Public funding for pharmaceuticals often follows repeated negotiation between manufacturers and a public agency, yet little empirical work examines how the timing of agreement reflects the economic structure of these interactions. Using a duration model of negotiations in Australia from 2005 to 2018, we assess whether observed patterns of delay and agreement align with dynamic bargaining theory under incomplete information. Agreements from 634 submissions for 400 therapies required a median of 16months, and 71% of negotiation rounds ended without agreement. Therapies with lower expected value to the agency-reflected in higher incremental cost per QALY, greater budget impact, or evidence uncertainty-experienced longer delays and lower agreement rates, while those with strong clinical importance, perceived need, or elevated public interest were listed more quickly. Delays also varied across therapeutic classes and with a therapy's position in the sequencing of available treatments. The observed patterns point to a systematic listing rule in which therapies are funded when expected health gains justify their opportunity costs. They also support the view that the timing of agreement reflects strategic negotiation under uncertainty, not simply procedural delay.

Organizations adopting generative AI (GenAI) face complex strategic tensions among management, departments, and employees that fundamentally determine adoption outcomes. This study develops a multi-level Bayesian game-theoretic framework modeling these multi-stakeholder interactions, identifying four distinct adoption patterns through formal equilibrium analysis. Our theoretical derivations establish that successful GenAI implementation requires three analytically-derived conditions: (1) strong strategic complementarity across departments, (2) efficient investment allocation, and (3) effective employee displacement mitigation. The formal model specifies explicit utility functions for three stakeholder groups — senior management, departmental units, and individual employees — and characterizes Bayesian Nash equilibria under incomplete information. Companies must simultaneously invest in cross-functional coordination mechanisms, establish shared governance structures, and implement workforce development programs that position GenAI as a capability enhancement rather than a job replacement. Our computational analysis, based on 10,000 Monte Carlo simulations with explicit parameter specifications and convergence criteria, demonstrates that coordination-focused strategies significantly outperform technology-focused approaches in organizational welfare, providing actionable guidance for AI transformation leadership. • Multi-level Bayesian game models GenAI adoption inside organizations. • Strategic complementarity drives coordinated GenAI value creation. • Employee displacement risks critically shape adoption equilibria. • Coordination strategies outperform technology-first GenAI adoption. • Formal thresholds distinguish value co-creation from co-destruction.

Spacecraft pursuit-evasion game with incomplete information: A pre-training approximate dynamic programming method

A two-stage three-way dynamic consensus approach for large-scale group decision-making in social networks with incomplete information and adjustment willingness

Large-scale studies examining the demographic, serological, and seasonal characteristics of myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) and aquaporin-4 immunoglobulin G-positive neuromyelitis optica spectrum disorder (AQP4+NMOSD) remain limited, despite their potential to provide crucial information for resource allocation, clinical trial design, and recruitment. To investigate demographic, serological, and seasonal variations in MOGAD and AQP4+NMOSD using a large neuroimmunology laboratory registry validated by clinical cohorts. We conducted a retrospective, laboratory-based study using data from the Mayo Clinic Neuroimmunology Laboratory registry and clinical cohorts between July 2014 and April 2024. The first available serum sample from each patient tested for MOG-IgG and AQP4-IgG was included. Analyses focused on age and sex distributions, antibody titers, and seasonal patterns of seropositivity, as well as the month and season of disease onset and attacks among the clinical cohorts of MOGAD and AQP4+NMOSD. We included 89,495 sera tested for MOG-IgG and 198,401 for AQP4-IgG, supplemented by validated clinical cohorts of 528 patients with MOGAD and 534 with AQP4+NMOSD. MOG-IgG was detected in 6,313 samples (7.1%), with 1,566 (24.8%) exhibiting high titers (≥1:1,000). AQP4-IgG was positive in 5,057 samples (2.5%). Individuals with MOG-IgG positivity were younger than those with AQP4-IgG positivity (mean age, 34.1 years [SD=20.0] vs 47.7 years [SD=17.9]; p<0.0001). The frequency of MOG-IgG positivity was highest among patients younger than 12 years (1,052 [17.9%]) and declined with older age, while AQP4-IgG positivity increased with older age. AQP4-IgG revealed a strong female predilection (female-to-male ratio 6.2:1), varying by age, whereas MOG-IgG showed a modest female predominance (female-to-male ratio 1.5:1), consistent across all ages. MOG-IgG-titers peaked in younger children and older adults, while AQP4-IgG titers remained stable across ages. Both diseases showed a winter peak in seropositivity, disease onset, and relapses. This large-scale registry analysis provides comprehensive demographic and serological characterization of MOGAD and AQP4+NMOSD. The modest winter peak suggests that seasonal infectious triggers may play an important role in disease pathogenesis. Limitations include incomplete clinical information within the laboratory registry and a referral-based testing population. These findings have important implications for healthcare planning and optimization of clinical trial design and recruitment.

