Group-by-Treatment Interaction Effects in Comparative Bioavailability Studies.

Comparative bioavailability studies often involve multiple groups of subjects for a variety of reasons, such as clinical capacity limitations. This raises questions about the validity of pooling data from these groups in the statistical analysis and whether a group-by-treatment interaction should be evaluated. We investigated the presence or absence of group-by-treatment interactions through both simulation techniques and a meta-study of well-controlled trials. Our findings reveal that the test falsely detects an interaction when no true group-by-treatment interaction exists. Conversely, when a true group-by-treatment interaction does exist, it often goes undetected. In our meta-study, the detected group-by-treatment interactions were observed at approximately the level of the test and, thus, can be considered false positives. Testing for a group-by-treatment interaction is both misleading and uninformative. It often falsely identifies an interaction when none exists and fails to detect a real one. This occurs because the test is performed between subjects in crossover designs, and studies are powered to compare treatments within subjects. This work demonstrates a lack of utility for including a group-by-treatment interaction in the model when assessing single-site comparative bioavailability studies, and the clinical trial study structure is divided into groups.

Patients' and physicians' beliefs and attitudes towards integrating personalized risk estimates into patient education about left ventricular assist device therapy.

BACKGROUND: Personalized risk (PR) estimates may enhance clinical decision making and risk communication by providing individualized estimates of patient outcomes. We explored stakeholder attitudes toward the utility, acceptability, usefulness and best-practices for integrating PR estimates into patient education and decision making about Left Ventricular Assist Device (LVAD). METHODS AND RESULTS: As part of a 5-year multi-institutional AHRQ project, we conducted 40 interviews with stakeholders (physicians, nurse coordinators, patients, and caregivers), analyzed using Thematic Content Analysis. All stakeholder groups voiced positive views towards integrating PR in decision making. Patients, caregivers and coordinators emphasized that PR can help to better understand a patient's condition and risks, prepare mentally and logistically for likely outcomes, and meaningfully engage in decision making. Physicians felt it can improve their decision making by enhancing insight into outcomes, enhance tailored pre-emptive care, increase confidence in decisions, and reduce bias and subjectivity. All stakeholder groups also raised concerns about accuracy, representativeness and relevance of algorithms; predictive uncertainty; utility in relation to physician's expertise; potential negative reactions among patients; and overreliance. CONCLUSION: Stakeholders are optimistic about integrating PR into clinical decision making, but acceptability depends on prospectively demonstrating accuracy, relevance and evidence that benefits of PR outweigh potential negative impacts on decision making quality.

Improved discovery of RNA-binding protein binding sites in eCLIP data using DEWSeq.

Enhanced crosslinking and immunoprecipitation (eCLIP) sequencing is a method for transcriptome-wide detection of binding sites of RNA-binding proteins (RBPs). However, identified crosslink sites can deviate from experimentally established functional elements of even well-studied RBPs. Current peak-calling strategies result in low replication and high false positive rates. Here, we present the R/Bioconductor package DEWSeq that makes use of replicate information and size-matched input controls. We benchmarked DEWSeq on 107 RBPs for which both eCLIP data and RNA sequence motifs are available and were able to more than double the number of motif-containing binding regions relative to standard eCLIP processing. The improvement not only relates to the number of binding sites (3.1-fold with known motifs for RBFOX2), but also their subcellular localization (1.9-fold of mitochondrial genes for FASTKD2) and structural targets (2.2-fold increase of stem-loop regions for SLBP. On several orthogonal CLIP-seq datasets, DEWSeq recovers a larger number of motif-containing binding sites (3.3-fold). DEWSeq is a well-documented R/Bioconductor package, scalable to adequate numbers of replicates, and tends to substantially increase the proportion and total number of RBP binding sites containing biologically relevant features.

Innate extracellular Hsp70 inflammatory properties are mediated by the interaction of Siglec-E and LOX-1 receptors.

