource-constrained settings often provide breakthrough solutions that enhance emergency medical systems globally, including in the United States. Guatemala has a high injury burden with no structured prehospital system. Care falls to "bomberos"—mostly volunteer firefighters without medical training or health system integration. Geographic challenges and traffic extend transport times while patients receive no care en route, arriving at overwhelmed hospitals unprepared for their management. This causes critical delays in definitive treatment for patients who could benefit from established interventions like tourniquets. Research shows that accurate prehospital triage with advance hospital notification significantly improves trauma outcomes. Our team previously collaborated with bomberos in Guatemala to create a mobile and physical-simulation platform for training on essential prehospital hemorrhage control techniques and successfully trained bomberos to apply tourniquets with high proficiency. To address critical prehospital needs worldwide, we propose Rapid Emergency Skills for Prehospital Optimization, Notification, and Delivery (RESPOND) mHealth, a comprehensive, gamified, AI-powered program that delivers three integrated solutions to bomberos: (1) context-adapted basic emergency prehospital skills training; (2) a reliable mechanism for accurate triage and prompt hospital notification; and (3) a bombero community-of-practice with AI-powered decision support and consensus guidelines to aid management decisions, triage classifications, and sustainable learning. We will focus on well established components of emergency prehospital care (e.g. basic airway management, spinal immobilization, hemorrhage control, accurate triage and transportation). The innovative AI-powered training, consensus-generation, and decision-support platform developed through this research will inform improvements to US emergency medical services, particularly benefiting rural American communities. We hypothesize that the RESPOND program will improve the proportion of patients who arrive at the hospital with appropriate prehospital interventions and with accurate prehospital notification for the sickest patients. In the R21 phase, we will Aim 1: Develop the RESPOND mHealth application software and context- adapted content; Aim 2: Assess RESPOND mHealth training efficacy, validity, and app usability in a sample of 35 bomberos; Aim 3: Establish considerations for implementation through stakeholder interviews at Roosevelt Hospital in Guatemala City. In the R33 phase, we will: Aim 1: Implement RESPOND with two bombero companies servicing Roosevelt Hospital; Aim 2: Measure RESPOND impact on trauma triage, patients, bomberos, and hospital stakeholders using a prospective cohort design and qualitative methods; and Aim 3: Assess sustainability and scalability of the RESPOND program. If successful, this work will enable accurate recognition of life threatening injuries, effective delivery of basic prehospital care, and efficient interface with the hospital in a challenging, resource-constrained environment while generating innovations that benefit prehospital systems worldwide.