Natural language processing (NLP) tasks can be addressed with several deep learning architectures, and many different approaches have proven to be efficient. This study aims to briefly summarize the use cases for NLP tasks along with the main architectures. This research presents transformer-based solutions for NLP tasks such as Bidirectional Encoder Representations from Transformers (BERT), and Generative Pre-Training (GPT) architectures. To achieve that, we conducted a step-by-step process in the review strategy: identify the recent studies that include Transformers, apply filters to extract the most consistent studies, identify and define inclusion and exclusion criteria, assess the strategy proposed in each study, and finally discuss the methods and architectures presented in the resulting articles. These steps facilitated the systematic summarization and comparative analysis of NLP applications based on Transformer architectures. The primary focus is the current state of the NLP domain, particularly regarding its applications, language models, and data set types. The results provide insights into the challenges encountered in this research domain.

UNASSIGNED: Although the Americas and Europe have historically dominated the global research landscape, emerging economies - Brazil, Russia, India, China, and South Africa (BRICS) have significantly increased their contributions in recent years. This article studies clinical trial trends in the BRICS nations between 2018 and 2022 and compares it with trends in the G7 nations (comprising Canada, France, Germany, Italy, Japan, the UK, the USA, and the European Union). This will help stakeholders in planning drug development strategies.
UNASSIGNED: Data were collected from the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) and the World Bank database. An electronic search was done for the total number of trials registered between January 1, 2018, and March 15, 2023. Information was analyzed based on the year of registration, therapeutic area, type of intervention, sponsorship, and type of special population. The trial density indices (TDIs) were calculated based on population (Xi) and gross domestic product (GDP) (Yi) using author-derived formulae.
UNASSIGNED: Altogether 2, 77, 536 trials from the BRICS and G7 were registered. China and the US had the most trials among the BRICS and G7, respectively. Between 2018 and 2022, the gap between the BRICS and G7 steadily reduced. The most common indication for clinical trials among the BRICS was cancer. Based on population, the TDI was the highest in China and the lowest in Russia. In proportion to the GDP, the TDI was maximum in Russia and minimum in India.
UNASSIGNED: There is a remarkable reduction in the gap in clinical trial trends between the BRICS and G7 nations. Among the BRICS, India and China are at the forefront in drug development. There is scope for improvement in trial density based on India\'s population and GDP. Stakeholders are likely to utilize the strengths of the BRICS as an attractive destination for investment in this area.

BACKGROUND: Healthcare delivery is undergoing radical changes that influence effective infection prevention and control (IPC). Futures research (short: Futures), the science of deliberating on multiple potential future states, is increasingly employed in many core societal fields. Futures might also be helpful in IPC to facilitate current education and organisational decisions. Hence, we conducted an initial survey as part of the IPC Crystal Ball Initiative.
METHODS: In 2019, international IPC experts were invited to answer a 10-item online questionnaire, including demographics, housekeeping, and open-ended core questions (Q) on the \"status of IPC in 2030\" (Q1), \"people in charge of IPC\" (Q2), \"necessary skills in IPC\" (Q3), and \"burning research questions\" (Q4). The four core questions were submitted to a three-step inductive and deductive qualitative content analysis. A subsequent cross-case matrix produced overarching leitmotifs. Q1 statements were additionally coded for sentiment analysis (positive, neutral, or negative).
RESULTS: Overall, 18 of 44 (41%) invited experts responded (from 11 countries; 12 physicians, four nurses, one manager, one microbiologist; all of them in senior positions). The emerging leitmotifs were \"System integration\", \"Beyond the hospital\", \"Behaviour change and implementation\", \"Automation and digitalisation\", and \"Anticipated scientific progress and innovation\". The statements reflected an optimistic outlook in 66% of all codes of Q1.
CONCLUSIONS: The first exercise of the IPC Crystal Ball Initiative reflected an optimistic outlook on IPC in 2030, and participants envisioned leveraging technological and medical progress to increase IPC effectiveness, freeing IPC personnel from administrative tasks to be more present at the point of care and increasing IPC integration and expansion through the application of a broad range of skills. Enhancing participant immersion in future Crystal Ball Initiative exercises through simulation would likely further increase the authenticity and comprehensiveness of the envisioned futures.

