Trade-off Preferences Research Articles

886: Enabling Tradeoff Preference in Deep Learning (DL) based HN IMRT Auto Planning

886: Enabling Tradeoff Preference in Deep Learning (DL) based HN IMRT Auto Planning

Patient Risk-Benefit Preferences for Transcatheter Versus Surgical Mitral Valve Repair.

Transcatheter edge-to-edge repair (TEER) of mitral regurgitation is less invasive than surgery but has greater 5-year mortality and reintervention risks, and leads to smaller improvements in physical functioning. The study objective was to quantify patient preferences for risk-benefit trade-offs associated with TEER and surgery. A discrete choice experiment survey was administered to patients with mitral regurgitation. Attributes included procedure type; 30-day mortality risk; 5-year mortality risk and physical functioning for 5 years; number of hospitalizations in the next 5 years; and risk of additional surgery in the next 5 years. A mixed-logit regression model was fit to estimate preference weights. Two hundred one individuals completed the survey: 63% were female and mean age was 74 years. On average, respondents preferred TEER over surgery. To undergo a less invasive procedure (ie, TEER), respondents would accept up to a 13.3% (95% CI, 8.7%-18.5%) increase in reintervention risk above a baseline of 10%, 4.6 (95% CI, 3.1-6.2) more hospitalizations above a baseline of 1, a 10.7% (95% CI, 6.5%-14.5%) increase in 5-year mortality risk above a baseline of 20%, or more limited physical functioning representing nearly 1 New York Heart Association class (0.7 [95% CI, 0.4-1.1]) over 5 years. Patients in general preferred TEER over surgery. When holding constant all other factors, a functional improvement from New York Heart Association class III to class I maintained over 5 years would be needed, on average, for patients to prefer surgery over TEER.

Treatment Outcome Priorities of Older Adults with Newly Diagnosed Blood Cancers Elicited with Best-Worst Scaling

Background: A clear understanding of patients' preferred treatment outcomes is necessary to guide treatment decision-making. In this study, we sought to quantify the treatment outcome priorities of older adults (≥ 60 years old) newly diagnosed with hematologic malignancies and to describe individual patient's willingness to tolerate specific treatment-related benefit-risk tradeoffs. Methods: We developed a values elicitation measure using best-worst (BW) scaling and direct elicitation through a community-centered approach engaging diverse stakeholders. The measure includes 7 treatment priorities: Living longer, Maintaining day-to-day activities, Avoiding becoming more dependent on others, Avoiding hospitalizations/Increasing time at home, Avoiding long-term side effects, Avoiding short-term side effects, and Avoiding high financial costs. Patients complete repeated choice tasks choosing the most and least important treatment priority among repeated subsets using a balanced-incomplete block design. Direct elicitation was used to determine specific benefit-risk tradeoff preferences. Recruitment occurred at a public safety-net hospital following pathologic diagnosis. Patients completed the measure prior to the initial treatment decision and longitudinally each month. Data from the BW scaling measure were analyzed using the count method (best minus worst) and standardized. Results: In total, 48 patients consented (48% lymphoma, 40% leukemia/myelodysplastic syndrome, 12% other). Mean age was 74 years (range 60 - 95); 59% were male; 96% were white; 67% had a bachelor's degree or higher. Fifty-two percent (n=25) fully completed the BWS instrument at least once. The ranked importance of priorities for the entire cohort was 1) Maintaining day-to-day activities (standardized best minus worst, 23.8), 2) Living longer (12.4), 3) Avoiding becoming more dependent on others (7.9), 4) Avoiding hospitalizations/Increasing time at home (2.2), 5) Avoiding long-term side effects (-5.2), 6) Avoiding short-term side effects (-19.5), and 7) Avoiding high financial costs (-21.5). The most important outcomes (including ties) for individual patients were Maintaining day-to-day activities (40% of patients), Living longer (40%), and Avoiding becoming dependent on others (20%), while the least important outcomes were Avoiding high financial costs (52%), Avoiding short term side-effects (32%), and Living longer (20%). Thirteen (52%) patients completed subsequent BWS surveys; 69% changed their most important outcome at least once. Direct elicitation demonstrated that patients were willing to accept additional hospital stays of 1 week (88%), 1 month (58%), and 3 months (25%) in exchange for an additional 6 months of life. Only 25% of patients indicated willingness to require a full-time caregiver in exchange for an additional 2 years of life, whereas patients were more willing to accept a part-time caregiver in exchange for an additional 1 month (38%) or 6 months (79%) of life. Some patients were unwilling to accept an additional 6 months of life if it meant requiring assistance dressing (33%) or bathing (54%). In exchange for an additional 6 months of life, patients were willing to spend $250 (83%), $500 (58%), or $1000 (42%) per month. Conclusions: Older adults with newly diagnosed blood cancers differed when prioritizing treatment outcomes. Maintaining function was highly valued and was frequently more important than prolonging survival. Some patients felt prolonging survival was least important. Many patients changed their treatment priorities over time. These results highlight the need to elicit treatment outcome priorities at the time of each significant treatment decision for older adults. Further, they suggest that endpoints other than overall survival should be considered for clinical trials in this population as prolonging survival may not be prioritized by most patients. This study also demonstrated that a combination of BW scaling and direct elicitation was able to quantify patients' treatment outcome priorities and benefit-risk tradeoff preferences, highlighting the potential for the development of similar values elicitation measures to be used to inform specific treatment decisions in various clinical contexts.