Identification of physiological parameters and missing inputs in glucose dynamics via PINN-constrained inference.

Cardiac dysfunction after acute ischaemic stroke: Long-term outcomes from the SICFAIL cohort.

Polygenic risk for depression is associated with post-stroke depression and stroke.

Shortest path problems of uncertain random networks under incomplete information environments: Models and algorithms

DNA as a data storage medium.

Decision-making (DM) problems in real-world environments are frequently described by ambiguity, expert hesitation, linguistic assessments, and incomplete information, which limit the effectiveness of traditional fuzzy set (FS) and intuitionistic fuzzy set (IFS) frameworks. To deal with such problems, this paper demonstrates an innovative and more expressive model, called linguistic cubic interval-valued intuitionistic fuzzy sets (LCuIVIFSs), which combines interval-valued intuitionistic fuzzy uncertainty, linguistic information, and cubic structures into a single framework. Some traditional operations of the newly defined LCuIVIFS model, such as union, intersection, and complement, are systematically introduced to ensure operational consistency and mathematical soundness. Within the framework of LCuIVIFSs, several aggregation operators (AOs), including arithmetic AO, geometric AO, weighted arithmetic AO, and weighted geometric AO, is presented to significantly combine complex and uncertain information. The key features of the proposed AOs are investigated. Moreover, a novel multi-criteria decision-making (MCDM) technique is developed using the newly defined AOs. To discuss the significance of the proposed approach, it is implemented to a case study of supplier selection problem in smart manufacturing, where both quantitative and qualitative criteria under uncertainty are considered. The final results ensure that the newly defined approach contributes reliable, flexible, and robust decision outcomes compared with existing FS-based models. The proposed study thus provides a valuable decision-support framework for complex DM problems under linguistic and cubic uncertainty.

Funding based on population need is key to equitable health care. The formula for general practice capitation payments in England has not been updated for over 20 years and is based on crude workload weights. To estimate and describe the uplifts in practice payments required to bring them in line with updated and more precise needs-based workload weights. Cross-sectional study assessing the most updated routinely available data on 6213 non-atypical general practices in England with >1000 registered patients and complete information on payments as of 1 April 2024. Updated workload weights were applied to publicly available practice data on patient age and gender, new registration, ethnicity, deprivation, and prevalence of 20 long-term conditions. The practice payment provided for input price variation was removed and the current workload payment per weighted patient for each practice and the funding required to uplift all practices to minimum thresholds were calculated. Workload payment per patient was £92.66 on average, varying from £86.72 in the lowest to £99.91 in the highest deprivation decile. Workload payment per weighted patient varied from £81.40 (5th percentile) to £107.10 (95th percentile), and from £89.69 in the lowest to £96.40 in the highest deprivation decile. Uplifting payments to a minimum of £102.46 per weighted patient (90th percentile value) would increase total payments by 11.6% (£677.77 million per year) and payments per patient to £99.75 in the lowest and £108.18 in the highest deprivation decile. The workload element of the general practice funding formula should use updated weights. Additional concerns related to equity in outcomes and underfunding of practices in deprived areas should be addressed by adjustments or revisions of other elements of the funding formula.