UNASSIGNED: Given the significance of addressing neonatal mortality in pursuing the 2030 Sustainable Development Goal on child health, research focus on this area is crucial. Despite the persistent high rates of neonatal mortality rate (NMR) in Bangladesh, there remains a notable lack of robust evidence addressing inequalities in NMR in the country. Therefore, this study aims to fill the knowledge gap by comprehensively investigating inequalities in NMR in Bangladesh.
UNASSIGNED: The Bangladesh Demographic and Health Survey (BDHS) data from 2000 to 2017 were analyzed. The equity stratifiers used to measure the inequalities were wealth status, mother\'s education, place of residence, and subnational region. Difference (D) and population attributable fraction (PAF) were absolute measures, whereas population attributable risk (PAR) and ratio (R) were relative measures of inequality. Statistical significance was considered by estimating 95% confidence intervals (CIs) for each estimate.
UNASSIGNED: A declining trend in NMR was found in Bangladesh, from 50.2 in 2000 to 31.9 deaths per 1000 live births in 2017. This study detected significant wealth-driven (PAF: -20.6, 95% CI: -24.9, -16.3; PAR: -6.6, 95% CI: -7.9, -5.2), education-related (PAF: -11.6, 95% CI: -13.4, -9.7; PAR: -3.7, 95% CI: -4.3, -3.1), and regional (PAF: -20.6, 95% CI: -27.0, -14.3; PAR: -6.6, 95% CI: -8.6, -4.6) disparities in NMR in all survey points. We also found a significant urban-rural inequality from 2000 to 2014, except in 2017. Both absolute and relative inequalities in NMR were observed; however, these inequalities decreased over time.
UNASSIGNED: Significant variations in NMR across subgroups in Bangladesh highlight the need for comprehensive, and targeted interventions. Empowering women through improved access to economic resources and education may help address disparities in NMR in Bangladesh. Future research and policies should focus on developing strategies to address these disparities and promote equitable health outcomes for all newborns.

BACKGROUND: Working mothers are in a situation where they have to manage both their job and maternal roles simultaneously. We aim to show the relationship between mothers\' breastfeeding behavior and working status, as well as the effect of their working conditions on the continuation of breastfeeding.
METHODS: This study examined 3557 (weighted 3490) child-mother pairs from Turkey\'s Demographic Health Survey data in 2008, 2013, and 2018 with a complex sample multiple logistic regression analysis to explore the relationship between breastfeeding and maternal employment.
RESULTS: In our study, 35.5% of the mothers had never worked, 18.6% were still working, and 45.9% had worked in the past but were not currently working. When breastfeeding percentages were examined based on maternal working status, mothers who worked in 2008 had significantly lower breastfeeding percentages than those who had never worked or had worked previously but not currently. In both 2013 and 2018, after legal regulations, maternal employment didn\'t affect breastfeeding. After adjusting for confounding factors, maternal employment in the industry sector had lower breastfeeding percentages [AOR:0.06 (95% CI: 0.01-0.48)] than those working in agriculture.
CONCLUSIONS: It seems that adhering to legal regulations such as paid maternity leave and lactation leave can promote breastfeeding. It is necessary to raise awareness of mothers working in the industry about the continuation of breastfeeding and to monitor their legal rights.

UNASSIGNED: With socioeconomic development, the increase of older pregnancies and multiparas has brought risks to mothers and infants.
UNASSIGNED: As parities increased, the proportion of women of advanced maternal age (AMA) and non-local domicile increased, while the proportion of women with higher education levels decreased. Women with ≥3 parities are more likely to have preterm birth (PTB) and macrosomia.
UNASSIGNED: A comprehensive analysis of pregnancy traits among women at different parities offers a robust foundation for tailored strategies against adverse pregnancy outcomes.

UNASSIGNED: Pneumonia is one of the most common causes of hospital admissions in the United States and remains a major cause of death. However, less is known regarding the mortality burden from pneumonia in the United States and how this burden has changed over time.
UNASSIGNED: Death rates from causes related to pneumonia were determined using the CDC Wide-ranging Online Data for Epidemiologic Research (WONDER) data from 1999-2019. Pneumonia deaths were calculated for the overall population as well as for sociodemographic subgroups. We also analysed changes in death rates over time.
UNASSIGNED: Overall, 2.1% of total US deaths during the period between 1999 and 2019 were due to pneumonia (2.6% in 1999 and 1.5% in 2019). Mortality declined over time for both men and women, and across most age cohorts, as well as all racial, urbanisation, and regional categories. Rates of pneumonia deaths were higher among males as compared to females (age-adjusted mortality rate ratio (AAMRR) = 1.35; 95% CI: 1.34-1.35). Compared to White Americans, Black Americans had the highest pneumonia-related mortality rates of any racial group (AAMRR = 1.11; 95% CI: 1.10-1.11).
UNASSIGNED: Rates of pneumonia-related death have decreased in the United States in recent decades. However, significant racial and gender disparities remain, indicating the need for more equitable care.

