Bayesian Network Research Articles

Application of artificial intelligence to study the causality in public health: a systematic review

Abstract Introduction In epidemiology, establishing causation requires meeting rigorous criteria beyond mere association. Traditional methodologies often struggle to provide definitive causal conclusions due to confounding variables. However, leveraging artificial Intelligence (AI) and machine Learning (ML) offers a promising solution by modeling complex interactions among relevant factors. This systematic review explores the AI techniques utilized to uncover causal relationships in public health applications. Methods A literature search was conducted in Pubmed, Web of Science, and Scopus, employing the following search strategy: (Causalit*[Title/abstract] OR Causation*[Title/abstract] OR Causal[Title/abstract]) AND (‘Machine Learning’[Title/abstract] OR “Deep Learning” [Title/abstract] OR ‘Artificial Intelligence’ [Title/abstract] OR Algorithm* [Title/abstract]). The search aimed to identify studies encompassing causal inference techniques utilizing ML and AI methodologies in public health. Results From a total of 28,230 articles published up to August 2023, after removing duplicates (11,158) and articles that did not meet the inclusion criteria (16,168 post title/abstract screening, and 856 post full-text screening), the systematic review included 48 articles. Bayesian additive regression trees, Bayesian network models, causal forests analysis, and graphical causal models emerged as the prevailing causal methodologies. AI-driven causal methodologies have been investigated across various domains within public health, encompassing disease prediction, evaluation of treatment efficacy, identification of risk factors, analysis of health behaviors, precision Public Health initiatives, and environmental health assessment. Discussion While our systematic review affirms the potential of AI-driven causal approaches across several public health domains, further research is warranted before these emerging techniques can be effectively integrated into real settings. Key messages • In epidemiology, establishing causation requires meeting rigorous criteria beyond mere association and traditional methodologies often struggle to provide definitive causal conclusions. • This systematic review explores the AI techniques utilized to uncover causal relationships in public health applications.

Comparison of Surgical Techniques in Managing Craniosynostosis: Systematic Review and Bayesian Network Meta-Analysis

Open total cranial vault reconstruction (CVR) is the common procedure in managing craniosynostosis, yet more techniques have been introduced as alternatives, namely endoscopic suturectomy (ES), endoscopy-assisted craniectomy (EC), spring-assisted surgery (SAS), strip craniectomy with helmet (SC), Pi craniectomy (PiC), Pi plasty (PiP), and Renier’s “H” technique (RH). The aim of this study was to compare the effectiveness of craniosynostosis surgeries in improving the cephalic index of the patients. Studies published until 7 March 2024 reporting CVR, ES, SAS, SC, RH, and PiP as definitive craniosynostosis management with the cephalic index as the outcome were included. Bayesian network meta-analysis and pair-wise meta-analysis were performed using a random-effects model based on standardized mean difference (SMD) and 95% confidence interval (CI). Nine studies published in 2008–2024 recruiting a total of 464 craniosynostosis patients (age: 18–61 months) were included in this meta-analysis. EC (SMD = 0.23 [95%CI: −5.47 to 5.63]; p = 0.935), PiP (SMD = −0.07 [95%CI: −9.27 to 8.79]; p = 0.988), ES (SMD = −0.59 [95%CI: −6.07 to 4.94]; p = 0.834), PiC (SMD = −1.16 [95%CI: −8.89 to 6.35]; p = 0.765), RH (SMD = −0.96 [95%CI: −6.62 to 4.53]; p = 0.736), SAS (SMD = −0.86 [95%CI: −8.25 to 6.18]; p = 0.815), and SC (SMD = −1.79 [95%CI: −9.05 to 5.28]; p = 0.624) were found to be as effective as CVR in improving the cephalic index. Network meta-analysis suggests that PiP is the most effective among these techniques (rank 1 probability = 0.273). According to the rank probabilities of our model the order of techniques from the most to the least effective is as follows: EC > CVR > PiP > ES > SAS > RH > PiC > SC.