ABSTRACT Spanish-speaking children face a myriad of cultural, linguistic, and systemic barriers and discrimination when seeking Special Education services. This pilot study assessed a brief seminar-based intervention to increase Latinx families’ knowledge about SPED services and explored families’ qualitative experiences with the school system (N = 10 families). Parents’ knowledge of special education increased pre- versus post-seminar (t([9]) = -4.846, p < 0.001). Children’s psychiatric diagnoses were positively correlated with the number of traumatic and stressor events (t(18) = -9.08, p = .006). Families reported that schools would provide them with incomplete information about their children’s needs and that persistence was a positive way to navigate these services.

The automatic retrieval of boundary information from image objects suffers from the problem of under and over-segmentation, where the former leads to missed object detection, while the latter delivers an improper object shape. A method to optimize the automatic retrieval of complete and proper boundary information is proposed in this research based on an unsupervised approach. The strategy is to utilize the trade-off between the coefficient of variation from shape distribution against the mean of entropy contribution from segmented regions. This mechanism relies on the assumption that the segmentation result of a natural image contains a prominent main object representation with its details which are presented as a normal distribution of segmented regions. The research also enhances the entropy-based segmentation evaluation by redefining the computation of image entropy and segmentation entropy. The experiment shows that the proposed approach is capable of reducing over-segmentation by 57.20% compared to the existing algorithm, while at the same time reducing the consumption time by 85.26%. The empirical evaluation shows that the proposed approach delivers the highest accuracy among other evaluated methods. Qualitative validation based on groups of human observers shows that the proposed approach is the most desired algorithm for producing boundary information and measuring segmentation quality. These findings suggest that the trade-off between the mean of entropy contribution from the segmented regions and the coefficient of variation from shape distribution becomes an effective feature for unsupervised retrieval of boundary information.

The interventional landscape for psoriasis has expanded considerably with the introduction of systemic and biologic agents, intensifying the importance of comprehensive adverse event (AE) documentation to inform safe clinical practice. While regulatory initiatives such as the Food and Drug Administration Amendments Act Final Rule (FDAAA Final Rule) were designed to enhance clinical trial transparency, disparities between AE data presented in ClinicalTrials.gov and peer-reviewed literature continue to occur, potentially complicating interpretation of publicly available harms information used in evidence synthesis and clinical decision-making. Variability in how serious adverse events (SAEs), other adverse events (OAEs), deaths, and treatment discontinuations due to AE are documented can distort safety perceptions and influence clinical practice. This study compares AE reporting between ClinicalTrials.gov and corresponding peer-reviewed publications for psoriasis clinical trials and evaluates whether completeness and concordance differ before versus after implementation of the FDAAA Final Rule. We performed a cross-sectional analysis using systematic retrieval of psoriasis trials from ClinicalTrials.gov with results posted between 2009 and 2024. Data on AEs-encompassing SAEs, OAEs, treatment discontinuation due to AE, and deaths-were independently abstracted by two reviewers, with disagreements resolved through adjudication. Trials were categorized according to FDA regulatory status, and variations in AE documentation were examined using descriptive statistics, chi-squared tests, Bland-Altman and funnel plots, and regression modeling. The study's primary endpoint was completeness and agreement of AE documentation between ClinicalTrials.gov and corresponding peer-reviewed publications across four predefined domains: SAEs, OAEs, discontinuation due to AE, and death. Pre-Final Rule AE reporting patterns were mixed across domains. After implementation of the FDAAA Final Rule, registry entries more consistently reported SAEs, OAEs, and mortality than corresponding publications. Although death documentation in registry entries showed improvement after the rule's enactment, publications demonstrated persistent inconsistency. Following implementation of the FDAAA Final Rule, SAEs were present in 53% of registry records versus 40% of publications. OAEs were captured in 51% of registry entries but only 22% of publications. Disparities in SAE totals occurred in over 90% of trials, frequently with registry entries recording higher counts. Numerous discrepancies stemmed from incomplete calculations, ambiguous characterizations, or AE information confined to narrative text. Both visual and statistical assessments revealed a pattern of underreporting in publications, with minimal temporal improvement irrespective of trial scale or sponsorship type. Notwithstanding current regulatory frameworks, AE documentation for psoriasis trials remains fragmented with substantial inconsistency between registry and published data. These shortcomings undermine rigorous safety assessment and evidence-based clinical recommendations. Improved patient care requires increased integration of registry data into evidence synthesis, better clarification of AE definitions, and standardized AE reporting practices. Preregistered on The International Prospective Register of Systematic Reviews (PROSPERO; CRD420251081207) and Open Science Framework (4jaz3).