Correction to: Sensitivity to scope in estimating the social benefits of prolonging lives for regulatory decisions using national stated preference tradeoffs

Correction to: Sensitivity to scope in estimating the social benefits of prolonging lives for regulatory decisions using national stated preference tradeoffs

Improving Pareto Front Learning via Multi-Sample Hypernetworks

Pareto Front Learning (PFL) was recently introduced as an effective approach to obtain a mapping function from a given trade-off vector to a solution on the Pareto front, which solves the multi-objective optimization (MOO) problem. Due to the inherent trade-off between conflicting objectives, PFL offers a flexible approach in many scenarios in which the decision makers can not specify the preference of one Pareto solution over another, and must switch between them depending on the situation. However, existing PFL methods ignore the relationship between the solutions during the optimization process, which hinders the quality of the obtained front. To overcome this issue, we propose a novel PFL framework namely PHN-HVI, which employs a hypernetwork to generate multiple solutions from a set of diverse trade-off preferences and enhance the quality of the Pareto front by maximizing the Hypervolume indicator defined by these solutions. The experimental results on several MOO machine learning tasks show that the proposed framework significantly outperforms the baselines in producing the trade-off Pareto front.

Social bias, not time bias

People seem to have pure time preferences about trade-offs concerning their own pleasures and pains, and such preferences contribute to estimates of people's individual time discount rate. Do pure time preferences also matter to interpersonal welfare trade-offs, including those concerning the welfare of future generations? Most importantly, should the intergenerational time discount rate include a pure time preference? Descriptivists claim that the intergenerational discount rate should reflect actual people's revealed preferences, and thus it should include a pure time preference. Prescriptivists claim that the intergenerational discount rate should be based on moral analysis, and thus they (often) claim that the rate of pure time preference should be zero. I argue that regardless of which view is correct, a focus on pure time preference is misplaced. First, the most plausible interpretation of descriptive preferences for intergenerational trade-offs is that people are socially biased and not time biased. Second, social bias is superior to time bias as a prescriptive reason to discount the welfare of future people. Third, recent advances in measuring social bias as a social discount rate make social bias a viable replacement for time bias in economic analyses of intergenerational welfare trade-offs.

Designing a transactive electric vehicle agent with customer’s participation preference

The proliferation of electric vehicles (EVs) and their inherent flexibility in charging timings make them an asset to improve grid performance. In contrast to direct control by a utility or autonomous price-based charging, the transactive control framework not only provides benefits to both grid and customers but also ensures customer autonomy. In this work, we design a transactive electric vehicle (TEV) agent that incorporates the EV owner’s willingness to trade-off between savings and amenity in form of a slider, where the EV owner’s amenity is characterized as vehicle readiness. Further, a privacy-preserving bidding formulation is proposed that also represents the customer’s transactive preference. A transactive market mechanism is discussed that integrates the TEV Agents into the local retail market and reconciles with the current day-ahead and real-time market structure. It is demonstrated that the proposed slider is able to provide a preferred trade-off between savings and amenity to individual customers. At the same time, the market mechanism is shown to successfully reduce both peak prices and peak demand. A comparative investigation of V1G and V2G technologies with respect to the battery prices is also discussed. It reveals that the V2G does not offer significant additional benefits with current battery prices, but could be promising if battery costs decline in the future.

Predicting the causative pathogen among children with pneumonia using a causal Bayesian network.

