Balanced Metrics Research Articles

Best practice peri-extubation bundle reduces neonatal and infant extubation failure after cardiac surgery.

Extubation failure after neonatal cardiac surgery is associated with increased intensive care unit length of stay, morbidity, and mortality. We performed a quality improvement project to create and implement a peri-extubation bundle, including extubation readiness testing, spontaneous breathing trial, and high-risk criteria identification, using best practices at high-performing centers to decrease neonatal and infant extubation failure by 20% from a baseline of 15.7% to 12.6% over a 2-year period. Utilising the transparency of the Pediatric Cardiac Critical Care Consortium database, five centres were identified as high performers, having better-than-expected neonatal extubation success rates with the balancing metric of as-expected or better-than-expected mechanical ventilation duration. Structured interviews were conducted with cardiac intensive care unit physician leadership at the identified centers to determine centre-specific extubation practices. Data from those interviews underwent qualitative content analysis which was used to develop a peri-extubation bundle. The bundle was implemented at a single-centre 17-bed cardiac intensive care unit. Extubation failure, defined as reintubation within 48 hours of extubation for anything other than a procedure, ventilator days and bundle compliance was tracked. There was a 41.4% decrease in extubation failure following bundle implementation (12 failures of 76 extubations pre-implantation; 6 failures of 65 extubations post-implementation). Bundle compliance was 95.4%. There was no difference in ventilator days (p = 0.079) between groups. Implementation of a peri-extubation bundle created from best practices at high-performing centres reduced extubation failure by 41.4% in neonates and infants undergoing congenital heart surgery.

CF3-Functionalized Side Chains in Nonfullerene Acceptors Promote Electrostatic Interactions for Highly Efficient Organic Solar Cells.

The advent of next-generation nonfullerene acceptors (NFAs) has propelled major advances in organic solar cells (OSCs). Here we report an NFA design incorporating CF3-terminated side chains having varying N-(CH2)n-CF3 linker lengths (n = 1, 2, and 3) which introduce new intermolecular interactions, hence strong modulation of the photovoltaic response. We report a systematic comparison and contrast characterization of this NFA series with a comprehensive set of chemical/physical techniques versus the heavily studied third-generation NFA, Y6, revealing distinctive and beneficial properties of this new NFA series. Single-crystal diffraction analyses reveal unusual two-dimensional mesh-like crystal structures, featuring strong interactions between the side chain CF3-terminal and NFA core F substituents. These atomistic and morphological features contribute to enhanced charge mobility and significantly enhanced photovoltaic performance. We show that varying the CF3-terminated side chain linker length strongly modulates light harvesting efficiency as well as charge recombination and the photovoltaic bandgap. The CF3-(CH2)2-based OSC demonstrates the most balanced performance metrics, achieving a remarkable 19.08% power conversion efficiency and an exceptional 80.09% fill-factor. These results imply that introducing CF3-terminated side chains into other OSC conjugated constituents may accelerate next-generation solar cell development.

A Reliable and Load Balancing Controller Placement Method in Software‐Defined Networks

ABSTRACTSoftware‐defined networking (SDN) achieves the programmability of the control plane by separating it from the data forwarding plane to provide flexible management of the network resources. The multicontroller architecture is required to be deployed to enhance the scalability and reliability of the control plane with the network traffic growth. However, the controller placement problem (CPP) is considered an important challenge in software‐defined networking, which should be addressed. The number of required controllers and their locations are the important challenges that affect various aspects of the separated controller plane such as the performance metrics, and ability to respond to failures. Also, unappropriated subdomain partitioning of the software‐defined network by multicontrollers may cause the unbalanced distribution of controller loads resulting in the reduction of communication performance of the network. In this paper, an optimization subdomain partitioning method based on the particle swarm optimization (PSO) algorithm is presented for deploying the CPP and allocating switches to controllers. The proposed control placement method aims to minimize the cost of the network known as the number of required controllers, to minimize the maximum load imbalance between controllers, and to improve resilience against a failure between each switch and its mapping controller. The presented method is evaluated using two widely used networks from the Internet Topology Zoo such as Aarnet, Oxford, Chinanet, Interoute, and ION topologies to show the scalability of the proposed method. The results show that the proposed method achieves better performance in the required number of controllers, propagation delay, and load balancing among controllers when compared to the controller placement methods based on the Varna, clustering‐based network partition algorithm (CNPA), and K‐means. Moreover, the proposed method improves load balancing when compared to the controller placement methods based on the Varna, CNPA, and K‐means, respectively. The proposed controller placement based on the PSO outperforms nearly 20% and 17% decline in the number of required controllers in comparison with the Varna‐based heuristic controller placement method and the CNPA for different scales of topologies, respectively. Moreover, the proposed controller placement method based on the particle swarm optimization enhances the load balancing metric by nearly 6% compared to the Varna‐based controller placement method in the case of load balancing scenario in the Interoute and ION topologies, which shows the improvement of the proposed method based on the PSO compared to the Varna‐based method. Also, in the proposed controller placement method based on the PSO, the load balancing scenario outperforms the load balancing metric among the assigned controllers by nearly 14%, 22%, 13%, and 18% compared to the K‐means‐based method in the ION, Interoute, Chinanet, Oxford, and Aarnet topologies. Furthermore, the proposed method achieves nearly 9%, 5%, and 15% decline in the average propagation delay compared to the Varna‐, CNPA‐, and K‐means‐based controller placement methods for different topologies. Furthermore, the proposed scheme achieves a higher resilience against controller failures compared to the existing approaches.

