Multivariate Modeling Approach Research Articles

Reimagining the future through the co-creation of social entrepreneurship in higher education: a multivariate prediction model approach

PurposeThis remastered analysis focuses on the impact of entrepreneurial interventions in higher education institutions (HEI), particularly in social entrepreneurship. The study evaluated the effectiveness of such interventions through a pre-and post-test approach, examining various skill sets in students. The primary goal was to analyze the influence of entrepreneurial training programs on students' competencies in social entrepreneurship by analyzing changes in personal behavior, leadership, innovation, social value and management skills before and after the educational interventions.Design/methodology/approachThe study employed a quasi-experimental design, analyzing pre-and post-test results in three distinct social entrepreneurship training experiences. The sample consisted of 304 participants, providing a comprehensive view of the impact of these interventions.FindingsThe main findings were: (1) Educational interventions in social entrepreneurship must emphasize strategies for self-awareness, emotional intelligence and personal development improvement. The analysis revealed significant improvements in these areas, indicating that targeted strategies in these domains are essential for enhancing the effectiveness of social entrepreneurship education. (2) The impact of educational interventions on these capabilities can be effectively evaluated using machine learning methods such as ordinary least squares (OLS) regression. This approach allows for the inclusion of variables such as gender, age or location, providing a comprehensive assessment of the interventions' impact. (3) The interventions were particularly effective in improving students' innovation and leadership competencies. The analysis demonstrated substantial enhancements in these areas, underscoring the success of the interventions in developing these critical skills. (4) The study highlighted the need for a more focused approach in future interventions, emphasizing the importance of management, social value and personal skills. Additionally, it pointed out the necessity of developing and utilizing appropriate tools to create and evaluate these interventions effectively.Practical implicationsThe study provides insights into improving educational interventions in social entrepreneurship to better develop essential skills in students.Originality/valueThis research introduces a significant approach to educational interventions for educational communities and decision-makers by demonstrating the effectiveness of entrepreneurial training for competencies in innovation and leadership, which are crucial for societal and economic development.

Impact of physicochemical and microbial drivers on the formation of disinfection by-products in drinking water distribution systems: A multivariate Bayesian network modeling approach

Impact of physicochemical and microbial drivers on the formation of disinfection by-products in drinking water distribution systems: A multivariate Bayesian network modeling approach

Impact of Internet and mobile communication on cyber resilience: A multivariate adaptive regression spline modeling approach

The spread of broadband Internet and the availability of mobile communication services expand access to digital services for businesses and the public alike. However, at the same time, it aggravates the problem of ensuring digital space security, protection against cyber threats, and the fight against cybercrime. This research aims to calculate the index of a country's resilience to cyber-hacking for 143 countries, to divide these countries into groups based on this resilience (high, above-average, average, below-average, and low), compare these results with those obtained on the basis of National Cyber Security Index (NCSI), and to identify the impact of the Internet and mobile communication prevalence in a country on this level. The selection of the countries is based on the availability of statistical data for 2022 in the databases of the Surfshark VPN service, and the International Telecommunication Union. The integral index of a country's resilience to cyber-hacking is calculated through the multiplicative convolution (with weighted geometric mean) of the number of breached accounts, the Internet penetration probability (penetration into users’ data through the Internet), and the breach density per thousand users. The influence of active mobile broadband subscriptions (per 100 inhabitants), mobile broadband basket (% of Gross National Income Per Capita), mobile cellular subscriptions (per 100 inhabitants), and total fixed broadband subscriptions on the integral index of a country's resilience to cyber-hacking is investigated using multivariate adaptive regression spline. According to the calculations, France, Iceland, Montenegro, the United States, and the United Arab Emirates were the least resistant to cyber hacking in 2022. For countries with high, above-average, and below-average levels of resilience to cyber-hacking, the most relevant factor is the number of active mobile broadband subscriptions (per 100 inhabitants). For countries with an average level, it is total fixed broadband subscriptions.