Pneumonia remains a leading cause of hospitalization and death among young children worldwide, and the diagnostic challenge of differentiating bacterial from non-bacterial pneumonia is the main driver of antibiotic use for treating pneumonia in children. Causal Bayesian networks (BNs) serve as powerful tools for this problem as they provide clear maps of probabilistic relationships between variables and produce results in an explainable way by incorporating both domain expert knowledge and numerical data. We used domain expert knowledge and data in combination and iteratively, to construct, parameterise and validate a causal BN to predict causative pathogens for childhood pneumonia. Expert knowledge elicitation occurred through a series of group workshops, surveys and one-on-one meetings involving 6-8 experts from diverse domain areas. The model performance was evaluated based on both quantitative metrics and qualitative expert validation. Sensitivity analyses were conducted to investigate how the target output is influenced by varying key assumptions of particular high degree of uncertainty around data or domain expert knowledge. Designed to apply to a cohort of children with X-ray confirmed pneumonia who presented to a tertiary paediatric hospital in Australia, the resulting BN offers explainable and quantitative predictions on a range of variables of interest, including the diagnosis of bacterial pneumonia, detection of respiratory pathogens in the nasopharynx, and the clinical phenotype of a pneumonia episode. Satisfactory numeric performance has been achieved including an area under the receiver-operating characteristics curve of 0.8 in predicting clinically-confirmed bacterial pneumonia with sensitivity 88% and specificity 66% given certain input scenarios (i.e., information that is available and entered into the model) and trade-off preferences (i.e., relative weightings of the consequences of false positive versus false negative predictions). We specifically highlight that a desirable model output threshold for practical use is very dependent upon different input scenarios and trade-off preferences. Three commonly encountered scenarios were presented to demonstrate the potential usefulness of the BN outputs in various clinical pictures. To our knowledge, this is the first causal model developed to help determine the causative pathogen for paediatric pneumonia. We have shown how the method works and how it would help decision making on the use of antibiotics, providing insight into how computational model predictions may be translated to actionable decisions in practice. We discussed key next steps including external validation, adaptation and implementation. Our model framework and the methodological approach can be adapted beyond our context to broad respiratory infections and geographical and healthcare settings.

Sensitivity to scope in estimating the social benefits of prolonging lives for regulatory decisions using national stated preference tradeoffs

Sensitivity to scope in estimating the social benefits of prolonging lives for regulatory decisions using national stated preference tradeoffs

Patient Preferences for Treatment of Bacillus Calmette-Guérin–unresponsive Non–muscle-invasive Bladder Cancer: A Cross-country Choice Experiment

BackgroundPatients with non–muscle-invasive bladder cancer (NMIBC) that is unresponsive to bacillus Calmette-Guérin (BCG) immunotherapy face a difficult choice. Immediate radical cystectomy (RC) is effective but might represent overtreatment. Continuing bladder preservation with medical therapy is an alternative, but it risks progression to muscle-invasive bladder cancer (MIBC) and a reduction in survival. ObjectiveTo understand the trade-offs patients are willing to make in selecting treatments for BCG-unresponsive NMIBC. Design, setting, and participantsAdults with NMIBC from the UK, France, Germany, and Canada who reported current receipt of BCG, disease unresponsive to BCG, or receipt of RC in the previous 12 mo after failure of BCG were recruited to participate in an online choice experiment. Patients were asked to make repeated choices between two hypothetical medical treatments and the option to undergo immediate RC. The medical treatments required trade-offs between the time to RC, the mode and frequency of administration, the risk of experiencing serious side effects, and the risk of disease progression. Outcome measurements and statistical analysisError component logit models were used to calculate relative attribute importance (RAI) scores as the maximum percentage contribution to a preference and acceptable benefit-risk trade-offs. Results and limitationsMost of the 107 participants (average age 63 yr) never selected RC (89%) as their preferred option in the choice experiment. Preferences were most affected by time to RC (RAI 55%), followed by risk of progressing to MIBC (RAI 25%), medication administration (RAI 12%), and the risk of serious side effects (RAI 8%). To increase the time to RC from 1 yr to 6 yr, patients accepted a 43.8% increase in the risk of progression and a 66.1% increase in the risk of serious side effects. ConclusionsPatients with BCG-treated NMIBC valued bladder-sparing treatments and were willing to make substantial benefit-risk trade-offs to delay RC. Patient summaryAdults with bladder cancer not invading the bladder muscle completed an online experiment in which they chose between hypothetical medications and bladder removal. The results show that patients would be willing to accept different risks associated with medications to delay bladder removal. Patients considered disease progression the most important risk of medicinal treatment.