Residue-Level Multiview Deep Learning for ATP Binding Site Prediction and Applications in Kinase Inhibitors.

Accurate identification of adenosine triphosphate (ATP) binding sites is crucial for understanding cellular functions and advancing drug discovery, particularly in targeting kinases for cancer treatment. Existing methods face significant challenges due to their reliance on time-consuming precomputed features and the heavily imbalanced nature of binding site data without further investigations on their utility in drug discovery. To address these limitations, we introduced Multiview-ATPBind and ResiBoost. Multiview-ATPBind is an end-to-end deep learning model that integrates one-dimensional (1D) sequence and three-dimensional (3D) structural information for rapid and precise residue-level pocket-ligand interaction predictions. Additionally, ResiBoost is a novel residue-level boosting algorithm designed to mitigate data imbalance by enhancing the prediction of rare positive binding residues. Our approach outperforms state-of-the-art models on benchmark data sets, showing significant improvements in balanced metrics with both experimental and AI-predicted structures. Furthermore, our model seamlessly transfers to predicting binding sites and enhancing docking simulations for kinase inhibitors, including imatinib and dasatinib, underscoring the potential of our method in drug discovery applications.

Dynamics of the Korteweg–de Vries Equation on a Balanced Metric Graph

Dynamics of the Korteweg–de Vries Equation on a Balanced Metric Graph

Comparative Analysis of Machine Learning Algorithms for Cross-Site Scripting (XSS) Attack Detection

Cross-Site Scripting (XSS) attacks pose a significant cybersecurity threat by exploiting vulnerabilities in web applications to inject malicious scripts, enabling unauthorized access and execution of malicious code. Traditional XSS detection systems often struggle to identify increasingly complex XSS payloads. To address this issue, this research evaluated the efficacy of Machine Learning algorithms in detecting XSS threats within online web applications. The study conducts a comprehensive comparative analysis of XSS attack detection using four prominent Machine Learning algorithms, which consist of Extreme Gradient Boosting (XGBoost), Random Forest (RF), K-Nearest Neighbors (KNN), and Support Vector Machine (SVM). This research utilizes a comparative methodology to assess the selected Machine Learning algorithms by analyzing their performance metrics, including confusion matrix, 10-fold cross-validation, and assessment of training time to thoroughly evaluate the models. By exploring dataset characteristics and evaluating the performance metrics of each selected algorithm, the study determined the most robust Machine Learning solution for XSS detection. Results indicate that Random Forest is the top performer, achieving 99.93% accuracy and balanced metrics across all criteria evaluated. These findings will significantly enhance web application security by providing reliable defenses against evolving XSS threats.

Limits of speech in connected homes: Experimental comparison of self-reporting tools for human activity recognition

Data annotation for human activity recognition is a well-known challenge for researchers. In particular, annotation in daily life settings relies on self-reporting tools with unknown accuracy. Speech is a promising interface for activity labeling. In this work, we compare the accuracy of two commercially available tools for annotation: voice diaries and connected buttons. We retrofit the water meters of thirty homes in the USA for infrastructure-mediated sensing. Participants are split into equal groups and receive one of the self-reporting tools. The balanced accuracy metric is transferred from the field of machine learning to the evaluation of the annotation performance. Our results show that connected buttons perform significantly better than the voice diary, with 92% median accuracy and 65% median reporting rate. Using questionnaire answers, we highlight that annotation performance is impacted by habit formation and sentiments toward the annotation tool. The use case for data annotation is to disaggregate water meter data into human activities beyond the point of use. We show that it is feasible with a machine-learning model and the corrected annotations. Finally, we formulate recommendations for the design of studies and intelligent environments around the key ideas of proportionality and immediacy.