Investigating Functional Connectivity in Adolescent Depression and Suicide Attempt during Neurofeedback Sessions: A Multivariate Random Covariance Model Approach

Background A recent neurofeedback functional magnetic resonance imaging (NF, fMRI) study on depressed vs. healthy adolescents elicited differential functional connectivity (FC) amongst brain regions of interest (ROIs). Previous results employed univariate methods and included only two seed areas of FC (amygdala and hippocampus). In this study, we propose a new multivariate analysis for whole-network FC estimation. Methods Primary analyses concerned a pre-identified network of 17 salient ROIs reflecting key regions in self-processing and emotion regulation. A random covariance model (RCM) was applied to jointly estimate participant- and group-specific connectivity, where FC was measured by partial correlation conditioned on or adjusted for rest-of-network connectivity patterns. Secondary analyses concerned participant-specific network association with mental functioning changes and the AAL3 whole-brain atlas. Results New findings suggested that depressed adolescents with a suicide attempt expressed significantly higher positive FC between the left temporal gyrus and the left amygdala during NF, compared to negative FC in non-attempting depressed youth, while healthy controls displayed negative FC between the insula, inferior frontal gyrus to inferior parietal lobe connection, compared to mild negative connectivity in depressed adolescents. Previous cross-hemispheric findings in depressed vs. healthy adolescents were corroborated. Conclusion A multivariate RCM uncovered key ROI-pairwise connections differentiating FC patterns between depressed youth vs. healthy controls and among depressed youth, with and without a suicide attempt. Findings were strengthened by enhanced inference vs. univariate methods, and corroboration of previous NF secondary analyses demonstrated future utility for participant-specific study in association with clinical outcomes and/or whole-brain analyses with larger sample sizes.

A multi-modal evaluation of experimental pain and psychological function in women with carpometacarpal osteoarthritis

ObjectiveThumb carpometacarpal osteoarthritis (CMC1 OA) is a prevalent and debilitating condition that lacks effective treatments. Understanding the multidimensional pain experience across CMC1 OA disease stages is crucial to improving treatment outcomes. This study examined how radiographic CMC1 OA severity is associated with physical, psychological, and somatosensory function. MethodThirty-one women with early-stage (Eaton-Littler 1–2) or end-stage (Eaton-Littler 3–4) radiographic CMC1 OA completed validated questionnaires to assess pain, disability, and psychological function. Additionally, experimental pain was measured in each participant using quantitative sensory testing (QST) (mechanical, pressure, vibratory, thermal) at seven body sites (thenar, hypothenar, brachioradialis bi-laterally; quadriceps on affected side). Cohort differences (early-vs. end-stage) across all variables were analyzed using a multivariable modeling approach that included fixed effects and interactions; notably, age was controlled as a confounder. ResultsEnd-stage CMC1 OA participants had higher scores in the pain (p ​= ​0.01) and function (p ​= ​0.02) portions of the AUSCAN assessment, self-reported disability of the DASH questionnaire (p ​= ​0.04), and painDETECT scores (p ​= ​0.03), indicating greater pain and disability compared to early-stage participants. Additionally, end-stage CMC1 OA participants demonstrated reduced vibratory detection and heat pain thresholds at multiple body sites (p's ​< ​0.05), with significant interactions observed across the mechanical and cold stimuli. ConclusionFindings revealed women with end-stage CMC1 OA exhibited increased neuropathic pain characteristics and somatosensory loss compared to those with early-stage CMC1 OA. These results underscore the importance of addressing both peripheral and centralized pain mechanisms and the need for multimodal approaches in the treatment of CMC1 OA.

MULTIVARIATE PROBIT MODELING OF CLIMATE-SMART ADAPTATION CHOICES AMONG FARMERS IN SUB-SAHARAN AFRICA: GHANAIAN STUDY OF COCOA FARMERS

Climate change and variability pose significant challenges to cocoa production in many sub-Saharan African countries, and Ghana is no exception. Therefore, knowledge of how farmers react to climate change and its variability effects is crucial in developing policies and programs to guide the design and implementation of effective climate-smart adaptation strategies at the farm level in Africa. This study aims to examine climate-smart adaptation choices cocoa farmers make as coping strategies for climate change and variability in Ghana using a multivariate probit modeling approach. Following a multistage sampling procedure, primary data was collected from 720 cocoa farmers across the major cocoa-growing regions in Ghana. The study found major adaptation choices for cocoa farmers to include improved and resistant varieties, optimal fertilizer application, optimal pesticide application, adjusting planting/harvesting dates, shade tree management, crop diversification, non-crop diversification, and off-farm economic diversification. We further observed that farmers’ adaptation choices were significantly influenced by climate change and adaptation-related variables such as rainfall variability index, temperature variability index, climate change impact perception index, and CSA awareness as well as agroclimatic zone effects. Concerning the influence of farmer and farm-specific variables, the study identifies that sex, age, education, farm size, quality of extension services delivery, access to credit, and membership in farmer groups were significant predictors of farmers’ CSA decision choices. This study offers beneficial insights for policymakers and stakeholders on how to promote climate-smart adaptation methods among cocoa farmers in Ghana.

