Emergency care-sensitive conditions (ECSCs) require rapid identification and treatment and are responsible for over half of all deaths worldwide. Prehospital emergency care (PEC) can provide rapid treatment and access to definitive care for many ECSCs and can reduce mortality in several different settings. The objective of this study is to propose a method for using artificial intelligence (AI) and machine learning (ML) to transcribe audio, extract, and classify unstructured emergency call data in the Serviço de Atendimento Móvel de Urgência (SAMU) system in southern Brazil. The study used all “1-9-2” calls received in 2019 by the SAMU Novo Norte Emergency Regulation Center (ERC) call center in Maringá, in the Brazilian state of Paraná. The calls were processed through a pipeline using machine learning algorithms, including Automatic Speech Recognition (ASR) models for transcription of audio calls in Portuguese, and a Natural Language Understanding (NLU) classification model. The pipeline was trained and validated using a dataset of labeled calls, which were manually classified by medical students using LabelStudio. The results showed that the AI model was able to accurately transcribe the audio with a Word Error Rate of 42.12% using Wav2Vec 2.0 for ASR transcription of audio calls in Portuguese. Additionally, the NLU classification model had an accuracy of 73.9% in classifying the calls into different categories in a validation subset. The study found that using AI to categorize emergency calls in low- and middle-income countries is largely unexplored, and the applicability of conventional open-source ML models trained on English language datasets is unclear for non-English speaking countries. The study concludes that AI can be used to transcribe audio and extract and classify unstructured emergency call data in an emergency system in southern Brazil as an initial step towards developing a decision-making support tool.

Introduction: Delays in prehospital care attributable to the call-taking process can often be traced back to miscommunication, including uncertainty around the call location. Geolocation applications have the potential to streamline the call-taking process by accurately identifying the caller’s location. Objective: To develop and validate an application to geolocate emergency calls and compare the response time of calls made via the application with those of conventional calls made to the Brazilian Medical Emergency System (Serviço de Atendimento Médico de Urgência—SAMU). Methods: This study was conducted in two stages. First, a geolocating application for SAMU emergency calls (CHAMU192) was developed using a mixed methods approach based on design thinking and subsequently validated using the System Usability Scale (SUS). In the second stage, sending time of the geolocation information of the app was compared with the time taken to process information through conventional calls. For this, a hypothetical case control study was conducted with SAMU in the Maringá, Paraná, Brazil. A control group of 350 audio recordings of emergency calls from 2019 was compared to a set of test calls made through the CHAMU192 app. The CHAMU192 group consisted of 201 test calls in Maringá. In test calls, the location was obtained by GPS and sent to the SAMU communication system. Comparative analysis between groups was performed using the Mann-Whitney test. Results: CHAMU192 had a SUS score of 90, corresponding to a “best imaginable” usability rating. The control group had a median response time of 35.67 seconds (26.00–48.12). The response time of the CHAMU192 group was 0.20 (0.15–0.24). Conclusion: The use of the CHAMU192 app by emergency medical services could significantly reduce response time. The results demonstrate the potential of app improving the quality and patient outcomes related to the prehospital emergency care services.

Out-of-hospital cardiac arrests (OHCAs) have high mortality rates, worsened by limited access to automated external defibrillators (AEDs). This study analyzed OHCA response times, identified areas with prolonged ambulance travel times, and proposed optimal AED locations in a medium-sized city in southern Brazil. Data from 278 non-traumatic OHCA cases (2019–2022) in patients over 18 years old, with ambulance response times under 20 min, were included. Spatial survival analysis assessed the probability of exceeding the recommended 5-min (300 s) ambulance response time. The maximal covering location problem identified 100 strategic AED sites within a 150-s reach for bystanders. AED and ambulance travel times were compared using the Wilcoxon test (p < 0.01). Defibrillation occurred in 89 cases (31.01%), and bystander CPR was performed in 149 cases (51.92%). Despite these efforts, 77% of patients died. The median ambulance response time was 11.63 min, exceeding 5 min in most cases, particularly at peak times like 11 a.m. AED placement in selected locations could cover 76% of OHCA occurrences, with a mean AED travel time of 320 s compared to 709 s for ambulances. Strategic AED placement could enhance early defibrillation and improve survival outcomes.