Abstract 4128201: Reducing Blood Product Waste in a Pediatric Cardiac Catheterization Laboratory – A Quality Improvement Initiative

Background/Aim: Transfusion can be a life-saving intervention, however routine pre-procedural blood products frequently go unused. Recent national blood product shortages highlight the importance of blood product stewardship. In our pediatric cardiac catheterization laboratory (PCCL), packed red blood cells (pRBCs) were ordered prior to all cases. These units are rarely transfused and sometimes cannot be reallocated, resulting in waste. As a quality improvement project, we set a primary aim to reduce the percentage of monthly pRBC waste from a baseline mean of 7.4% to <5% over 12 months. Methods: Baseline data on all cardiac catheterization cases, excluding electrophysiology, lymphatic, and hybrid cases, were collected from 1/1/2022 to 2/28/2023. Interventions (Table 1) were implemented and outcomes measured from 3/1/2023 to 3/31/2024. Primary outcome measures were monthly percentage of blood wasted, and percent of unused pRBC orders. Emergency blood product activation was measured as a balancing metric. Statistical process control charts were used to display and analyze data. Results: During the intervention period 1,318 cases were performed of which 32% met low-risk criteria for pRBC order de-implementation. The percent of blood that was subsequently wasted was reduced from 7.4% to 5.5% (Figure 1A), primarily driven by the reduction in the anecdotal practice of pRBC unit splitting in patients <1 year old (Figure 1B). The percent of unused pRBC orders was reduced from 97% to 79% with sustained improvement following two cycles of routine blood order de-implementation for low-risk cases. No emergency blood activation occurred (0%, 97.5% CI: 0-0.28%). Conclusion: De-implementation of default pRBC orders for low-risk catheterization procedures and reduction in unit splitting reduced the percent of blood wasted in our PCCL, saving time and cost without compromising patient safety. A data-driven approach to blood stewardship may be broadly applicable across procedural areas.

Impact of nutrition-related laboratory tests on mortality of patients who are critically ill using artificial intelligence: A focus on trace elements, vitamins, and cholesterol.

This study aimed to understand the collective impact of trace elements, vitamins, cholesterol, and prealbumin on patient outcomes in the intensive care unit (ICU) using an advanced artificial intelligence (AI) model for mortality prediction. Data from ICU patients (December 2016 to December 2021), including serum levels of trace elements, vitamins, cholesterol, and prealbumin, were retrospectively analyzed using AI models. Models employed included category boosting (CatBoost), extreme gradient boosting (XGBoost), light gradient boosting machine (LGBM), and multilayer perceptron (MLP). Performance was evaluated using area under the receiver operating characteristic curve (AUROC), accuracy, precision, recall, and F1-score. The performance was evaluated using 10-fold crossvalidation. The SHapley Additive exPlanations (SHAP) method provided interpretability. CatBoost emerged as the top-performing individual AI model with an AUROC of 0.756, closely followed by LGBM, MLP, and XGBoost. Furthermore, the ensemble model combining these four models achieved the highest AUROC of 0.776 and more balanced metrics, outperforming all models. SHAP analysis indicated significant influences of prealbumin, Acute Physiology and Chronic Health Evaluation II score, and age on predictions. Notably, the ratios of selenium to age and low-density lipoprotein to total cholesterol also had a notable impact on the models' output. The study underscores the critical role of nutrition-related parameters in ICU patient outcomes. Advanced AI models, particularly in an ensemble approach, demonstrated improved predictive accuracy. SHAP analysis offered insights into specific factors influencing patient survival, highlighting the need for broader consideration of these biomarkers in critical care management.