Prognostic impact of secondary prevention medical therapy in patients with Takotsubo cardiomyopathy: a Bayesian network meta-analysis

Abstract Background Takotsubo cardiomyopathy (TTC) is a form of transient left ventricular systolic dysfunction without evidence of complicated coronary artery disease. Although initially considered a benign condition, recent studies have demonstrated a significant risk of acute phase complications and a mortality rate between 1-8%. Efficacy of medical therapy in secondary prevention of all-cause mortality and TTC recurrence is not well established. We performed a systematic review and network meta-analysis to evaluate prognostic impact of secondary prevention medical therapy in patients with TTC. Methods PubMed, MEDLINE, EMBASE, CENTRAL, and Web of Science were searched up to January 6, 2024. Eligible studies included multivariable-adjusted or propensity matched studies of patients receiving medical therapy with B-blockers (BB), renin–angiotensin–aldosterone system (RAAS) inhibitors, aspirin and statins after an index presentation with TTC. The primary outcome was all-cause mortality at any time point. Secondary outcome was TTC recurrence. Random-effect hierarchical Bayesian meta-analyses was performed. Results We identified 15 observational studies. TTC mortality was reported in 435 (4.4%) out of 9735 patients. Mean age was 70.2 ± 4 years and 8858 patients (91.5%) were females. There were no differences in all-cause mortality when comparing BB versus control (OR 0.83; 95% CI 0.47 to 1.47, p=0.52); BB versus RAAS inhibitors (OR 1.03; 95% CI 0.61-1.63), BB versus aspirin (OR 1.02; 95% CI 0.471-2.25) and BB versus statins (OR 0.871; 95% 0.432-1.75). No significant differences were also observed for TTC recurrence. Conclusion This network meta-analysis found no significant difference in all-cause mortality or recurrence rates in Takotsubo cardiomyopathy with the use of secondary prevention medical therapies, including beta-blockers, RAAS inhibitors, aspirin, and statins

Forecasting cognitive functioning: an artificial intelligence approach with Dynamic Bayesian Networks

Abstract Background Several potentially modifiable risk factors are associated with subjective cognitive decline (SCD). However, developmental patterns of these risk factors have not been used before to forecast later SCD. Practical tools for the prevention of cognitive decline are needed. We examined multifactorial trajectories of risk factors, and their associations with SCD using an artificial intelligence approach to build a score that forecasts later SCD. Methods Five repeated surveys (2000-2022) of the Helsinki Health Study (n = 8960, 79% women, aged 40-60 at Phase 1) were used to build dynamic Bayesian networks (DBN) for estimating the odds of SCD. A score-based approach was implemented for learning DBN using the quotient normalized maximum likelihood criterion. The model was used to predict SCD based on the history of consumption of fruit and vegetables, smoking, alcohol consumption, leisure-time physical activity, body mass index, and insomnia symptoms, adjusting for sociodemographic covariates. Results Of the participants, 31-48% reported decline in memory, learning, and concentration in 2022. Physical activity was the strongest predictor of SCD in a 5-year interval, with an odds ratio of 0.76 (95% Bayesian credible interval 0.59-0.99) for physically active compared to inactive participants. Alcohol consumption showed a U-shaped relationship with SCD. Other risk factors had minor effects. Conclusions A new online risk score tool was developed that enables individuals to inspect their own risk profiles, as well as explore potential targets for interventions and their estimated contributions to later SCD. Dynamic decision heatmap was presented as a communication tool to be used at healthcare consultations. Key messages • Potentially modifiable health-related behaviours such as physical activity can contribute to subjective cognitive decline. • Artificial intelligence can be applied to discover the complex interactions between risk factors and subjective cognitive decline, to support preventive healthcare at individual and population levels.

