Search

Objective We described the clinical profile and outcomes of patients with SLE and RA diseases reported to the Brazilian Health Information System with primary dengue infection. Methods Databases from the Brazilian Public Health Informatics System (SUS) were linked as the source of information. Three databases comprising different longitudinal information of lupus or rheumatoid arthritis (RA) patients under treatment and care through the Brazilian Health System were linked. Patients who had lupus ICD-9 code or RA ICD-9 code and their treatment approved by SUS were included in the study. In Study 1, we described the clinical characteristics of RA/lupus patients who had dengue infection. In Study 2, we compared RA/lupus patients with or without dengue for hospitalization rates after index dengue diagnosis for dengue-exposed or matching date for dengue-unexposed. Results We included 69 SLE and 301 RA patients with dengue. In the RA/lupus with dengue case series, hospitalization was found in 24.6% of lupus subjects and of 11.2% of RA subjects. It differed by geographic region ( p = 0.03), gender ( p = 0.05) and the use of azathioprine ( p = 0.02). Dengue was the most frequent reason for hospitalization reported (43.0%). Hospitalization due to dengue was noted in 12 (42.9%) dengue-exposed patients ( p = 0.02), while rheumatoid arthritis was reported as the cause of hospitalization in 22.2% of dengue-unexposed ( p = 0.005). Five deaths were reported among the dengue-exposed and none among dengue-unexposed. Bacterial infection was the most frequent cause of death. We found that the dengue exposure was associated with an increased risk of hospitalization outcome in RA and lupus patients (RR = 6.2; 95% CI: 2.99–12.94). Summary We found that when comparing RA/lupus patients with or without dengue, dengue-exposed patients had an increased rates of hospitalization and death. [ABSTRACT FROM AUTHOR]

Nowadays soccer is the most practiced sport in the world and moves a multimillionaire market. Therefore, a club that is able to recruit and develop talented players to theirs fullest potential has a lot of advantages and economic benefits. However, in most clubs the players are selected through scouts and coaches recommendation, with predictive success based mostly on intuition than other objective criteria. In addition, it is known that talent development and identification is a multifactorial process involving many characteristics. To this end, this paper proposes the creation of performance indicators based on multivariate statistical analysis. Usual principal components and factor analysis are performed to construct physical, technical and general score and copula modeling is proposed to create the consistency index, which generalizes the Z score method. With these indicators, a web-oriented expert system for analyzing sport data in real time via R software is proposed as a powerful tool for talent identification in soccer. This system, the so called iSports, allows the monitoring and continuous comparison of athletes in a simple and efficient way, taking into account essentials aspects, as well as identifying candidate talented that have above the average performance, that is, who stand out from the studied population of soccer players. In order to promote and popularize the access of information and the statistical science applied in the sports context, the iSports system can be used in any training center of the country, impacting the increase of knowledge of the athletes in training phase at any school, city or region. [ABSTRACT FROM AUTHOR]

Emergency department (ED) overcrowding is a global condition that severely worsens attention to patients, increases clinical risks and affects hospital cost management. A correct and early prediction of ED's admission is of high value and a motivation to adopt machine learning models. However, several of these studies do not consider data collected in textual form, which is a feature set that contains detailed information about patients and presents great potential for medical health care improvement. To this end, we propose and compare predictive models for admission that use both structured and unstructured data available at triage time. In total, our dataset comprised 499,853 pediatric ED's presentations (with an admission rate of 5.76%) of patients with age up to 18 years old observed over 3.5 years. Our best model consists of a 2-stage architecture with a deep neural network (DNN) to extract information from textual data followed by a gradient boosting classifier. This combined model achieved a value of 0.892 for the Area Under the Curve (AUC) in the test data. We highlight the importance of DNN-based text processing for better prediction, since the absence of text features resulted in AUC reduction of approximately two percentage points. Also, the feature importance of text was higher than that of the Manchester Triage System (MTS), which is a widely used risk classification protocol. These results suggest that activations from a trained DNN should be used in transfer learning setups in future studies. • Structured and unstructured triage data utilized to predict admission with high AUC. • Dataset from pediatric patients, with higher label imbalance than comparable works. • Experiments show that texts hold important signals that enhance model performance. • A new 2-stage architecture proposed: DNN for texts, followed by a boosting classifier. [ABSTRACT FROM AUTHOR]