Background: Mortality resulting from coronary artery disease (CAD) among women is a complex issue influenced by many factors that encompass not only biological distinctions but also sociocultural, economic, and healthcare-related components. Understanding these factors is crucial to enhance healthcare provisions. Therefore, this study seeks to identify the social and clinical variables related to the risk of mortality caused by CAD in women aged 50 to 79 years old in Paraná state, Brazil, between 2010 and 2019. Methods: This is an ecological study based on secondary data sourced from E-Gestor, IPARDES, and DATASUS. We developed a model that integrates both raw and standardized coronary artery disease (CAD) mortality rates, along with sociodemographic and healthcare service variables. We employed Bayesian spatiotemporal analysis with Markov Chain Monte Carlo simulations to assess the relative risk of CAD mortality, focusing specifically on women across the state of Paraná. Results: A total of 14,603 deaths from CAD occurred between 2010 and 2019. Overall, temporal analysis indicates that the risk of CAD mortality decreased by around 22.6% between 2010 (RR of 1.06) and 2019 (RR of 0.82). This decline was most prominent after 2014. The exercise stress testing rate, accessibility of cardiology centers, and IPARDES municipal performance index contributed to the reduction of CAD mortality by approximately 4%, 8%, and 34%, respectively. However, locally, regions in the Central-West, Central-South, Central-East, and Southern regions of the Central-North parts of the state exhibited risks higher-than-expected. Conclusion: In the last decade, CAD-related deaths among women in Paraná state decreased. This was influenced by more exercise stress testing, better access to cardiology centers, improved municipal performance index. Yet, elevated risks of deaths persist in certain regions due to medical disparities and varying municipal development. Therefore, prioritizing strategies to enhance women’s access to cardiovascular healthcare in less developed regions is crucial.

Purpose: Developing technological solutions to reduce care delay and mortality in out-of-hospital cardiac arrest (OHCA) cases is a challenge both in Brazil and worldwide. The aim of this study was to develop and validate a geolocation application for locating automated external defibrillators (AEDs) in the urban area of a medium-sized city in southern Brazil. Methods: A mobile application was developed to locate AEDs in public places, indicating the best route to the nearest AED with an intuitive design for emergency use. The app provides information on cardiopulmonary resuscitation (CPR), initial care, and AED use. The research steps included a literature review on the double diamond design process, prototyping, expert validation, and simulation tests. Results: The Local DEA app, developed for Android and iOS, offers features such as route calculation to the nearest AED and emergency services. In an evaluation conducted by 30 experts, the app was well accepted for its ease of use and efficiency (Cronbach’s alpha = 0.92). Simulation tests demonstrated that the strategic installation of accessible and well-located AEDs improves coverage and significantly reduces travel time to emergency care as compared with ambulance travel times (p < 0.001). Conclusion: The Local DEA app is a promising tool for reducing mortality in OHCA, enabling AED geolocation, use instructions, and CPR guidelines. When combined with the strategic placement of public-access AEDs, the application can increase the likelihood of survival in OHCA cases.

The aim of this study was to conduct a detailed geospatial analysis of mobile phone signal coverage in the northwest macro-region of Paraná State, Brazil, seeking to identify areas where limitations in coverage may be related to lengthy travel times of the helicopter emergency medical service (HEMS) for the assistance of victims of road traffic injuries (RTIs). An observational study was conducted to examine mobile phone signal coverage and HEMS travel times from 2017 to 2021. HEMS travel times were categorized into four groups: T1 (0-15 min), T2 (16-30 min), T3 (31-45 min), and T4 (over 45 min). Empirical Bayesian Kriging was used to map areas with low mobile signal coverage. The Kruskal-Wallis test and Dwass-Steel-Critchlow-Fligner comparative analyses were performed to explore how mobile signal coverage relates to HEMS travel times to RTI locations. There were 470 occurrences of RTIs attended by HEMS, of which 108 (23%) resulted in on-site fatalities. Among these deaths, 47 (26.85%) occurred in areas with low mobile phone signal coverage ("shadow areas"). Low mobile phone signal coverage identified at 175 (37.24%) RTIs locations, was unevenly distributed across the macro-region. The lowest medians of mobile signal quality were predominantly found in areas with HEMS travel times exceeding 30 min, corresponding to signal strength values of -98.44 (T3) and -100.75 (T4) dBm. This scenario represents a challenge for effective communication to activate HEMS. In the multiple comparison analysis among travel time groups, significant differences were observed between T1 and T2 (p < 0.001), T1 and T3 (p < 0.001), T1 and T4 (p < 0.001), and T2 and T3 (p < 0.001), indicating a potential association between lower mobile phone signal coverage and longer HEMS travel times. It can be concluded that poor mobile phone signals in remote areas can hinder HEMS activation, potentially delaying the start of treatment for RTIs. Identification of the shadow areas can help communication and health managers in designing and implementing the necessary changes to improve mobile phone signal coverage and consequently reduce delays in the initial response to RTIs.