Deep Learning in Finance: A Survey of Applications and Techniques

Machine learning (ML) has transformed the financial industry by enabling advanced applications such as credit scoring, fraud detection, and market forecasting. At the core of this transformation is deep learning (DL), a subset of ML that is robust in processing and analyzing complex and large datasets. This paper provides a comprehensive overview of key deep learning models, including Convolutional Neural Networks (CNNs), Long Short-Term Memory networks (LSTMs), Deep Belief Networks (DBNs), Transformers, Generative Adversarial Networks (GANs), and Deep Reinforcement Learning (Deep RL). Beyond summarizing their mathematical foundations and learning processes, this study offers new insights into how these models are applied in real-world financial contexts, highlighting their specific advantages and limitations in tasks such as algorithmic trading, risk management, and portfolio optimization. It also examines recent advances and emerging trends in the financial industry alongside critical challenges such as data quality, model interpretability, and computational complexity. These insights can guide future research directions toward developing more efficient, robust, and explainable financial models that address the evolving needs of the financial sector.

Unveiling the impact of siren noise exposure on cognitive function and mental health among firefighters

Firefighters reside in inherently noisy environments with sirens and emergency alarms. This noise can negatively affect the performance of firefighters and further pose a safety risk. This study investigates the relationship between siren noise exposure and its mental health disorders. In doing so, the study focuses on cognitive function among firefighters, using a Bayesian network model for the first time. This case-control study involved 92 firefighters and was conducted in 2023. Participants were categorized into case and control groups based on their exposure levels to siren noise. Noise exposures were initially measured according to ISO 9612 standards. The questionnaires covered demographic information, scales of their depression, anxiety, and stress, the Buss-Perry aggression questionnaire, and the cognitive failure questionnaire. GeNIe academic software version 2.3 was employed for Bayesian network analysis. The study results indicated a strong association that heightened siren noise exposure with a 100% probability increases depression (10%), stress (14%), anxiety (13%), and aggression (16%). In addition, each variable–noise exposure, depression, stress, anxiety, and aggression was linked to cognitive failure by 8%, 28%, 8%, and 13%, respectively. The combination of variables including noise exposure, depression, anxiety, and other mental health disorders had a 57% increase in cognitive failure. The most influence values were observed between depression and cognitive failures (0.564). The results of this study demonstrated the effect of siren noise exposure and its mental health disorders on cognitive function among firefighters. The results further highlight proper implementation of measures to mitigate these effects, and further required investigation on their effectiveness.

Uncertainty-driven dynamic ensemble framework for rotating machinery fault diagnosis under time-varying working conditions

Multi-scale ensemble learning combines different scales of feature resolution, thereby improving fault diagnostic accuracy. However, the effectiveness of different information scales in characterizing fault features under time-varying speed conditions varies with speed. It is difficult for existing ensemble strategies to ensure the effectiveness of feature information when ensemble multi-scale feature information is involved. Accordingly, we propose an uncertainty-driven dynamic ensemble Bayesian convolutional neural network (DEBCNN) framework. The uncertainty of the results of different scale models was used to dynamically determine their weights in the ensemble framework, which reduced the influence of irrelevant features on the diagnostic results. By employing the proposed dynamic ensemble strategy, the ensemble framework can utilize fault feature information corresponding to different rotational speeds in the final diagnostic results. Experiments on motor and bearing datasets illustrate the superiority of this strategy over other techniques. This study provides useful insights for further research in the field of fault diagnosis of rotating machinery at time-varying speeds.

A real-world pharmacovigilance study of FDA Adverse Event Reporting System (FAERS) for gemcitabine