Multivariate analysis of metabolic state vulnerabilities across diverse cancer contexts reveals synthetically lethal associations.

Targeting the distinct metabolic needs of tumor cells has recently emerged as a promising strategy for cancer therapy. The heterogeneous, context-dependent nature of cancer cell metabolism, however, poses challenges in identifying effective therapeutic interventions. Here, we utilize various unsupervised and supervised multivariate modeling approaches to systematically pinpoint recurrent metabolic states within hundreds of cancer cell lines, elucidate their association with tumor lineage and growth environments, and uncover vulnerabilities linked to their metabolic states across diverse genetic and tissue contexts. We validate key findings via analysis of data from patient-derived tumors and pharmacological screens, and by performing new genetic and pharmacological experiments. Our analysis uncovers new synthetically lethal associations between the tumor metabolic state (e.g., oxidative phosphorylation), driver mutations (e.g., loss of tumor suppressor PTEN), and actionable biological targets (e.g., mitochondrial electron transport chain). Investigating the mechanisms underlying these relationships can inform the development of more precise and context-specific, metabolism-targeted cancer therapies.

The effect of total factor productivity on the food security and livelihood vulnerability of farm households in Bangladesh

Food security and livelihood vulnerability are important issues for the economic sustainability of developing countries like Bangladesh. This study examines the influence of total factor productivity (TFP) on the livelihood vulnerability and food security of rice farming households in Bangladesh. Data from 1,841 rice farming households were extracted from the Bangladesh Integrated Household Survey (2015 and 2018) conducted by the International Food Policy Research Institute. Various statistical methods, such as the stochastic frontier model, principal component analysis, path analysis using structural equation modeling, and multivariate regression, were employed to analyze the data. The study utilizes a multivariate modeling approach that combines the stochastic frontier model to determine TFP and sophisticated methodologies to estimate the livelihood vulnerability index (LVI) and women’s empowerment in agriculture index (WEAI). The LVI, household dietary diversity Score (HDDS), TFP, and WEAI scores were 0.454, 10.72, 0.703, and 0.717, respectively. The results indicate a significant relationship between TFP and both LVI and HDDS. Higher TFP is associated with lower LVI and higher HDDS among rice farming households, suggesting that improving TFP can enhance food security and reduce vulnerability. The multivariate regression analysis reveals that TFP, household wealth index, women’s empowerment in agriculture index, per capita food expenditure, household level welfare, and household size have a positive significant impact on HDDS, while TFP is negatively associated with LVI, per capita food expenditure and household size. The findings underscore the importance of increasing TFP to improve food security, reduce livelihood vulnerability, and achieve sustainable development goals in countries like Bangladesh. Higher TFP yields positive outcomes regarding household dietary diversity and livelihood vulnerability, highlighting the need for agricultural policies that prioritize TFP enhancement. Policymakers and professionals can use these findings as a roadmap to implement advanced agricultural policies to achieve food security and reduce livelihood vulnerability. Improving household dietary diversity and reducing livelihood vulnerability can be achieved by focusing on increasing TFP, enhancing household wealth, women’s empowerment, per capita food expenditure, household welfare, and household size. Therefore, increasing TFP should be considered in the design of policies aiming to achieve SDGs Goal 2.

Do try this at home: Age prediction from sleep and meditation with large-scale low-cost mobile EEG