In response to concerns regarding overprescribing of psychotropic medication in children/adolescents, this study examined trends in psychotropic medication use in Ireland by age group and gender. A retrospective, repeated, cross-sectional study of the Irish pharmacy claims database was conducted. Yearly prevalence of children/adolescents receiving dispensed psychotropic medications was analysed from January 2017 to December 2021 and compared across years, age groups (5-15 years, and stratified as 5-11 and 12-15 years) and gender. Yearly prevalence was defined as the mean number of patients in receipt of medication per month per 1000 eligible population during a given calendar year. Negative binomial regression was used to examine the association of year, age group and gender on prevalence. Prevalence ratios (PRs) per year (average change in prevalence between each year) were presented with 95% confidence intervals (CIs). The prevalence of included psychotropic medications dispensed in the 5-15 years group increased from 6.41 (95% CI: 6.22, 6.59) in 2017 to 8.46 (95% CI: 8.26, 8.68) in 2021 per 1000 eligible population (32% increase). The PR per year (adjusting for age category and gender) was 1.07 (95% CI: 1.035, 1.107; p < 0.001). An increasing trend over time was also observed for all individual drug classes. These findings suggest increased use of psychotropic medication in children/adolescents from 2017 to 2021. However, despite increased prevalence over time, comparison with the literature shows that psychotropic medication use in Ireland remains lower than international comparators.

UNASSIGNED: As the volume of technology-assisted total hip arthroplasty (THA) increases, there is a need to characterise the outcomes of robotic-assisted (RA) and computer-navigated (CN) THA. The goal of this study was to assess outcomes and opioid consumption following CN-THA and RA-THA compared to conventionally-instrumented (CON) THA.
UNASSIGNED: The Premier Database was queried for all patients who underwent primary, elective THA from 2015-2020. Patients were divided into 3 groups: CN, RA, or CON-THA. Yearly usage trends were assessed. Univariate and multivariate analyses were performed to assess the 90-day risk of postoperative complications. Opioid consumption was reported in morphine milligram equivalents (MME) for postoperative days (POD) 0 and 1.
UNASSIGNED: Overall, 474,707 elective THAs were identified (95.7% CON, 2.1% CN, 2.2% RA. After accounting for confounders, CN-THA patients were at decreased risk for periprosthetic joint infection (PJI) (aOR: 0.55, p < 0.001) and dislocation (aOR 0.45, p < 0.001), but increased risk for blood transfusion (aOR 1.97, <0.001) compared to CON-THA. RA-THA patients were at decreased risk of dislocation (aOR:0.66, p < 0.001) but increased risk for transfusion (aOR 1.20, p < 0.001), prosthesis breakage (aOR 3.88, p < 0.001), and periprosthetic fracture (aOR 1.72, p < 0.001). Opioid consumption for CN-THA patients was lower on POD1 and lower for RA-THA patients POD0 and 2 compared to CON-THA.
UNASSIGNED: CN-THA was associated with reduced rates of PJI and dislocation, but increased rates of blood transfusion while RA-THA was associated with decreased rates of dislocation, but increased rates of blood transfusion, prosthesis complications, and periprosthetic fracture compared to CON-THA. Technology-assisted THA was associated with lower postoperative opioid consumption.

UNASSIGNED: To investigate 20-year trends in social participation among the oldest old (77+ years) in Sweden and assess the extent to which changes in educational attainment and functional abilities explain these trends.
UNASSIGNED: Seven waves of the Swedish Panel Study of Living Conditions of the Oldest Old (SWEOLD) spanning 2002-2021 were used with a repeated cross-sectional design. To analyse the association between time and social participation we employed the Karlson-Holm-Breen method of decomposition. The study focused on informal social participation (contact with friends and family), leisure participation (public or semi-public gatherings), and formal participation (organisational engagement and study circle attendance).
UNASSIGNED: Both leisure and informal participation peaked in 2014 and declined in 2021, influenced by the COVID-19 pandemic, whereas formal participation showed a slight increase in 2021. Total participation increased at least until 2011. Overall, older adults have increased their levels of social participation in recent decades, disregarding the influence of the pandemic. Decomposition analysis revealed that population-level changes in educational attainment and functional abilities explained a substantial portion of the observed trends.
UNASSIGNED: As the proportion of older adults continues to rise, it becomes increasingly important to understand the developments and drivers of behavioural change in the older population. As more people are socially active, there may be increasing differences between those participating and those not - which could lead to increased inequalities. The observed trend in increasing participation, influenced by changes in education and health, emphasises the importance of fostering age-friendly environments and addressing potential social inequalities among older adults.