G4 & the balanced metric family – a novel approach to solving binary classification problems in medical device validation & verification studies

BackgroundIn medical device validation and verification studies, the area under the receiver operating characteristic curve (AUROC) is often used as a primary endpoint despite multiple reports showing its limitations. Hence, researchers are encouraged to consider alternative metrics as primary endpoints. A new metric called G4 is presented, which is the geometric mean of sensitivity, specificity, the positive predictive value, and the negative predictive value. G4 is part of a balanced metric family which includes the Unified Performance Measure (also known as P4) and the Matthews’ Correlation Coefficient (MCC). The purpose of this manuscript is to unveil the benefits of using G4 together with the balanced metric family when analyzing the overall performance of binary classifiers.ResultsSimulated datasets encompassing different prevalence rates of the minority class were analyzed under a multi-reader-multi-case study design. In addition, data from an independently published study that tested the performance of a unique ultrasound artificial intelligence algorithm in the context of breast cancer detection was also considered. Within each dataset, AUROC was reported alongside the balanced metric family for comparison. When the dataset prevalence and bias of the minority class approached 50%, all three balanced metrics provided equivalent interpretations of an AI’s performance. As the prevalence rate increased / decreased and the data became more imbalanced, AUROC tended to overvalue / undervalue the true classifier performance, while the balanced metric family was resistant to such imbalance. Under certain circumstances where data imbalance was strong (minority-class prevalence < 10%), MCC was preferred for standalone assessments while P4 provided a stronger effect size when evaluating between-groups analyses. G4 acted as a middle ground for maximizing both standalone assessments and between-groups analyses.ConclusionsUse of AUROC as the primary endpoint in binary classification problems provides misleading results as the dataset becomes more imbalanced. This is explicitly noticed when incorporating AUROC in medical device validation and verification studies. G4, P4, and MCC do not share this limitation and paint a more complete picture of a medical device’s performance in a clinical setting. Therefore, researchers are encouraged to explore the balanced metric family when evaluating binary classification problems.

A Kummer construction for Chern–Ricci flat balanced manifolds

Given a non-Kähler Calabi–Yau compact orbifold with isolated singularities endowed with a Chern–Ricci flat balanced metric, we study, via a gluing construction, the existence of Chern–Ricci flat balanced metrics on its crepant resolutions, and discuss applications to the search of solutions for the Hull–Strominger system. We also describe the scenario of singular threefolds with ordinary double points, and see that similarly is possible to obtain balanced approximately Chern–Ricci flat metrics.

Watershed-scale spatial prediction of agricultural land phosphorus mass balance and soil phosphorus metrics: A bottom-up approach.

Analysis of nutrient balance at the watershed scale, including for phosphorus (P), is typically accomplished using aggregate input datasets, resulting in an inability to capture the variability of P status across the study region. This study presents a set of methods to predict and visualize partial P mass balance, soil P saturation ratio (PSR), and soil test P for agricultural parcels across a watershed in the Lake Champlain Basin (Vermont, USA) using granular, field-level data. K-means cluster analyses were used to group agricultural parcels by soil texture, average slope, and crop type. Using a set of parcels accounting for ∼21% of the watershed's agricultural land and having known soil test and nutrient management parameters, predictions of partial P mass balance, PSR, and soil test P for agricultural land across the watershed were made by cluster, incorporating uncertainty. This resulted in an average partial P balance of 5.5±0.2kg P ha-1year-1 and an average PSR of 0.0399±0.0002. Furthermore, approximately 30% of agricultural land had predicted soil test P values above optimum levels. Results were used to visualize areas with high P loss potential. Such data and visualizations can inform watershed P modeling and assist practitioners in nutrient management decision making. These techniques can also serve as a framework for bottom-up modeling of nutrient mass balance and soil metrics in other regions.

Prognostic score-based model averaging approach for propensity score estimation

BackgroundPropensity scores (PS) are typically evaluated using balance metrics that focus on covariate balance, often without considering their predictive power for the outcome. This approach may not always result in optimal bias reduction in the treatment effect estimate. To address this issue, evaluating covariate balance through prognostic scores, which account for the relationship between covariates and the outcome, has been proposed. Similarly, using a typical model averaging approach for PS estimation that minimizes prediction error for treatment status and covariate imbalance does not necessarily optimize PS-based confounding adjustment. As an alternative approach, using the averaged PS model that minimizes inter-group differences in the prognostic score may further reduce bias in the treatment effect estimate. Moreover, since the prognostic score is also an estimated quantity, model averaging in the prognostic scores can help identify a better prognostic score model. Utilizing the model-averaged prognostic scores as the balance metric for constructing the averaged PS model can contribute to further decreasing bias in treatment effect estimates. This paper demonstrates the effectiveness of the PS model averaging approach based on prognostic score balance and proposes a method that uses the model-averaged prognostic score as a balance metric, evaluating its performance through simulations and empirical analysis.MethodsWe conduct a series of simulations alongside an analysis of empirical observational data to compare the performances of weighted treatment effect estimates using the proposed and existing approaches. In our examination, we separately provid four candidate estimates for the PS and prognostic score models using traditional regression and machine learning methods. The model averaging of PS based on these candidate estimators is performed to either maximize the prediction accuracy of the treatment or to minimize intergroup differences in covariate distributions or prognostic scores. We also utilize not only the prognostic scores from each candidate model but also an averaged score that best predicted the outcome, for the balance assessment.ResultsThe simulation and empirical data analysis reveal that our proposed model-averaging approaches for PS estimation consistently yield lower bias and less variability in treatment effect estimates across various scenarios compared to existing methods. Specifically, using the optimally averaged prognostic scores as a balance metric significantly improves the robustness of the weighted treatment effect estimates.DiscussionThe prognostic score-based model averaging approach for estimating PS can outperform existing model averaging methods. In particular, the estimator using the model averaging prognostic score as a balance metric can produce more robust estimates. Since our results are obtained under relatively simple conditions, applying them to real data analysis requires adjustments to obtain accurate estimates according to the complexity and dimensionality of the data.ConclusionsUsing the prognostic score as the balance metric for the PS model averaging enhances the performance of the treatment effect estimator, which can be recommended for a wide variety of situations. When applying the proposed method to real-world data, it is important to use it in conjunction with techniques that mitigate issues arising from the complexity and high dimensionality of the data.