Abstract EEG is an established method for quantifying large-scale neuronal dynamics which enables diverse real-world biomedical applications including brain-computer interfaces, epilepsy monitoring and sleep staging. Advances in sensor technology have freed EEG from traditional laboratory settings, making low-cost ambulatory or at-home assessments of brain function possible. While ecologically valid brain assessments are becoming more practical, the impact of their reduced spatial resolution and susceptibility to noise remain to be investigated. This study set out to explore the potential of at-home EEG assessments for biomarker discovery using the brain age framework and four-channel consumer EEG data. We analyzed recordings from more than 5200 human subjects (18-81 years) during meditation and sleep, to predict age at the time of recording. With cross-validated R2 scores between 0.3 - 0.5, prediction performance was within the range of results obtained by recent benchmarks focused on laboratory-grade EEG. While age prediction was successful from both meditation and sleep recordings, the latter led to higher performance. Analysis by sleep stage uncovered that N2-N3 stages contained most of the signal. When combined, EEG features extracted from all sleep stages gave the best performance, suggesting that the entire night of sleep contains valuable age-related information. Furthermore, model comparisons suggested that information was spread out across electrodes and frequencies, supporting the use of multivariate modeling approaches. Thanks to our unique dataset of longitudinal repeat sessions spanning 153 to 529 days from eight subjects, we finally evaluated the variability of EEG-based age predictions, showing that they reflect both trait- and state-like information. Overall, our results demonstrate that state-of-the-art machine learning approaches based on age prediction can be readily applied to real-world EEG recordings obtained during at-home sleep and meditation practice.

Size of Extranodal Extension in the Sentinel Lymph Node as a Predictor of Prognosis in Early-Stage Breast Cancer

IntroductionThe presence of extranodal extension (ENE) in sentinel lymph nodes (SLNs) can predict non-SLN metastases in breast cancer (BC) patients; however, the prognostic relevance of its extent remains controversial. The purpose of this study was to examine the predictive role of ENE in SLNs measured by its widest dimension (WD), highest dimension (HD), and the WD/HD ratio for non-SLN involvement, overall, and disease-free survival (OS, DFS) in cT1-2N0 BC patients with positive SLNs. Materials and MethodsA total of 511 women with cT1-2N0 BC and positive SLNs undergoing axillary lymph node dissection were retrospectively enrolled. The associations of ENE's WD, HD, and WD/HD ratio with non-SLN metastases, 5-year OS, and DFS were established through a multivariable modeling approach. ResultsSLNs were ENE-positive in 149 (29.16%) participants, and 133 (26.03%) had non-SLN metastases. During the median 60 (16-60)-month follow-up, 69 (13.50%) patients experienced recurrences, and 62 (12.13%) died. The numbers of SLNs, non-SLNs, and total axillary LNs involved differed between the ENE-negative and ENE-positive groups, as well as between the WD/HD ≤ 1.2 and WD/HD > 1.2 subgroups (all P-values were < .001). Multivariable analyses showed significant associations of the WD/HD ratio > 1.2 with non-SLN involvement, OS, and DFS (P-values were .003, < .001, and .005, respectively). DiscussionDespite no predictive value of ENE's WD and HD, the WD/HD ratio > 1.2 was an independent predictor of non-SLN involvement, mortality, and recurrence.ENE's WD/HD ratio could be a valuable indicator for cT1-2N0 BC individuals with positive SLNs for whom further axillary treatment may be beneficial.

PathIntegrate: Multivariate modelling approaches for pathway-based multi-omics data integration.

As terabytes of multi-omics data are being generated, there is an ever-increasing need for methods facilitating the integration and interpretation of such data. Current multi-omics integration methods typically output lists, clusters, or subnetworks of molecules related to an outcome. Even with expert domain knowledge, discerning the biological processes involved is a time-consuming activity. Here we propose PathIntegrate, a method for integrating multi-omics datasets based on pathways, designed to exploit knowledge of biological systems and thus provide interpretable models for such studies. PathIntegrate employs single-sample pathway analysis to transform multi-omics datasets from the molecular to the pathway-level, and applies a predictive single-view or multi-view model to integrate the data. Model outputs include multi-omics pathways ranked by their contribution to the outcome prediction, the contribution of each omics layer, and the importance of each molecule in a pathway. Using semi-synthetic data we demonstrate the benefit of grouping molecules into pathways to detect signals in low signal-to-noise scenarios, as well as the ability of PathIntegrate to precisely identify important pathways at low effect sizes. Finally, using COPD and COVID-19 data we showcase how PathIntegrate enables convenient integration and interpretation of complex high-dimensional multi-omics datasets. PathIntegrate is available as an open-source Python package.