"Rapid" "Fast" or "Abbreviated" MRI examinations, which leverage fewer sequences to answer a targeted clinical question, are increasingly being explored as clinical tools. The purpose of this study was to evaluate and compare payor costs and patient out-of-pocket costs for brain MRI without contrast compared to a rapid brain MRI among commercially insured patients at a quaternary academic children's hospital. We performed a retrospective search to identify patients with private insurance who underwent an outpatient brain MRI without contrast (standard MRI) or rapid brain MRI at our quaternary academic children's hospital. Examinations lacking complete payment information were excluded, and only the first exam per patient was included. All rapid MRIs were included; an equal number of standard MRIs were randomly sampled by month. For rapid MRI examinations, we calculated the frequency of cases in which the payor did not recognize the limited modifier, as indicated by the payor reason codes. For all examinations, we calculated the (1) payor cost (the amount reimbursed for the exam) and (2) the patient's total out-of-pocket cost, calculated as the sum of the deductible, coinsurance, and co-payment. Descriptive statistics were used and means were compared between groups using Student's t test. Our sample included 147 standard MRIs and 166 rapid MRIs (coded with the 52 modifier). Most examinations included patient cost sharing: 77% (113/147) of standard MRIs and 69% of rapid MRIs (115/166). Payor reimbursement differed (P<0.001) by examination, with higher payor costs for standard MRI (mean, $2,760; SD, $1,187 vs. mean, $1,986; SD, $198) for rapid MRI. Among examinations with cost sharing, the mean total out-of-pocket costs were similar between examination types (standard, $1,206 vs rapid, $1,285; P=0.55). In 43% (71/166) of rapid MRI examinations, the payor did not recognize the limited modifier. Although rapid brain MRIs reduced payor reimbursement, patient out-of-pocket costs remained unchanged. Inconsistent recognition of the limited modifier underscores the need for updated CPT codes and reimbursement policies aligned with evolving imaging practices.

Acrylamide is a widespread environmental and dietary contaminant, and hemoglobin adducts of acrylamide (HbAA) and its metabolite glycidamide (HbGA) serve as objective biomarkers of internal exposure. However, the association between acrylamide exposure and rheumatoid arthritis (RA) has not been well characterized at the population level. We analyzed data from 4 cycles of the US National Health and Nutrition Examination Survey (2003–2006 and 2013–2016), including 1972 adults aged ≥20 years with complete information on RA status and acrylamide hemoglobin adduct measurements. Multivariable logistic regression models accounting for the complex survey design were used to examine associations between HbAA, HbGA, total adducts (HbAA + HbGA), the metabolic ratio (HbGA/HbAA), and RA prevalence. Restricted cubic spline models were applied to explore potential nonlinear exposure–response relationships. Nonlinear associations were observed between acrylamide-related hemoglobin adducts and the prevalence of RA. HbAA exhibited a J-shaped association, with higher odds observed at intermediate exposure levels. In contrast, HbGA and total adduct concentrations showed increasing associations across higher exposure ranges. The HbGA/HbAA ratio was positively associated with RA prevalence. Restricted cubic spline analyses supported these non-monotonic exposure–response patterns. Subgroup analyses showed variability in point estimates across behavioral and clinical strata, but estimates were often imprecise and should be interpreted as exploratory. In this nationally representative cross-sectional study, HbAA was nonlinearly associated with the prevalence of RA among US adults. These findings support an epidemiological association between internal biomarkers of acrylamide exposure and RA. Prospective studies with repeated exposure assessment and mechanistic measurements are needed to clarify temporal relationships and underlying biological pathways.