THE TRANSFORMATIVE IMPACT OF JUDO ON ATHLETES WITH DISABILITIES: INCLUSION, ADAPTATION, AND PERFORMANCE INSIGHTS

Judo is celebrated for its inclusivity and adaptability, offering significant benefits for athletes with disabilities (PwD). It fosters physical, emotional, social, and cognitive growth, enhancing motor skills, coordination, and balance. For athletes with physical disabilities, mastering techniques such as falling safely promotes strength and agility. Moreover, engaging in judo boosts self-esteem and confidence as athletes achieve milestones like advancing through belt ranks and earning medals. The supportive environment encourages personal accomplishment and helps develop social skills through interaction with peers. Research highlights the diverse range of disabilities in judo, including visual, auditory, and intellectual impairments. Studies by Oblak et al. (2020) reveal the therapeutic benefits of judo for individuals with intellectual disabilities, emphasizing the need for more research in this area. Bocioacă and Marin (2023) explore the role of adapted judo for individuals with Autism Spectrum Disorder (ASD) and Down syndrome, demonstrating improvements in integration skills through tailored practices. Additional research by Kons et al. (2021) and Krabben et al. (2017) addresses the dynamics of Paralympic judo, revealing challenges related to classification systems for visually impaired athletes. Kons et al. (2022) identify a concerning rise in injury rates among judo athletes with disabilities, indicating the need for improved safety measures. Lastly, the comparative study of neuromuscular control by Kons et al. (2019) shows that visually impaired judokas can excel in balance and certain strength metrics. Overall, these studies underscore judo’s role in promoting physical development and social integration for PwD, while also highlighting the need for further research to ensure equitable and inclusive practices in the sport.

Energy‐Water Asynchrony Principally Determines Water Available for Runoff From Snowmelt in Continental Montane Forests

ABSTRACTChanges in the volume, rate, and timing of the snowmelt water pulse have profound implications for seasonal soil moisture, evapotranspiration (ET), groundwater recharge, and downstream water availability, especially in the context of climate change. Here, we present an empirical analysis of water available for runoff using five eddy covariance towers located in continental montane forests across a regional gradient of snow depth, precipitation seasonality, and aridity. We specifically investigated how energy‐water asynchrony (i.e., snowmelt timing relative to atmospheric demand), surface water input intensity (rain and snowmelt), and observed winter ET (winter AET) impact multiple water balance metrics that determine water available for runoff (WAfR). Overall, we found that WAfR had the strongest relationship with energy‐water asynchrony (adjusted r2 = 0.52) and that winter AET was correlated to total water year evapotranspiration but not to other water balance metrics. Stepwise regression analysis demonstrated that none of the tested mechanisms were strongly related to the Budyko‐type runoff anomaly (highest adjusted r2 = 0.21). We, therefore, conclude that WAfR from continental montane forests is most sensitive to the degree of energy‐water asynchrony that occurs. The results of this empirical study identify the physical mechanisms driving variability of WAfR in continental montane forests and are thus broadly relevant to the hydrologic management and modelling communities.