DilateFormer: Multi-Scale Dilated Transformer for Visual Recognition

As a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">de facto</i> solution, the vanilla Vision Transformers (ViTs) are encouraged to model long-range dependencies between arbitrary image patches while the global attended receptive field leads to quadratic computational cost. Another branch of Vision Transformers exploits local attention inspired by CNNs, which only models the interactions between patches in small neighborhoods. Although such a solution reduces the computational cost, it naturally suffers from small attended receptive fields, which may limit the performance. In this work, we explore effective Vision Transformers to pursue a preferable trade-off between the computational complexity and size of the attended receptive field. By analyzing the patch interaction of global attention in ViTs, we observe two key properties in the shallow layers, namely locality and sparsity, indicating the redundancy of global dependency modeling in shallow layers of ViTs. Accordingly, we propose Multi-Scale Dilated Attention (MSDA) to model <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">local</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">sparse</i> patch interaction within the sliding window. With a pyramid architecture, we construct a Multi-Scale Dilated Transformer (DilateFormer) by stacking MSDA blocks at low-level stages and global multi-head self-attention blocks at high-level stages. Our experiment results show that our DilateFormer achieves state-of-the-art performance on various vision tasks. On ImageNet-1 K classification task, DilateFormer achieves comparable performance with 70% fewer FLOPs compared with existing state-of-the-art models. Our DilateFormer-Base achieves 85.6% top-1 accuracy on ImageNet-1 K classification task, 53.5% box mAP/46.1% mask mAP on COCO object detection/instance segmentation task and 51.1% MS mIoU on ADE20 K semantic segmentation task.

Factorization Vision Transformer: Modeling Long-Range Dependency With Local Window Cost.

Transformers have astounding representational power but typically consume considerable computation which is quadratic with image resolution. The prevailing Swin transformer reduces computational costs through a local window strategy. However, this strategy inevitably causes two drawbacks: 1) the local window-based self-attention (WSA) hinders global dependency modeling capability and 2) recent studies point out that local windows impair robustness. To overcome these challenges, we pursue a preferable trade-off between computational cost and performance. Accordingly, we propose a novel factorization self-attention (FaSA) mechanism that enjoys both the advantages of local window cost and long-range dependency modeling capability. By factorizing the conventional attention matrix into sparse subattention matrices, FaSA captures long-range dependencies, while aggregating mixed-grained information at a computational cost equivalent to the local WSA. Leveraging FaSA, we present the factorization vision transformer (FaViT) with a hierarchical structure. FaViT achieves high performance and robustness, with linear computational complexity concerning input image spatial resolution. Extensive experiments have shown FaViT's advanced performance in classification and downstream tasks. Furthermore, it also exhibits strong model robustness to corrupted and biased data and hence demonstrates benefits in favor of practical applications. In comparison to the baseline model Swin-T, our FaViT-B2 significantly improves classification accuracy by 1% and robustness by 7% , while reducing model parameters by 14% . Our code will soon be publicly available: at https://github.com/q2479036243/FaViT.

How Does the Public Evaluate Vaccines for Low-Incidence, Severe-Outcome Diseases? A General-Population Choice Experiment.

Because immunizing large numbers of healthy people could be required to reduce a relatively small number of infections, disease incidence has a large impact on cost effectiveness, even if the infection is associated with very serious health outcomes. In addition to cost effectiveness, the US Advisory Committee on Immunization Practices requires evidence of stakeholders' values and preferences to help inform vaccine recommendations. This study quantified general-population preferences for vaccine trade-offs among disease severity, disease incidence, and other vaccine features. We developed a best-practice discrete choice experiment survey and administered it to 1185 parents of children aged 12-23 years and 1203 young adults aged 18-25 years from a national opt-in consumer panel. The data were analyzed using exploded-logit latent-class analysis. Latent-class analysis identified two classes with similar relative-importance weights in both samples. One of the two classes represented about half the samples and had preferences consistent with well-structured, logically ordered, and acceptably precise stated-preference utility. Preferences for the other half of the samples were poorly defined over the ranges of vaccine and disease attributes evaluated. Both parents and young adults in the first class evaluated protection from a disease with 1 in 100 incidence and full recovery at home as having statistically the same preference utility as a disease with 1 in 1 million incidence requiring hospitalization and resulting in permanent deafness. The results suggest that vaccines that protect against low-incidence, severe-outcome diseases, provide 'peace of mind' benefits not captured by standard health-outcome metrics. The fact that half the respondents had poorly defined vaccine preferences is a reminder of the challenges of implementing patient-centric vaccine decision making.