ABSTRACT Background Gemcitabine is widely used in the treatment of various cancers. This study aims to evaluate gemcitabine-associated adverse events (AEs) using the Food and Drug Administration Adverse Event Reporting System (FAERS) database. Research design and methods We analyzed data spanning from January 2004 to June 2023. Employing reporting odds ratio (ROR) and Bayesian confidence propagation neural network (BCPNN) algorithms, we identified AEs with positive signals in patients administered gemcitabine. Results Out of 16,623,939 reports, 23,645 involved gemcitabine as the ‘primary suspected (PS)’ resulting in 74,306 AEs. Consistent with the reports in the specification and clinical trials, thrombocytopenia, pyrexia, neutropenia, and anemia were the most common AEs. Notably, our study identified some unexpected AEs such as abdominal pain, pleural effusion, ascites, and gastrointestinal hemorrhage, among others. The most significant SOC was ‘Blood and lymphatic system disorders’. The median onset time for gemcitabine-related AEs was 24 days (interquartile range [IQR] 6–82 days), with most cases occurring within the initial 30 days following gemcitabine administration. Conclusion Gemcitabine is associated with a broad spectrum of AEs affecting multiple organ systems, with a notable incidence of hospitalization. The study highlights both expected and unexpected AEs, which could enhance future clinical applications and safety of gemcitabine.

Intravenous thrombolysis or antiplatelet therapy for acute nondisabling ischemic stroke: A systematic review and network meta-analysis.

Uncertainties remain on the optimal treatment for acute minor stroke with nondisabling symptoms. The two most common therapeutic approaches are intravenous thrombolysis (IVT) and antiplatelet therapy, notably dual antiplatelet therapy (DAPT). We synthesized data from the literature to compare IVT to DAPT and identify the best treatment for this population. We systematically searched Pubmed, Web of Science and the Cochrane Library for randomized trials and observational studies comparing IVT, aspirin, and/or DAPT, started within 24 h of symptom onset in patients with minor stroke (NIHSS ⩽ 5) and nondisabling symptoms. Random-effects Bayesian network meta-analysis was conducted. The primary outcome was excellent functional outcome at 3 months (mRS 0-1). Secondary outcomes included mRS 0-2, symptomatic intracranial hemorrhage, mortality, and recurrent stroke. Four randomized trials and 2 observational studies (5897 patients for the analysis of the primary outcome) were included. Compared with IVT (alteplase), DAPT (clopidogrel + aspirin) was significantly associated with higher odds of mRS 0-1 (OR = 1.52, 95% CrI, 1.09-2.35), but aspirin alone was not (OR = 1.36, 95% CrI, 0.87-2.30). DAPT was also associated with lower odds of symptomatic intracranial hemorrhage than alteplase (OR = 0.14, 95% CrI, 0.03-0.91). There were no significant differences between treatment groups regarding the other outcomes. For each outcome, the ranking for the best treatment was DAPT, then aspirin, and then IVT. This network meta-analysis suggests that DAPT may be the optimal treatment for acute nondisabling stroke, with higher odds of excellent functional outcome compared with IVT.Registration: PROSPERO ID: CRD42024522038.

Data-driven Bayesian risk assessment of factors influencing the severity of marine accidents in port areas

The most prevalent types of ship accidents in port areas are allisions, collisions, and groundings. A comprehensive risk assessment is needed to prevent and mitigate these accidents and their consequences. This study evaluates the risk of such accidents in port areas by presenting a model that elucidates the relationship between risk-identifying factors (RIFs) and accident severity. In this context, the RIFs are determined by analyszing the reports of 528 accidents that occurred in port areas between 1995 and 2023. Subsequently, the model is created by analysing the data derived from these reports using the Tree Augmented Naive Bayes (TAN) algorithm, which is an approach of the data-driven Bayesian network method. The findings of the study indicate that accident type, wind, ship age, and vessel type are the most influential factors in predicting the severity of accidents in port areas. It is thought that the model will assist port authorities in identifying operational risks contributing to accidents and in formulating preventive regulations.