Modeling rainfall extremes along the coastal and Northern parts of Ghana

The main objective of the study was to determine the appropriate distribution for extreme rainfall along the coastal and northern sectors of Ghana. For stakeholders and policymakers to make appropriate risk-mitigating measures to lessen the damage caused by flood and drought, it is necessary to make proper inferences about extreme rainfall. In this study, we used both the multivariate and univariate extreme value data analysis approaches. The Generalized Extreme Value (GEV) with the Block Maxima approach and Generalized Pareto Distribution (GPD) with the Peak over the threshold (that is all excesses and decluster peaks approaches) were used in this study. Historical gridded monthly maximum rainfall data from 1970 to 2020 were obtained from the Climatic Research Unit and were grouped as the coastal and northern stations. The Maximum Likelihood Estimation method was used to estimate the model parameters, and both the unit root test and the Mann-Kendall tests were used to test for trend in the data. With the multivariate extreme modelling approach, the logistic bivariate GEV model was chosen as the “best” model. However, the dependence value was 0.965, so the extreme rainfall should be modelled independently using the univariate extreme value approaches. Hence, based on the information criteria and analysis of deviance approaches, the GEV distribution was considered the “best” fit for the extreme rainfall dataset for the northern part of Ghana. In contrast, the GPD distribution was the “best” fit for the coastal station. Comparatively, for the volume of rainfall in the year 2020, the extreme rainfall is expected to be higher in the coastal station of Ghana in the next two years. Also, extreme rainfall in 2 years would not exceed the maximum occurrence of rainfall (279.267), which happened in September 2020 at the northern station of Ghana.

Identifying prognostic factors for survival in intensive care unit patients with SIRS or sepsis by machine learning analysis on electronic health records.

Systemic inflammatory response syndrome (SIRS) and sepsis are the most common causes of in-hospital death. However, the characteristics associated with the improvement in the patient conditions during the ICU stay were not fully elucidated for each population as well as the possible differences between the two. The aim of this study is to highlight the differences between the prognostic clinical features for the survival of patients diagnosed with SIRS and those of patients diagnosed with sepsis by using a multi-variable predictive modeling approach with a reduced set of easily available measurements collected at the admission to the intensive care unit (ICU). Data were collected from 1,257 patients (816 non-sepsis SIRS and 441 sepsis) admitted to the ICU. We compared the performance of five machine learning models in predicting patient survival. Matthews correlation coefficient (MCC) was used to evaluate model performances and feature importance, and by applying Monte Carlo stratified Cross-Validation. Extreme Gradient Boosting (MCC = 0.489) and Logistic Regression (MCC = 0.533) achieved the highest results for SIRS and sepsis cohorts, respectively. In order of importance, APACHE II, mean platelet volume (MPV), eosinophil counts (EoC), and C-reactive protein (CRP) showed higher importance for predicting sepsis patient survival, whereas, SOFA, APACHE II, platelet counts (PLTC), and CRP obtained higher importance in the SIRS cohort. By using complete blood count parameters as predictors of ICU patient survival, machine learning models can accurately predict the survival of SIRS and sepsis ICU patients. Interestingly, feature importance highlights the role of CRP and APACHE II in both SIRS and sepsis populations. In addition, MPV and EoC are shown to be important features for the sepsis population only, whereas SOFA and PLTC have higher importance for SIRS patients.

Analysis of The Influence of Digital Financial Literacy Understanding on The Profitability of Small and Medium Micro Enterprises

The objective of this study is to impact the financial knowledge of micro, small, and medium firms in order to improve their profitability. This study utilizes primary data obtained through interviews and questionnaires delivered to participants. This study is based on causal research. The research focuses on Micro, Small, and Medium Enterprises (MSMEs). This study employs non-probability sampling methods. Purposive sampling is the approach utilized. The study used the multivariate structural equation model approach. The research results indicate a notable correlation between financial attitude, financial conduct, financial knowledge, and profitability. Financial attitude significantly influences profitability, with higher levels of financial attitude leading to increased profitability. Furthermore, improved financial behavior positively impacts profitability, indicating that those with better financial habits can attain higher levels of profitability. Financial knowledge significantly impacts profitability, with better financial understanding leading to increased profitability. Thus, improving financial attitude, financial behavior and financial knowledge can increase individual or company profitability.

