Patricia Takako Endo

Background: The state of Amazonas, located in the Brazilian Amazon, was disproportionately affected during the first waves of the COVID-19 pandemic. Methods: We conducted a population-based cohort study including all confirmed COVID-19 cases (2020–2) in the state of Amazonas, located in northern Brazil, linking data from three surveillance systems with cases of influenza-like illness (ILI) and severe acute respiratory syndrome (SARS). Outcomes included death, intensive care unit (ICU) admission, and a composite severity outcome. Results: Among 420 819 cases, 378 142 (89.9%) had ILI and 42 677 (10.1%) had SARS. A substantial burden of severe outcomes was observed among SARS cases, with ∼25% (n = 8910/35 909) of patients admitted to the ICU, 21.1% (n = 7229/34 323) requiring invasive mechanical ventilation, and 36.7% (n = 14 971/40 837) dying. Male sex [odds ratio (OR) = 1.23; 95% confidence interval (CI): 1.15–1.31; P < .001], age of ≥70 years (OR = 13.73; 95% CI: 11.97–15.76; P < .001), Indigenous ethnicity (OR = 2.32; 95% CI: 1.85–2.91; P < .001), dyspnea (OR = 1.99; 95% CI: 1.82–2.19; P < .001), loss of smell (OR = 1.54; 95% CI: 1.31–1.81; P < .001), and comorbidities, including chronic kidney disease (OR = 1.94; 95% CI: 1.58–2.39; P < .001), chronic respiratory disease (OR = 1.38; 95% CI: 1.08–1.77; P < .05), diabetes (OR = 1.12; 95% CI: 1.03–1.22; P < .05), obesity (OR = 1.31; 95% CI: 1.13–1.51; P < .001), and immunosuppression (OR = 1.90; 95% CI: 1.48–2.44; P < .001), were independently associated with a higher risk of death. Lethality increased significantly during the Gamma wave (OR = 3.76; 95% CI: 3.34–4.24; P < .001). Conclusion: The severity of COVID-19 in Amazonas was shaped by biological, structural, and viral factors, highlighting the need for equitable strengthening of the health system.

The rapid evolution of Generative Artificial Intelligence (GenAI) presents significant opportunities to transform healthcare, particularly in generating personalized treatment recommendations. This systematic literature review explores the current state of GenAI language models applications in various medical domains, assessing their effectiveness, applicability, and limitations. The review addresses nine specific research questions to understand the potential and challenges of integrating GenAI into clinical practice. We use the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. From a pool of 3237 studies, 42 were selected based on inclusion and exclusion criteria. These studies were analyzed to evaluate the use of generative language models, such as GPT-3 and GPT-4, in various medical domains including oncology, cardiovascular, gastrointestinal, and ophthalmological care. The analysis revealed that most GenAI applications in healthcare rely on general purpose LLMs to provide treatment recommendations. Fine-tuning with domain-specific data and prompt engineering were found to significantly improve output quality and reliability. However, persistent challenges include lack of clinical validation, ethical concerns such as bias, and issues related to transparency and regulatory compliance. While GenAI demonstrates strong potential to support clinical decision-making, real-world deployment remains limited due to unresolved ethical and validation issues. Future research should prioritize the development of interpretable, domain-specific models and rigorous clinical trials to ensure safe and effective integration into healthcare settings.

The study aimed to discuss howArtificial Intelligence (AI) has been applied to improve the care of pulmonary diseases in preterm neonates. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines, a Systematic Literature Review(SLR)was conducted across six scientific databases, covering studies published between 2014 and 2024. From 7,565 initially identified studies, 35 were selected for analysis. Most studies focus on predicting Bronchopulmonary Dysplasia (BPD), whereas others address chronic and acute pulmonary diseases and additional clinical outcomes. There is a predominance of well-established AI models, particularly Random Forest and Logistic Regression, likely due to their recognition in the literature and consistent performance across clinical applications. Their familiarity may facilitate adoption in clinical contexts, especially when combined with commonly used evaluation metrics such as the Area Under the Receiver Operating Characteristic Curve (AUC-ROC), sensitivity, and accuracy. The most frequently used data for model training include gestational attributes (e.g., gestational age and gestational diabetes), followed by neonatal attributes (e.g., birth weight and sex) and clinical attributes (e.g., comorbidities and indicators of pulmonary function). Recent studies demonstrate promising progress in applying AI to predict pulmonary diseases in preterm neonates. However, methodological limitations, including small sample sizes and inconsistent handling of missing data, remain prevalent and hinder clinical implementation. Despite these challenges, the trends and findings identified in this review may contribute meaningfully to future research.