Applying Bayesian Networks for Detecting Bank Fraudulent Transactions in Nigeria

Bank fraud is an increasingly prevalent issue, causing substantial financial losses for both financial institutions and customers. Also, with the increasing prevalence of bank fraud, especially facilitated by online banking and digital payments, there is a pressing need for potential expertise deployment and advanced fraud detection techniques in real-world banking systems. This study specified the Bayesian Network models for detecting fraudulent bank transactions. Bayesian Networks offer a probabilistic graphical modeling approach that can effectively capture complex relationships and dependencies within financial data. Thus, the research aimed at developing a custom Bayesian network model trained on a large dataset of bank transactions comprising over one million bank transactions to classify instances as fraudulent or non-fraudulent. Python 3.11.10 was used for the graphical representations. Down-sampling was employed to reduce the dataset from its initial size of 1 million observations to 17,650 observations, ensuring a balanced representation of both fraud and non-fraud instances. Various estimation techniques: Maximum Likelihood, Bayesian Networks, and Expectation Maximization were employed and evaluated to learn the model parameters. The model's performance was measured using metrics: accuracy, precision, recall, F1-score, and ROC-AUC. The Bayesian Networks estimator achieved an overall accuracy of 66.18%, precision of 67.73%, F1-score of 64.06%, ROC-AUC of 66.13%, Recall of 60.77%, Sensitivity of 60.77% and Specificity of 71.50%. Also, the Maximum Likelihood achieved an overall Accuracy of 66.77%, Precision of 69.22%, F1-score of 63.96%, ROC-AUC of 66.71%, Recall of 59.45%, Sensitivity of 59.45% and Specificity of 73.97%. Likewise, the Expectation Maximization achieved an overall Accuracy of 66.83%, Precision of 69.31%, F1-score of 64.00%, ROC-AUC of 66.77%, Recall of 59.45%, Sensitivity of 59.45% and Specificity of 74.09%. On the confusion matrix, the model correctly classified 1272 instances as Non-Fraudulent transactions (True Negative). Also, it was observed that the model incorrectly classified 507 instances as fraudulent transactions when they were Non-Fraudulent (False Positive). The model similarly incorrectly classified 687 instances as Non-Fraudulent transactions when they were Fraudulent (False Negative). Finally, it correctly classified 1064 instances as fraudulent transactions (True Positive). These results demonstrated the Bayesian Network's ability to identify fraudulent transactions while minimizing false alarms accurately. The findings highlight the potential of Bayesian Networks as a robust framework for fraud detection in the banking sector, contributing to enhanced security and reduced financial losses. The developed model can be introduced into existing banking systems to strengthen fraud prevention strategies.

Characterization of drought propagation over the Tibetan Plateau

Study regionThis study was carried out on the Tibetan Plateau (TP), which contains multiple important ecosystems, is the source of many rivers in China, and is experiencing significant climate change. Study focusClarifying drought propagation characteristics is crucial for understanding the mechanism of drought development and benefiting drought mitigation and early warning schemes. However, there is currently a notable lack of research on droughts on the TP and drought propagation characteristics on the TP have yet to be investigated. In this study, we investigated drought propagation time, probability, threshold, and the climatic factors associated with drought development based on the copula probability model and correlation, Bayesian network, and attribution analyses, allowing us to fill current research gaps. New hydrological insights for the regionThe research framework proposed in this study effectively reflected the characteristics of drought propagation. Our results showed that seasonality and aridity controlled the propagation time from meteorological drought (MD) to agricultural drought (AD), i.e., the time was shortest in summer (2.311.0 months) and in the humid zone (2.37 months); the same pattern was found for the propagation of MD to hydrological drought (HD), with the propagation time being 2.33.0 months in summer and 2.36 months in humid zones. The propagation probability for both AD and HD generally increased synchronously with the severity of MD, and also exhibited seasonal patterns, with the highest probability values in summer (0.590.90 for AD and 0.510.82 for HD). The highest propagation thresholds were also found in summer (−1.68 to −1.09 for AD and −1.68 to −1.36 for HD). The trends of both propagation time and probability were significant, and the probability of HD exhibited significant downward trends (−0.074 to −0.030/decade) across the TP. Precipitation was the dominant factor controlling the development of drought in most cases; however, other climatic factors, such as maximum temperature, solar radiation, and specific humidity, contributed 14.0 %43.9 % of the variances of AD and HD.

Drug-associated glaucoma: A real-world study based on the Food and Drug Administration adverse event reporting system database.