A comparison of methods to suppress electrocardiographic artifacts in local field potential recordings

ObjectiveLocal field potential (LFP) recordings from deep brain stimulation (DBS) electrodes are often contaminated with electrocardiographic (ECG) artifacts that hinder the detection of disease-specific electrical brain activity. MethodsThree ECG suppression methods were evaluated: (1) QRS interpolation of the Perceive toolbox, (2) template subtraction, and (3) singular value decomposition (SVD). LFPs were recorded with the Medtronic PerceptTM PC system in nine Parkinson’s disease patients with stimulation OFF (“OFF-DBS”; anode disconnected) and ON at 0 mA (“ON-DBS 0 mA”; anode connected). Findings were verified with simulated ECG-contaminated time series. ResultsECG artifacts were present in 10 out of 18 ON-DBS 0 mA recordings. All ECG suppression methods drastically reduced artifact-induced beta band (13–35 Hz) power and at least partly recovered the beta peak and beta burst dynamics. Using external ECG recordings and lengthening artifact epoch length improved the performance of the suppression methods. Increasing epoch length, however, elevated the risk of flattening the beta peak and losing beta burst dynamics. ConclusionsThe SVD method formed the preferred trade-off between artifact cleaning and signal loss, as long as its parameter settings are adequately chosen. SignificanceECG suppression methods enable analysis of disease-specific neural activity from signals affected by ECG artifacts.

Dupuytren contracture: using qualitative data to inform a conceptual framework for shared decision-making.

Multiple treatment options are available to patients with Dupuytren contracture, making shared decision-making complex. Our rigorous qualitative analysis sought to understand patient perceptions of shared decision-making in Dupuytren contracture treatment and create a conceptual framework to optimize patient-physician communication. We interviewed 30 patients with Dupuytren contracture to learn about their experience with treatment selection. The following themes were integral to shared decision-making for Dupuytren contracture treatment: discussing disease progression and treatment initiation, presenting all available treatment options, assessing patients' pre-existing biases towards treatment, patient values and preferences for treatment trade-offs, treatment risks and benefits, physician recommendation and active patient participation. This model can optimize communication about treatment options and expectations for relevant outcomes including, recovery time, contracture recurrence, complications, and treatment-related expenses.Level of evidence: V.

Student COVID-19 vaccination preferences in China: A discrete choice experiment.

This study uses a discrete choice experiment (DCE) questionnaire to investigate student vaccination preferences for both intrinsic and extrinsic attributes. A two-part DCE questionnaire was distributed to 1,138 students through face-to-face interviews at vaccination centers in Qingdao, China. Conditional logit models were used to understand student preference trade-offs. Mixed logit models (MLM) and sub-group analysis were conducted to understanding student preference heterogeneity. We found that students preferred vaccines with fewer side effects (β = 0.845; 95% CI, 0.779-0.911), administered through third level health facilities (β = 0.170; 95% CI, 0.110-0.230), and had at least 1 year duration of protection (β = 0.396; 95% CI, 0.332-0.461. Higher perception of COVID-19 risks (β = 0.492; 95% CI, 0.432-0.552) increased the likelihood of student vaccination uptake. Surprisingly, vaccine effectiveness (60%) and percentages of acquaintances vaccinated (60%) reduced vaccination utility, which points to free-rider problems. In addition, we find that student study majors did not contribute to preference heterogeneity, and the main disparities in preferences were attributed to student risk tolerances. Both intrinsic and extrinsic attributes were influential factors shaping student preferences for COVID-19 vaccines. Our results inform universities and local governments across China on targeting their vaccination programs.

Evidence, values, and masks for control of COVID-19: 2 years later

Public health could have given better mask wearing guidance by acknowledging uncertainty regarding effectiveness, highlighting value and preference tradeoffs, and pointing out how tradeoffs differ across risk groups. In the absence of clear statements regarding each of these issues as part of public health guidance, for whom and when the benefits of mask wearing outweigh its burdens have remained contentious [1].