Innate immune responses to cell damage-associated molecular patterns induce a controlled degree of inflammation, ideally avoiding the promotion of intense unwanted inflammatory adverse events. When released by damaged cells, Hsp70 can stimulate different responses that range from immune activation to immune suppression. The effects of Hsp70 are mediated through innate receptors expressed primarily by myeloid cells, such as dendritic cells (DCs). The regulatory innate receptors that bind to extracellular mouse Hsp70 (mHsp70) are not fully characterized, and neither are their potential interactions with activating innate receptors. Here, we describe that extracellular mHsp70 interacts with a receptor complex formed by inhibitory Siglec-E and activating LOX-1 on DCs. We also find that this interaction takes place within lipid microdomains, and Siglec-E acts as a negative regulator of LOX-1-mediated innate activation upon mHsp70 or oxidized LDL binding. Thus, HSP70 can both bind to and modulate the interaction of inhibitory and activating innate receptors on the cell surface. These findings add another dimension of regulatory mechanism to how self-molecules contribute to dampening of exacerbated inflammatory responses.

Trust criteria for artificial intelligence in health: normative and epistemic considerations.

Rapid advancements in artificial intelligence and machine learning (AI/ML) in healthcare raise pressing questions about how much users should trust AI/ML systems, particularly for high stakes clinical decision-making. Ensuring that user trust is properly calibrated to a tool's computational capacities and limitations has both practical and ethical implications, given that overtrust or undertrust can influence over-reliance or under-reliance on algorithmic tools, with significant implications for patient safety and health outcomes. It is, thus, important to better understand how variability in trust criteria across stakeholders, settings, tools and use cases may influence approaches to using AI/ML tools in real settings. As part of a 5-year, multi-institutional Agency for Health Care Research and Quality-funded study, we identify trust criteria for a survival prediction algorithm intended to support clinical decision-making for left ventricular assist device therapy, using semistructured interviews (n=40) with patients and physicians, analysed via thematic analysis. Findings suggest that physicians and patients share similar empirical considerations for trust, which were primarily epistemic in nature, focused on accuracy and validity of AI/ML estimates. Trust evaluations considered the nature, integrity and relevance of training data rather than the computational nature of algorithms themselves, suggesting a need to distinguish 'source' from 'functional' explainability. To a lesser extent, trust criteria were also relational (endorsement from others) and sometimes based on personal beliefs and experience. We discuss implications for promoting appropriate and responsible trust calibration for clinical decision-making use AI/ML.

A systematic review of the development and application of home cage monitoring in laboratory mice and rats.

BACKGROUND: Traditionally, in biomedical animal research, laboratory rodents are individually examined in test apparatuses outside of their home cages at selected time points. However, the outcome of such tests can be influenced by various factors and valuable information may be missed when the animals are only monitored for short periods. These issues can be overcome by longitudinally monitoring mice and rats in their home cages. To shed light on the development of home cage monitoring (HCM) and the current state-of-the-art, a systematic review was carried out on 521 publications retrieved through PubMed and Web of Science. RESULTS: Both the absolute (~ × 26) and relative (~ × 7) number of HCM-related publications increased from 1974 to 2020. There was a clear bias towards males and individually housed animals, but during the past decade (2011-2020), an increasing number of studies used both sexes and group housing. In most studies, animals were kept for short (up to 4 weeks) time periods in the HCM systems; intermediate time periods (4-12 weeks) increased in frequency in the years between 2011 and 2020. Before the 2000s, HCM techniques were predominantly applied for less than 12 h, while 24-h measurements have been more frequent since the 2000s. The systematic review demonstrated that manual monitoring is decreasing in relation to automatic techniques but still relevant. Until (and including) the 1990s, most techniques were applied manually but have been progressively replaced by automation since the 2000s. Independent of the year of publication, the main behavioral parameters measured were locomotor activity, feeding, and social behaviors; the main physiological parameters were heart rate and electrocardiography. External appearance-related parameters were rarely examined in the home cages. Due to technological progress and application of artificial intelligence, more refined and detailed behavioral parameters have been investigated in the home cage more recently. CONCLUSIONS: Over the period covered in this study, techniques for HCM of mice and rats have improved considerably. This development is ongoing and further progress as well as validation of HCM systems will extend the applications to allow for continuous, longitudinal, non-invasive monitoring of an increasing range of parameters in group-housed small rodents in their home cages.