This study aims to assess the risk of drug-associated glaucoma and track its epidemiological characteristics using real-world data. Adverse event reports from the Food and Drug Administration Adverse Event Reporting System (FAERS) from January 2004 to December 2023 were analysed. Disproportionality analysis and the Bayesian Confidence Propagation Neural Network algorithm were used. The study classified drugs associated with glaucoma, assessed risk levels, and compared drug-induced times across different categories. Eight hundred and five drugs were linked to glaucoma in the FAERS database. Disproportionality analysis identified 46 drugs with significant risk, mainly adrenergic medications (clobetasol propionate, fluocinolone acetonide), antihypertensives (hydrochlorothiazide), insulin (insulin human), anticholinergics (umeclidinium, darifenacin), VEGF inhibitors (brolucizumab, faricimab), and psychotropics (topiramate, ziprasidone). The top three high-risk drugs were clobetasol propionate, umeclidinium, and fluocinolone acetonide. The shortest drug-induced times were observed with indacaterol, salmeterol, and umeclidinium. Anticholinergic medications had the shortest drug-induced time among all categories. Females (62.5%) and the elderly (average age 63.5 ± 16.8 years) were predominantly affected. Reports of drug-associated glaucoma increased over the years. Preventing drug-associated glaucoma is more effective than treatment. Identifying the risk and drug-induced times of systemic and ophthalmic drugs can reduce occurrence risk. Clinical practitioners should be vigilant and inform patients of these risks.

Safety assessment of sapropterin dihydrochloride: real-world adverse event analysis based on the FDA adverse event reporting system (FAERS).

Sapropterin dihydrochloride is the first drug for the therapy of phenylketonuria, which is a rare disease that occurs one of 10,000-15,000 newborns. As a result, detailed and comprehensive reports on the safety of sapropterin in large, real-world populations are required. The purpose of this study is to undertake a complete analysis of sapropterin's adverse events (AEs) using the FDA Adverse Event Reporting System (FAERS) database. We retrieved reports of adverse events with sapropterin as the principal suspect from FAERS between the first quarter of 2008 and the first quarter of 2024. The Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), and Bayesian Confidence Propagation Neural Network (BCPNN) were utilized to detect AE signals. The study collected 4,953 suspected AE cases from the FAERS database, with sapropterin as the major suspect. A total of 130 positive signals were obtained utilizing the ROR, PRP, and BCPNN. The FAERS database revealed that common clinical AEs of sapropterin included vomiting, upper respiratory infection, rhinorrhea, and a reduction in amino acid concentrations. Furthermore, we detected probable unexpected adverse events (AEs) using disproportionality analysis, including gastroesophageal reflux disease, flatulence, influenza, ear infection, viral infection, pharyngitis streptococcal, spontaneous abortion, and nephrolithiasis. By analyzing huge amounts of real-world data from the FAERS database, we found potential novel AEs of sapropterin using disproportionate analysis. It is advantageous for healthcare professionals and pharmacists to focus on efficiently managing sapropterin's high-risk adverse events, improving drug levels in clinical settings, and ensuring patient medication safety.

Efficient probabilistic inference in biochemical networks

Biochemical networks are usually modeled by ordinary differential equations that describe the time evolution of the concentrations of the interacting (biochemical) species for specific initial concentrations and certain values of the interaction rates. They consist, in general, of many variables and parameters. Parameter estimation of such systems using standard optimization algorithms is computationally expensive since a large number of numerical simulations must be performed for numerous values of initial conditions and parameters, making these approaches either inefficient or inaccurate. In this paper, we explain how biochemical networks can be approximated by dynamic Bayesian networks, a class of discrete probabilistic models. This type of approximation enables the use of Bayesian inference for performing tasks such as parameter estimation. We explain how to make this type of approximations and the subsequent parameter estimation task as accurate and computationally efficient as possible. Our approach is applied on concrete examples such as the EGF-NGF cellular signaling pathway.

