EDXL-LD and Architectural Tactics towards Information Sharing and Interoperability in Emergency Context

Abstract

s of Scientific Papers-WADEM Congress on Disaster and Emergency Medicine 2017 EDXL-LD and Architectural Tactics towards Information Sharing and Interoperability in Emergency Context Julia Kantorovitch, Jarmo Kalaoja, Ilkka Niskanen VTT Technical Research Center of Finland, Espoo/Finland Study/Objective: The objective of this study is to propose a new architectural approach supported by information models, to manage knowledge in the dynamic emergency context towards interoperable knowledge sharing and reuse. Background: The knowledge systems for emergency management are based on evolvable information provided by various actors, by diverse collections of sensors and information supplied by human volunteers. In order to achieve a common operational picture situation awareness, various knowledge, vocabulary and information models need to be aligned. This requires extendable time, application context architecture and models representing detailed evolvable knowledge about the types of adverse events, their potential impact and the means and resources that are best suited for response. The existing semantic research has a potential to address the identified needs, however the reported ontologies are rarely publically available, and they are also disconnected from widely used standard data models, data-exchange formats, and protocols related to emergency management. Methods: The literature review and the inputs provided by domain experts in the CONCORDE consortium and WHO, have facilitated the addressing of shortcomings and challenges identified above. Results: The Emergency Data Exchange Language (EDXL) based domain specific standards are taken as a base to create domain specific vocabularies. Vocabularies are published as a Linked Data (LD) and can be downloaded from GitHub software repository https://github.com/OntoRep/EDXL. The Model-View-Presentation (MVP) based architectural tactics as a software engineering pattern (see below) are exploited to achieve a desired extensibility and dynamicity of the system at its deployment stage. Conclusion: By keeping applications’ business logic separate from data and semantics, the underlying knowledge models can evolve without necessarily requiring changes to the interfaces and applications built on top of the models. Prehosp Disaster Med 2017;32(Suppl. 1):s228 doi:10.1017/S1049023X17005878 The J-SPEED: A Medical Relief Activities Reporting System for Emergency Medical Teams in Japan Tatsuhikiko Kubo, Hisayoshi Kondo, Yuichi Koido 1. Department Of Public Health, University of Occupational and Environmental Health, Kitakyusyu/Japan 2. National Disaster Medical Center, Tokyo/Japan Study/Objective: To introduce the J-SPEED; medical relief activities reporting system for Emergency Medical Teams (EMTs) of Japan. Background: During a disaster, information gathering and analysis are key elements for better coordination and timely response. Previous cases revealed that EMTs sometimes became the only capacity which could report medical, or more broadly health situations to a coordination body, and standardization of the reporting process from EMTs to the EMT Cortication Cell (EMTCC) will allow for better coordination, and for strengthening of the disease early warning system, since EMTs will act as additional sentinel reporting sites. One good existing model for this issue is the Surveillance in Post Extreme Emergencies and Disasters (SPEED) system employed in the Philippines. The SPEED was developed by Philippine’s Department of Health and the WHO in 2010. Based on the lessons learned from relief mission of the Japan Disaster Relief Medical Team against the super typhoon Yolanda in 2013, a Japanese version of the SPEED, so called J-SPEED has been developed and published in 2015. Methods: Field study. Results: The J-SPEED was first activated at the Kumamoto earthquake which occurred on April 14, 2016. During the 48 days of response, EMTs from various affiliation sent 1,828 daily reports to the EMTCC, which represented medical demand of 8,089 patients. Standardized information processing and quantitative information made communications among stakeholders efficient, and supported evidence, consensus based decision making by the local authority. Conclusion: Employment of the J-SPEED drastically changed the EMT coordination in Japan. Countries which don’t have a relevant system can easily set up a national reporting system utilizing the SPEED framework. Prehosp Disaster Med 2017;32(Suppl. 