Artificial intelligence decision systems are increasingly deployed in safety-, policy-, and human-sensitive settings where actions must satisfy feasibility constraints under incomplete information. Existing post-deployment monitoring approaches can detect observable failures, distribution shifts, or performance degradation, but they cannot, by themselves, determine whether feasibility can be guaranteed when safety-relevant latent states remain indistinguishable at decision time. This paper develops a formal framework for reliable deployment and monitoring of AI decision systems under fixed observation structures. We model deployment through latent states, observations, observation-consistent state sets, state-wise feasibility constraints, and observation-based policies. The framework characterizes when feasibility-guaranteed deployment is structurally possible, when it requires intervention, and when it is impossible without modifying the information structure. We prove that every task falls into one of three regimes: deployable and automatically measurable systems, non-automatically deployable but remediable systems, and hard non-deployable systems. We further introduce an operator-assisted review and rollback mechanism for remediable cases and show that additional data or monitoring alone is insufficient unless such measurements refine feasibility-relevant latent-state ambiguity. Empirical examples and a digital health case study illustrate how the framework supports practical deployment assessment, monitoring design, and human-in-the-loop safeguards for AI systems operating under partial observability.

Early identification of acute hematologic toxicity (HT) in locally advanced rectal cancer (LARC) patients undergoing radiotherapy is crucial for optimizing clinical outcomes. Here, we retrospectively collected multi-center LARC patients (n = 464, n = 56, and n = 79) with complete CT images, dose maps, hematologic biomarkers, and demographic information. A Transformer-based multimodal fusion model was constructed to combine the visual and non-visual representation features for HT prediction, and the study also testified to the modality-specific and region-specific contributions to HT. The multimodal fusion model achieved a state-of-the-art HT prediction performance in LARC patients: with an area under the curve (AUC) of 0.828 (95% confidence interval [CI]: 0.820-0.835), 0.757 (95% CI: 0.750-0.766), and 0.756 (95% CI: 0.752-0.762) in internal and two external testing datasets. The initial hematologic biomarkers were the best unimodal risk indicator, while the planning target volume served as the most sensitive region. The study confirms the sole and combined contributions of each modality to the radiotherapy-induced HT in LARC patients, and the multimodal fusion model shows promising interpretability and generalization for HT occurrence, which offers valuable insights to optimize personalized treatment plans for high-risk patients.

With global population aging and an increasing prevalence of cognitive decline, the association between cognitive change as a dynamic process and all-cause mortality remains insufficiently investigated. We therefore aimed to assess the association of cognitive decline on all-cause mortality in older women. We analyzed data originating from the cognitive cohort of the Women's Health Study established in 1998. 6,377 US-based women aged ≥ 65 years and currently or formerly employed in health-related professions were enrolled at baseline. Women with complete information on 4-year global cognitive performance change were eligible for our analysis and assigned to quintiles according to change in global cognitive performance. Follow-up for all-cause mortality was administered until December 31, 2022. Multivariable adjusted Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% CIs of the effect of cognitive change on the risk of all-cause mortality. Secondary analyses evaluated the risk in the lowest 20% and 10% of the cognitive change distribution, and according to verbal memory changes. 5,214 women with a mean age of 66.1 years (SD 4.0) were included in the analyses. Women were followed for an average of 13.4 person years and 3,333 deaths were observed. When compared to the 5th quintile with the greatest improvement, the adjusted HR for all-cause mortality was 0.96 (95%CI [0.86-1.08]) for the 4th, 1.04 (95%CI [0.93-1.16]) for the 3rd, 1.15 (95%CI [1.04-1.29]) for the 2nd, and 1.37 (95%CI [1.23-1.52]) for the 1st quintile characterized by the greatest decline. Similar patterns were observed for verbal memory change. Risk was further elevated when comparing the worst 20% (HR: 1.32 (95%CI [1.21-1.43])) and 10% (HR: 1.44 (95%CI [1.29-1.60])) to all other participants. Cognitive decline over four years was associated with an increased risk of mortality among older women. Further studies should explore whether declines earlier in life or among men are also associated with an increased risk of mortality.