Application of probabilistic and stochastic models and data mining for forecasting the contingent of old age pension recipients in the context of systemic uncertainty

The object of the research is mathematical models for forecasting the contingent of pension recipients in conditions of uncertainty caused by both the reform of the pension system and the impact of armed aggression. Based on the study of statistical information on the structure and dynamics of the contingent of pension recipients, an approach to uncovering systemic uncertainty in the task of forecasting the contingent of pensioners is proposed. This work is part of a study of the application of data mining methods of intellectual data analysis and mathematical modeling in information technology intended for use in the pension system. The main focus of this work is on forecasting the dynamics of the contingent of pension recipients by age, in particular, forecasting the number of newly appointed pensions. The difficulty of forecasting the contingent of pension recipients, in particular by age, is connected with the fact that it is necessary to ensure the representativeness and variability of data sets. In addition, it should be taken into account that a significant number of factors must be included in the model in accordance with the requirements of regulatory documents. Another problematic issue is that the time series of the investigated indicators, such as data on the insurance experience of insured persons (based on the results of a sample survey), may contain significant (more than 40 %) gaps that can be filled only on the basis of primary (paper) documents. Therefore, the input data sets are formed with assumptions about the probability of the accumulation of insurance experience in various groups of insured persons. The paper proposes an analytical toolkit based on the use of probabilistic and statistical models in the form of Bayesian networks, intended for use in specialized decision-making support systems of the Ukrainian pension system. In the course of the study, a number of numerical experiments were carried out, in which the correctness of the proposed method was investigated. The proposals presented in the paper will improve the stability of the pension system of Ukraine, including through a more accurate definition of the dynamics of the contingent of pension recipients and, accordingly, the costs of pension payments. The proposed models and methods can be used as part of decision-making support systems of state and public administration bodies to analyze the results of reforming the pension system.

Co-developing frameworks towards environmentally directed pharmaceutical prescribing in Scotland – A mixed methods study

The presence of human pharmaceuticals in the aquatic environment is recognised internationally as an important public health and environmental issue. In Scotland, healthcare sustainability targets call for improvements to medicine prescribing and use to reduce healthcare's impact on the environment. This proof-of-concept study aimed to develop a framework on the environmental impact of pharmaceuticals to use as a knowledge support tool for healthcare professionals, focussing on pharmaceutical pollution. Nominal Group Technique was applied to achieve consensus on pharmaceuticals and modelling factors for the framework, working with a panel of cross-sector stakeholders. Bayesian Belief Network modelling was applied to predict the environmental impact (calculated from hazard and exposure factors) of selected pharmaceuticals, with Scotland-wide mapping for visualisation in freshwater catchments. The model calculated the pollution risk score of the individual pharmaceuticals, using the ratio of prescribed mass vs. mass that would not exceed the predicted no-effect concentration in the freshwater environment. The pharmaceuticals exhibited different risk patterns, and spatial variation of risk was evident (generally related to population density), with the most catchments predicted to exceed the pollution risk score for clarithromycin (probability >80 % in 35 of 40 modelled catchments). Simulated risk scores were compared against observed risk calculated as the ratio of measured environmental concentrations from national regulatory and research monitoring and predicted no-effect concentrations. The model generally overpredicted risk, likely due to missing factors (e.g. solid-phase sorption, temporal variation), low spatial resolution, and low temporal resolution of the monitoring data. This work demonstrates a novel, trans-disciplinary approach to develop tools aiding collation and integration of environmental information into healthcare decision-making, through application of public health, environmental science, and health services research methods. Future work will refine the framework with additional clinical and environmental factors to improve model performance, and develop electronic interfaces to communicate environmental information to healthcare professionals.

Systematic review and indirect treatment comparisons of ritlecitinib against baricitinib in alopecia areata.