Implementation of Latent Gaussian Model in the Forecasting Process

This research aims to determine the implementation of the Latent Gaussian Model in the forecasting process. This research focuses on developing a forecasting model using the Multivariate Latent Gaussian Model (LGM) approach with shared components. which offers a more accurate representation without the assumption of stationarity and cointegration as it accommodates random components in the model. The forecasting results for the five KPPs are considered to have a very good level of accuracy with MAPE values &lt; 10%. This shows that LGM can achieve reliable forecasting when applied to the real life problems. This condition supports forecasting and can be an effective and targeted benchmark. The Latent Gaussian Model using the Bayesian Approach in parameter estimation can be utilized in forecasting Personal Income Tax Article 25/29. This is supported by the highly accurate MAPE value of 0.01%. The implementation of the developed model is not limited to forecasting Personal Income Tax Article 25/29. but can also be used in various other fields. With its hierarchical structure. the Bayesian approach proves to be an effective method for addressing complex modeling challenges.

In Search of Suitable Breeding Sites: Habitat Heterogeneity and Environmental Filters Determine Anuran Diversity of Western Madagascar

Environmental filtering shapes animal communities by preventing the colonization and persistence of certain species in a given habitat. More heterogenous environments are presumed to support a greater number of species and, consequently, increased species diversity, as environmental filters are also likely more heterogenous. Amphibians are especially sensitive to environmental influences due to distinct characteristics like permeable skin and low mobility. By analyzing the species richness and assemblage composition of tadpoles in 132 breeding ponds, we examined how the interplay of environmental variables shapes anuran species assemblages in breeding habitats of the dry forest of Western Madagascar. We found that environmental filtering is prevalent and habitat heterogeneity not only increases larval species richness but also alters species composition between these assemblages. Our study highlights the need for conserving heterogenous habitats to maintain local diversity. Furthermore, we recommend including multivariate modelling approaches to conservation efforts to acknowledge differences between specific habitats and beta diversity.

Modelling Latent Environmental Modifiers of Polygenic Risk for Late‐Life Depression Using Structural Equation Modelling

AbstractBackgroundOur understanding of the interplay between genetic and environmental factors (Gene x Environment Interaction, or GxE) determining mental health disorders has improved through the proliferation of genome‐wide interaction association studies (GWIAS) and targeted GxE analyses. Moreover, multivariate modelling approaches, such as structural equation modelling (SEM) and polygenic risk scores (PRS), offer opportunities for the integration of clinical and genome‐wide genotype data in building improved biopsychosocial models of mental illness aetiology and their response to treatment.MethodWe propose to construct a SEM framework to uncover the inter‐correlation and directed structure of mental health phenotypes by leveraging the joint predictive capacity of PRS for comorbid traits that share underlying biological and environmental risk pathways. The proposed model will be capable of linking latent constructs to their observed measurements; these will include disease severity, comorbidities and clinical histories, and behaviours and lifestyle factors such as physical and social activity.ResultOur gene‐by‐environment SEM (GESEM) will be initially developed and tested using four well‐characterized clinical cohorts for older adults diagnosed with late‐life depression and treated with antidepressants (CAN‐BIND, IRL‐GREY, STOP‐PD II and IMPACT; n = 1,238). The primary outcome will be antidepressant remission. Multiple PRS will be calculated to capture underlying genetic risk across vulnerable pathways which contribute to comorbidities. This selection will be made based on new, largely unpublished work from our group on the impact of PRS and targeted GxE studies on psychiatric outcomes across the lifespan. Each PRS will be calculated using both clumping and thresholding (PRSice‐2) and continuous shrinkage (PRS‐CS‐auto) methods across selected cohorts using well‐powered publicly available GWAS summary statistics. The multilevel GESEM model will include interactions between symptoms and comorbidities (i.e., observed measurements), which are caused by unobserved factors (i.e., latent constructs), and are subject to modification by background PRS. We will compare our GESEM model against existing SEM‐based approaches to GxE, including local SEM (LOSEM).ConclusionAn open‐source R package of the analytical code will be created and shared with the research community. This work has the potential to improve upon existing PRS‐based predictive models in a clinical setting.

