| dc.description.abstract | Triage is a process performed in the emergency department of hospitals aimed at sorting the patients according to their needs of care. On the other hand, an early warning system is a protocol used by hospitals to detect the deterioration of patient’s vital signs. When well performed, these processes can potentially increase the chances of life of patients with a high degree of complications, guiding their treatment and the correct diagnosis. Mobility is a musthave requirement for healthcare professionals to perform their daily activities and this is in the same way with the rise of mobile and ubiquitous computing. Mobile and wearable devices are increasingly present in our daily lives. This study aims to develop a computational model, called UbiTriagem 2, to support the triage process, supporting an early warning system, using the concepts of mobile and ubiquitous computing and Internet of things related to healthcare. The main scientific contribution of this study was to propose the interoperability between different protocols of triage and deterioration of vital signs (severity) used today in hospitals using an ontology. In addition, another concern was to aggregate, also in the ontology, information collected from different IoT sensors (as data source) to infer the triage and severity of patients. The model was evaluated through scenarios, which showed that the model is apt to be used in an emergency department. In relation to the triage, it was possible to conclude that the model was able to correctly determine the patient’s classification in 93.33% of the evaluated situations and, with minor adjustments, reached 100% of the cases. The early warning system was assertive in 86.71% of the cases, on the other hand we can conclude that it closely resembles the qualitative evaluation carried out by an emergency medical regulator. In addition, 63.61% of all cases from SAMU in the emergency department could benefit from this model. Finally, with the evaluation made through the focus group methodology, we can highlight as positive points of the developed model: the use of already validated protocols; the follow-up of the service queues; the use of mobile devices; the decrease in errors in the use of protocols; the use of wearable devices to monitor patients; a non-intrusive model; the aid in recording attendance data; greater support for nurses’ decisions; the reduction of mortality rates and major complications; and the decrease in the cost of care per patient. | en |