Stated-preference tradeoffs between regulatory costs and benefits: testing unit asking and double framing effects

A novel stated-preferences approach to imputing the value of life to estimate regulatory benefits elicits people’s preferred tradeoffs on behalf of the nation between national regulatory costs and nation-wide regulatory benefits, in contrast to the conventional approach of seeking estimates of the ‘value of a statistical life’ (VSL) by asking subjects how much they would be willing to pay personally for a small reduction in their own mortality risk. Two national-preference survey experiments were pursued here. The first U.S. experiment (n = 396) offered a between-person test of the effect of asking people to evaluate a hypothetical single life prolonged by regulation, before assessing the tolerable cost to the national economy of a hypothetical regulation prolonging 100 lives (LF frame). This unit asking task increased imputed means for the social benefit of a life prolonged (SB1LP*) in national tradeoffs. Cautions to respondents about responses that generated particularly low implicit SB1LP* values did not substantively reduce implausible values. The second U.S. experiment (n = 505) had people respond to both the lives-first (LF) frame, preceded by a unit asking task, and the costs-first (CF) frame (i.e. eliciting ‘reasonable’ numbers of lives prolonged if estimated regulatory cost is $1 billion each year). These frames both mimic the kinds of decisions that regulators face, as the VSL stated preference method does not. Higher LF values in earlier between-person studies were replicated in the unit asking within-person design here. A partial order effect occurred for the second experiment: starting with the CF frame yielded a subsequent LF mean four times higher. Open-ended probing found beliefs that regulatory costs are justified only by prolonging many lives may explain lower CF values. Using both frames can inform both conventional stated preference research (which uses only the LF frame) and regulators.

Cash holding and bankruptcy risk in Egyptian firms: the moderating effect of corporate social responsibility

PurposeThis study aims to investigate the impact of cash holding (CH) on bankruptcy (BR) risk. This study also examines the moderating effect of corporate social responsibility (CSR) practices on this relationship.Design/methodology/approachThe data were extracted from firms' annual reports. The panel data were used for 68 firms listed at the Egyptian Stock Exchange (EGX) with a total of 340 observations from 2015 to 2019. The research hypotheses were tested using the panel corrected standards errors (PCSE) method and the feasible generalized least squares (FGLS) method.FindingsThe results reveal that (1) CH has a positive effect on the Z-score (decreasing bankruptcy risk) of the Egyptian listed firms. (2) Egyptian firms that practice CSR have a low level of bankruptcy risk. (3) CSR practices in Egyptian listed firms support the positive relationship between CH and Z-score (declining bankruptcy risk).Research limitations/implicationsThe limitations of this study include a relatively small sample size. In addition, the analysis doesn't include other measures of bankruptcy risk due to a lack of data.Practical implicationsThe findings of this study will help investors and creditors to evaluate and predict the firms' bankruptcy risk. This study highlights the importance of cash holding for firms in emerging economies. Firms may hold cash to support liquidity, overcome financial distress risk, lower the cost of capital, increase future investment opportunities and reduce uncertainty. Additionally, the results would also help the policymakers, regulators at the EGX and Financial Regulatory Authority and stakeholders to realize the importance of cash holding, evaluate the cash liquidity in Egyptian listed firms, predict the firms' financial distress and consider the consequences of the CSR practices in accordance with Egypt's vision 2030.Originality/valueConsistent with liquidity preference theory and trade-off theory, this study adds evidence to the literature on bankruptcy risk by investigating the effect of cash holding on bankruptcy risk in emerging economies. According to Egypt's vision 2030, the empirical findings in this study extend previous findings by providing strong additional evidence in emerging economies regarding the moderating effect of CSR practices on the association between cash holding and bankruptcy risk. To the best of our knowledge, this study is the first to investigate the relationship between CSR, CH and BR risk in Egypt.

Transport network downsizing based on optimal sub-network

Transportation networks are sized to efficiently achieve some set of service objectives. Under particular circumstances, such as the COVID-19 pandemic, the demand for transportation can significantly change, both qualitatively and quantitatively, resulting in an over-capacitated and less efficient network. In this paper, we address this issue by proposing a framework for resizing the network to efficiently cope with the new demand. The framework includes a model to determine an optimal transportation sub-network that guarantees the following: (i) the minimal access time from any node of the urban network to the new sub-network has not excessively increased compared to that of the original transportation network; (ii) the delay induced on any itinerary by the removal of nodes from the original transportation network has not excessively increased; and (iii) the number of removed nodes from the transportation network is within a preset known factor. A solution is optimal if it induces a minimal global delay. We modelled this problem as a Mixed Integer Linear Program and applied it to the public bus transportation network of Lyon, France, in a case study. In order to respond to operational issues, the framework also includes a decision tool that helps the network planners to decide which bus lines to close and which ones to leave open according to specific trade-off preferences. The results on real data in Lyon show that the optimal sub-network from the MILP model can be used to feed the decision tool, leading to operational scenarios for network planners.