Artificial intelligent-driven decision-making for automating root fracture detection in periapical radiographs

BackgroundThe detection and early diagnosis of root fractures can be challenging; this difficulty applies particularly to newly qualified dentists. Aside from clinical examination, diagnosis often requires radiographic assessment. Nonetheless, human fallibility can introduce errors due to a lack of experience.AimThe proposed system aimed to assist in detecting root fractures through the integration of artificial intelligence techniques into the diagnosis process as a step for automating dental diagnosis and decision-making processes.Materials and methodA total of 400 radiographic images of fractured and unfractured teeth were obtained for the present research. Data handling techniques were implemented to balance the distribution of the samples. The AI-based system used the voting technique for five different pretrained models namely, VGG16, VGG19, ResNet50. DenseNet121, and DenseNet169 to perform the analysis. The parameters used for the analysis of the models are loss and accuracy curves.ResultsVGG16 exhibited notable success with low training and validation losses (0.09% and 0.18%, respectively), high specificity, sensitivity, and positive predictive value (PPV). VGG19 showed potential overfitting concerns, while ResNet50 displayed progress in minimizing loss but exhibited bias toward unfractured cases. DenseNet121 effectively addressed overfitting and noise issues, achieving balanced metrics and impressive PPVs for both fractured and unfractured cases (0.933 and 0.898 respectively). With increased depth, DenseNet169 demonstrated enhanced generalization capability.ConclusionThe proposed AI- based system demonstrated high precision and sensitivity for detecting root fractures in endodontically treated teeth by utilizing the voting method.

Frölicher spectral sequence of compact complex manifolds with special Hermitian metrics

In this paper we focus on the interplay between the behaviour of the Frölicher spectral sequence and the existence of special Hermitian metrics on the manifold, such as balanced, SKT or generalized Gauduchon. The study of balanced metrics on nilmanifolds endowed with strongly non-nilpotent complex structures allows us to provide infinite families of compact balanced manifolds with Frölicher spectral sequence not degenerating at the second page. Moreover, this result is extended to non-degeneration at any arbitrary page. Similar results are obtained for the Frölicher spectral sequence of compact generalized Gauduchon manifolds. We also find a compact SKT manifold whose Frölicher spectral sequence does not degenerate at the second page, thus providing a counterexample to a conjecture by Popovici.

Utilizing Cobit 4.1 and Balance Scorecard to Manage Information Technology Business Process in Higher Education Institution

Pamulang University's Information Technology Strategy Plan aims to develop a comprehensive strategy to maximize IT efficiency and effectiveness in the institution. Using the COBIT 4.1 approach, SWOT analysis, and Balanced Scorecard, this plan identifies strengths, weaknesses, opportunities, and threats in the current IT infrastructure and develops an integrated strategy for IT development. COBIT 4.1 provides a framework for managing and controlling IT by ensuring that IT processes support business objectives and meet stakeholder needs. The SWOT analysis explores the internal and external factors that influence IT at Pamulang University, while the Balanced Scorecard helps in establishing balanced strategic metrics and objectives across multiple perspectives, including financial, customer, internal processes, and learning and growth. This strategic plan is expected to improve IT management, support the achievement of Pamulang University's goals and facilitate data-based decisions for innovation and continuous improvement.

The Impact of Global Oil Market Shocks on India’s Merchandise Trade Balances

This paper examines the influence of global oil market shocks on India’s aggregate, non-oil, and oil trade balances using the structural vector autoregression model. The impulse response analysis suggests that aggregate and non-oil trade balances deteriorate in response to adverse oil-specific demand shocks, while the oil trade balance improves. Among oil market shocks, only oil-specific demand shocks have a significant impact on all trade balance metrics. This research has crucial policy implications for the Indian economy.

Evaluation and comparison of covariate balance metrics in studies with time-dependent confounding.

Marginal structural models have been increasingly used by analysts in recent years to account for confounding bias in studies with time-varying treatments. The parameters of these models are often estimated using inverse probability of treatment weighting. To ensure that the estimated weights adequately control confounding, it is possible to check for residual imbalance between treatment groups in the weighted data. Several balance metrics have been developed and compared in the cross-sectional case but have not yet been evaluated and compared in longitudinal studies with time-varying treatment. We have first extended the definition of several balance metrics to the case of a time-varying treatment, with or without censoring. We then compared the performance of these balance metrics in a simulation study by assessing the strength of the association between their estimated level of imbalance and bias. We found that the Mahalanobis balance performed best. Finally, the method was illustrated for estimating the cumulative effect of statins exposure over one year on the risk of cardiovascular disease or death in people aged 65 and over in population-wide administrative data. This illustration confirms the feasibility of employing our proposed metrics in large databases with multiple time-points.