1):s228 doi:10.1017/S1049023X1700588X How do we Measure Severity? An Assessment of Five Indexes used in Sudden Onset Disasters and Complex Emergencies to Measure Severity and Risk Anneli Eriksson, Martin Gerdin, Thorkild Tylleskär, Johan Von Schreeb 1. Public Health Science, Karolinska Institutet, Stockholm/Sweden 2. Centre For International Health, Bergen University, Bergen/Norway 3. Department Of Public Health Sciences, Karolinska Institutet, Stockholm/Sweden SITUATIONAL AWARENESS SYSTEMS Prehospital and Disaster Medicine Vol. 32, Supplement 1 Study/Objective: The aim was to, 1) study the relation between disaster outcomes after earthquakes, expressed as number of dead and injured, and the performance of five preidentified severity, and risk-scoring indexes, 2) to inform a model that in an initial phase of a disaster can be used to predict severity and levels of need, and thereby guide toward the appropriate levels of response. Background: A disaster is as an event that overwhelms local capacity, necessitating national or international assistance. Disasters can be categorized, based on the type of hazard causing them. An earthquake is a hazard that can lead to a disaster. The disaster-severity depends on the magnitude of the hazard, underlying vulnerability, the level of exposure, coping capacity and the disaster response. While assistance should be based on needs, determined by the severity of a situation, there is no recognized way to compare severity between disaster contexts. Several initiatives have been developed to provide information on global severity and risks in disaster situations. In this study we compare five indexes and their ability to define severity: GDACs, GEO, KI’s 7-need, INFORM and ECHO’s Crisis index. Methods: We did a mapping of the existing indexes and indicators used. Index-scores were standardized and then compared with the number of dead and injured as an absolute outcome, in earthquakes with magnitude equal to or higher than 6,5 that occurred in populated areas, between year 2001 and November 2016. Results: The five indexes evaluated were all indicating the severity after the examined earthquakes. There was not one single index that gave an absolute correlation. Indexes built on higher numbers of indicators had several indicators that gave identical information. Conclusion: It is possible to predict the severity of a disaster through proxy indicators. The number of indicators used is not automatically increasing the preciseness or validity of the outcome. Prehosp Disaster Med 2017;32(Suppl. 1):s228–s229 doi:10.1017/S1049023X17005891 Enhanced Situational Awareness through a Decision Support Service for Optimal Allocation of Resources and Response Capacity Irene Christodoulou, George M. Milis, Panayiotis Kolios, Christos Panayiotou, Marios Polycarpou, Ilkka Niskanen 1. Kios Research Center For Intelligent Systems And Networks, University of Cyprus, Nicosia/Cyprus 2. VTT Technical Research Center of Finland, Espoo/Finland Study/Objective: We designed and developed e-services, aiming to support the decision makers during various contexts of medical emergency response, offering them machine-aided enhanced situational awareness. Background: Currently, decisions are being made by human experts with hands-on experience in emergency fields. However, in most cases, experts do not have the required computational capacity to process the relevant heterogeneous information and perform informed decisions. Evidently, time is a very critical parameter in emergency situations, especially in large-scale incidents with large number of casualties. Methods: Taking this into account the services we develop, are based on the mathematical modeling of optimization problems for timely resources’ allocation, addressing different phases of the response. The formulated problems address: i) the optimal allocation of Emergency Medical Services (EMS) units (in terms of demand satisfaction and time), to active emergency incident fields, ii) the optimal allocation (in terms of exploiting their capacities and response time) of EMS staff to tasks on the incident field such, as triage and retrieval running, transferring of patients to medical treatment area, offering medical treatment, and iii) the optimal allocation (in terms of profile matching, demand satisfaction and time) of patients to EMS vehicles and subsequently to first receivers (hospitals). The services are supported by semantic modeling of EMS vehicles, hospital, staff and patients profiles, as well as by machine learning tools that estimate demand for resources given historical emergency incident data. The services offer clear interfaces, so as to be interoperable with existing emergency management systems, as long as access to the necessary information is given. Results: Our solution achieves the recommendation on allocation of resources, based on real-time collected information from the emergency field. Conclusion: Further work will focus on modeling different cost functions in the optimization, so as to customize the recommendations based on incident and/or decisionmaker needs. Prehosp Disaster Med 2017;32(Suppl. 1):s229 doi:10.1017/S1049023X17005908 Comparison of UAV Technology vs No UAV Technology in Identification of Hazards at a MCI Scenario in Primary Care Paramedic Students Trevor N. Jain, Aaron Sibley, Henrik Stryhn, Ives Hubloue 1. Biology (paramedicine), University of Prince Edward Island/ Holland College, Charlottetown/PE/Canada 2. H