During the neonatal period, newborns are more susceptible to developing conditions and diseases due to their fragility of the transition and adaptation to the extrauterine environment. Neonatal sepsis is one of the leading causes of morbidity and mortality in newborns, particularly among preterm and low birth weight infants, requiring early diagnosis to reduce complications and deaths. In this work, we evaluate the performance of artificial intelligence models in predicting neonatal sepsis and also identify the attributes that most contribute impact on models' learning and their relationship with the disease, using real data from the state of Pernambuco, Brazil. The six machine learning models evaluated were AdaBoost, CatBoost, Gradient Boosting, LightGBM, Random Forest and XGBoost. Performance metrics ranged from 0.7213 to 0.8548, with AdaBoost and LightGBM achieving the best results, reaching a sensitivity above 0.8197 and a specificity of 0.8397 in all three experiments. SHAPley Additive exPlanations (SHAP) analysis revealed strong relationships between sepsis and attributes such as intracranial hemorrhage, prematurity, CPAP use, TTN presence, and epicutaneous access, all of which were highly associated with sepsis cases. We conclude that the artificial intelligence models demonstrated promising results in predicting neonatal sepsis, highlighting critical clinical attributes associated with the disease and identifying the most relevant predictors.

Chikungunya is an arbovirus capable of affecting the musculoskeletal system of infected individuals. Furthermore, it has the potential to progress from the

The reduction of fetal, neonatal, and infant mortality rates is crucial in maternal and neonatal care. The UN’s 2030 Agenda aims for a better world by 2030, highlighting goal 3: healthy lives for all ages, with a focus on reducing global maternal and child mortality. In this context, prenatal care plays a vital role in identifying pregnant women at risk and enabling proactive interventions to minimize adverse outcomes, whether mortality or morbidity. This work presents the ANGELS (An iNtelligent GEstational foLlow-up System). This work presents ANGELS, an intelligent gestational follow-up platform designed to integrate and operationalize multiple machine learning models within a unified decision support system for maternal and neonatal care. The platform provides integrated services through a modular, API-based architecture, enabling the incorporation of different predictive models throughout pregnancy, childbirth, and postpartum periods. The system is validated in this study through an exemplar case focused on congenital syphilis risk prediction using real-world Brazilian data, demonstrating howpreviously validated models can be integrated, extended, and deployed within the ANGELS platform. In the exemplar case of congenital syphilis risk prediction, the complementary version of ANGELS achieved substantial improvements, with specificity and precision increasing from approximately 60% in the basic version to values close to 95%, reflecting known trade-offs across evaluation metrics. The approach proved promising for practical incorporation into healthcare systems, aiming to provide more accurate and personalized monitoring tailored to individual needs. It is hoped that the ANGELS system may contribute to reduce preventable stillbirths, improve the quality of maternal and child care, and help achieve the UN’s 2030 Agenda goals. Additionally, the integration of ANGELS services into healthcare systems and/or programs will strengthen efforts to promote the well-being of pregnant women and children.

Leprosy remains a significant public health issue in Brazil, with a high number of cases reported annually, posing challenges in monitoring and control. Data-driven strategies are essential for enhancing disease surveillance and supporting the goals outlined in the Global Leprosy Strategy 2021-2030. Current tools to monitor leprosy in Brazil have several limitations and there is no documentation on the user requirements and empirical evaluation results of such tools. This can compromise leprosy monitoring effectiveness, timely response and data-informed decision-making in public health. This study presents the specification, development and evaluation of an interactive dashboard that leverages data from Brazil’s Disease Notification System (SINAN) to provide a detailed visualization of leprosy cases. The dashboard includes critical metrics on patient demographics, geographical distributions, and disease progression, facilitating data-driven decision-making. The process ensured that the final dashboard aligned with the specific needs and tasks of health professionals managing leprosy cases. The dashboard supports data analysis and trend visualization, empowering users to track leprosy patterns, manage cases more effectively, and anticipate resource needs. Usability tests indicated that users were able to complete analytical tasks using the dashboard with satisfactory degree of success. This work demonstrates the value of data visualization in public health information systems, providing a replicable model for tracking leprosy. The attributes, tasks and visual representations proposed can be reused by other researchers and practitioners to build similar tools. Also, by offering rapid access to actionable insights, the tool can enhance response capabilities and resource planning for leprosy and similar health challenges.

During the neonatal period, newborns are more susceptible to developing conditions and diseases due to their fragility ofthe transition and adaptation to the extrauterine environment. Neonatal sepsis is one of the leading causes of morbidityand mortality in newborns, particularly among preterm and low birth weight infants, requiring early diagnosis to reducecomplications and deaths. In this work, we evaluate the performance of artificial intelligence models in predicting neonatalsepsis and also identify the attributes that most contribute impact on models’ learning and their relationship with the disease,using real data from the state of Pernambuco, Brazil. The six machine learning models evaluated were AdaBoost, CatBoost,Gradient Boosting, LightGBM, Random Forest and XGBoost. Performance metrics ranged from 0.7213 to 0.8548, withAdaBoost and LightGBM achieving the best results, reaching a sensitivity above 0.8197 and a specificity of 0.8397 in all threeexperiments. SHAPley Additive exPlanations (SHAP) analysis revealed strong relationships between sepsis and attributessuch as intracranial hemorrhage, prematurity, CPAP use, TTN presence, and epicutaneous access, all of which were highlyassociated with sepsis cases. We conclude that the artificial intelligence models demonstrated promising results in predictingneonatal sepsis, highlighting critical clinical attributes associated with the disease and identifying the most relevant predictors.