Ritlecitinib and baricitinib are recently approved systemic treatments for severe alopecia areata (AA). Both demonstrated superiority over placebo in hair regrowth measured by the Severity of Alopecia Tool (SALT), but they have not been directly compared in randomized controlled trials (RCTs). We conducted a systematic review of RCTs evaluating treatments in AA and estimated the efficacy and safety of ritlecitinib and baricitinib at Week 24 using Bayesian network meta-analysis. To adjust and explore effect modifiers, population-adjusted indirect comparison was performed via multilevel network meta-regression (ML-NMR) using ritlecitinib individual patient data (IPD). Co-primary endpoints were SALT ≤20 and SALT ≤10 at Week 24. Unanchored population adjusted ITCs were also computed to evaluate SALT ≤10 and SALT ≤20 endpoints at Week 48/52. Four RCTs (ALLEGRO 2a [NCT02974868], ALLEGRO 2b/3 [NCT03732807], BRAVE-AA1 [NCT03570749] and BRAVE-AA2 [NCT03899259]) were included. No evidence of a difference between ritlecitinib 50 mg and baricitinib 4 mg on SALT ≤10 (odds ratio, OR: 0.96, 95% credible interval, CrI: 0.18-7.21) and SALT ≤20 (OR: 2.16, 95% CrI: 0.48-16.46) at Week 24 was found. ML-NMR using ALLEGRO IPD adjusted for sex, SALT score at baseline, duration of current episode and disease duration found evidence of effect modification, although relative efficacy between ritlecitinib 50 mg and baricitinib 4 mg remained unchanged. Unanchored population-adjusted ITC at Week 48/52 was consistent with previous results. We found similar efficacy between ritlecitinib 50 mg and baricitinib 4 mg. These ITCs was informed by only four RCTs, uncertainty was considerable, and there was evidence of effect modification, highlighting the need for further quality research in AA.

Leveraging Network Insights into Positive Emotions and Resilience for Better Life Satisfaction

Introduction/Objective Within a network analysis framework, this study explored the relationships between positive emotions, resilience, and life satisfaction. Positive emotions and resilience are critical to enhancing psychological well-being, but their complex interplay requires further investigation, particularly in a Greek adult population. Methods We conducted a cross-sectional study with 1,230 Greek adults (67.6% females, 32.4% males), using a network analysis to assess the relationships among positive emotions, resilience, and life satisfaction. Data were collected through the Scale of Positive and Negative Experience (SPANE-8), the Brief Resilience Scale (BRS), and the Satisfaction with Life Scale (SWLS). Network structures were constructed using the graphical LASSO technique for partial correlation analysis, and Bayesian networks were applied to generate Directed Acyclic Graphs (DAGs) to identify directional pathways. Centrality metrics were used to determine critical variables in the network. Results Happiness, pleasure, and contentment are examples of positive emotions that are central nodes in the network that are highly correlated with higher levels of life satisfaction. Resilience, primarily as a recovery mechanism from stress, showed weaker direct associations with life satisfaction. The DAG revealed that positive emotions significantly influenced life satisfaction, suggesting their pivotal role in improving well-being. Conclusion The findings underscore the importance of positive emotions in enhancing life satisfaction, suggesting that interventions targeting emotional well-being may be more effective than those focused solely on resilience. These insights offer a foundation for developing psychological interventions to improve life satisfaction.

On the effectiveness of partially deterministic Bayesian neural networks

AbstractBayesian neural networks (BNNs) with computationally expensive Hamiltonian Monte Carlo sampling methods are often considered to provide better predictive performance than the maximum a posterior (MAP) solution. Here, as an alternative to sampling all parameters of a BNN (full-random), we experimentally evaluate partially deterministic BNNs that fix some part of the neural network parameters to their MAP solution. In particular, we consider various strategies for fixing half, or all parameters of a layer to the MAP-solution. Over a wide variety of regression and classification tasks, we find that partially deterministic BNNs often significantly improve predictive performance over the MAP-solution, with up to around 24% reduction in negative log-likelihood. Notably, we also find that partially deterministic BNNs that fix half of the parameters in each layer can also reduce under-fitting of full-random BNNs, resulting in up to 7% reduction in negative log-likelihood.