The role of vaccination and environmental factors on outbreaks of high pathogenicity avian influenza H5N1 in Bangladesh

High Pathogenicity Avian Influenza (HPAI) H5N1 outbreaks continue to wreak havoc on the global poultry industry and threaten the health of wild bird populations, with sporadic spillover in humans and other mammals, resulting in widespread calls to vaccinate poultry. Bangladesh has been vaccinating poultry since 2012, presenting a prime opportunity to study the effects of vaccination on HPAI H5N1circulation in both poultry and wild birds. We investigated the efficacy of vaccinating commercial poultry against HPAI H5N1 along with climatic and socio-economic factors considered potential drivers of HPAI H5N1 outbreak risk in Bangladesh. Using a multivariate modeling approach, we estimated that the rate of outbreaks was 18 times higher before compared to after vaccination, with winter months having a three times higher chance of outbreaks than summer months. Variables resulting in small but significant increases in outbreak rate were relatively low ambient temperatures for the time of year, literacy rate, chicken and duck density, crop density, and presence of highways; this may be attributable to low temperatures supporting viral survival outside the host, higher literacy driving reporting rate, density of the host reservoir, and spread of the virus through increased connectivity. Despite the substantial impact of vaccination on outbreaks, we note that HPAI H5N1 is still enzootic in Bangladesh; vaccinated poultry flocks have high rates of H5N1 prevalence, and spillover to wild birds has increased. Vaccination in Bangladesh thus bears the risk of supporting “silent spread,” where the vaccine only provides protection against disease and not also infection. Our findings underscore that poultry vaccination can be part of holistic HPAI mitigation strategies when accompanied by monitoring to avoid silent spread.

Snakebite epidemiology, outcomes and multi-cluster risk modelling in Eswatini.

Halving snakebite morbidity and mortality by 2030 requires countries to develop both prevention and treatment strategies. The paucity of data on the global incidence and severity of snakebite envenoming causes challenges in prioritizing and mobilising resources for snakebite prevention and treatment. In line with the World Health Organisation's 2019 Snakebite Strategy, this study sought to investigate Eswatini's snakebite epidemiology and outcomes, and identify the socio-geographical factors associated with snakebite risk. Programmatic data from the Ministry of Health, Government of Eswatini 2019-2021, was used to assess the epidemiology and outcomes of snakebite in Eswatini. We developed a snake species richness map from the occurrence data of all venomous snakes of medical importance in Eswatini that was subjected to niche modelling. We formulated four risk indices using snake species richness, various geospatial datasets and reported snakebites. A multivariate cluster modelling approach using these indices was developed to estimate risk of snakebite and the outcomes of snakebite in Eswatini. An average of 466 snakebites was recorded annually in Eswatini. Bites were recorded across the entire country and peaked in the evening during summer months. Two cluster risk maps indicated areas of the country with a high probability of snakebite and a high probability of poor snakebite outcomes. The areas with the highest rate of snakebite risk were primarily in the rural and agricultural regions of the country. These models can be used to inform better snakebite prevention and treatment measures to enable Eswatini to meet the global goal of reducing snakebite morbidity and mortality by 50% by 2030. The supply chain challenges of antivenom affecting southern Africa and the high rates of snakebite identified in our study highlight the need for improved snakebite prevention and treatment tools that can be employed by health care workers stationed at rural, community clinics.

Integrating burned area as a complementary performance measure for daily fire danger assessment: A large-scale test

Fire indices are used to describe the weather conditions that influence fire ignition and fire behavior. Although many studies analyzed their performance on fire occurrence at daily resolution, few focused on their ability to capture the burned area, which is usually analyzed at the weekly or monthly scale. Cumulative Logarithmic Area Ranking Efficiency (CLARE) is a newly developed metric that takes burned area into account when assessing daily fire danger. The use of CLARE in addition to the Area Under the receiver operating characteristic Curve (AUC) in the selection process of fire indices or fire occurrence models provides a complementary metric that allows for the evaluation of a model's ability to assess burned area. We evaluated the CLARE performance in 11 regions ranging from the European Alps to the Mediterranean basin. We also assessed the impact of (i) different groups of input variables (meteorological variables vs. fire indices), (ii) model complexity in terms of number of variables, and (iii) the modeling approach (Generalized Linear Models vs. Maxent) on the performance of CLARE. We found that models that achieve a high AUC for predicting fire occurrence may fail to show a high performance when predicting burned area. Using a multi-variable modeling approach is likely to provide higher CLARE performance than using single-variable fire index models, especially among models that have high AUC. Moreover, using this approach led to better multi-variable meteorological model performance than single-variable fire index models for some regions. This may be particularly valuable for regions where the calculation of fire indices is not possible. Finally, the differences between the modeling approaches were mainly related to the region or input variable groups. Overall, our results highlight that including burned area in the fire danger assessment process is feasible across a wide range of environmental conditions and provides valuable insights.