Challenges and advanced concepts for the assessment of learning and memory function in mice.

The mechanisms underlying the formation and retrieval of memories are still an active area of research and discussion. Manifold models have been proposed and refined over the years, with most assuming a dichotomy between memory processes involving non-conscious and conscious mechanisms. Despite our incomplete understanding of the underlying mechanisms, tests of memory and learning count among the most performed behavioral experiments. Here, we will discuss available protocols for testing learning and memory using the example of the most prevalent animal species in research, the laboratory mouse. A wide range of protocols has been developed in mice to test, e.g., object recognition, spatial learning, procedural memory, sequential problem solving, operant- and fear conditioning, and social recognition. Those assays are carried out with individual subjects in apparatuses such as arenas and mazes, which allow for a high degree of standardization across laboratories and straightforward data interpretation but are not without caveats and limitations. In animal research, there is growing concern about the translatability of study results and animal welfare, leading to novel approaches beyond established protocols. Here, we present some of the more recent developments and more advanced concepts in learning and memory testing, such as multi-step sequential lockboxes, assays involving groups of animals, as well as home cage-based assays supported by automated tracking solutions; and weight their potential and limitations against those of established paradigms. Shifting the focus of learning tests from the classical experimental chamber to settings which are more natural for rodents comes with a new set of challenges for behavioral researchers, but also offers the opportunity to understand memory formation and retrieval in a more conclusive way than has been attainable with conventional test protocols. We predict and embrace an increase in studies relying on methods involving a higher degree of automatization, more naturalistic- and home cage-based experimental setting as well as more integrated learning tasks in the future. We are confident these trends are suited to alleviate the burden on animal subjects and improve study designs in memory research.

Emergency Department Pediatric Readiness and Disparities in Mortality Based on Race and Ethnicity.

Importance: Presentation to emergency departments (EDs) with high levels of pediatric readiness is associated with improved pediatric survival. However, it is unclear whether children of all races and ethnicities benefit equitably from increased levels of such readiness. Objective: To evaluate the association of ED pediatric readiness with in-hospital mortality among children of different races and ethnicities with traumatic injuries or acute medical emergencies. Design, Setting, and Participants: This cohort study of children requiring emergency care in 586 EDs across 11 states was conducted from January 1, 2012, through December 31, 2017. Eligible participants included children younger than 18 years who were hospitalized for an acute medical emergency or traumatic injury. Data analysis was conducted between November 2022 and April 2023. Exposure: Hospitalization for acute medical emergency or traumatic injury. Main Outcomes and Measures: The primary outcome was in-hospital mortality. ED pediatric readiness was measured through the weighted Pediatric Readiness Score (wPRS) from the 2013 National Pediatric Readiness Project assessment and categorized by quartile. Multivariable, hierarchical, mixed-effects logistic regression was used to evaluate the association of race and ethnicity with in-hospital mortality. Results: The cohort included 633 536 children (median [IQR] age 4 [0-12] years]). There were 557 537 children (98 504 Black [17.7%], 167 838 Hispanic [30.1%], 311 157 White [55.8%], and 147 876 children of other races or ethnicities [26.5%]) who were hospitalized for acute medical emergencies, of whom 5158 (0.9%) died; 75 999 children (12 727 Black [16.7%], 21 604 Hispanic [28.4%], 44 203 White [58.2%]; and 21 609 of other races and ethnicities [27.7%]) were hospitalized for traumatic injuries, of whom 1339 (1.8%) died. Adjusted mortality of Black children with acute medical emergencies was significantly greater than that of Hispanic children, White children, and of children of other races and ethnicities (odds ratio [OR], 1.69; 95% CI, 1.59-1.79) across all quartile levels of ED pediatric readiness; but there were no racial or ethnic disparities in mortality when comparing Black children with traumatic injuries with Hispanic children, White children, and children of other races and ethnicities with traumatic injuries (OR 1.01; 95% CI, 0.89-1.15). When compared with hospitals in the lowest quartile of ED pediatric readiness, children who were treated at hospitals in the highest quartile had significantly lower mortality in both the acute medical emergency cohort (OR 0.24; 95% CI, 0.16-0.36) and traumatic injury cohort (OR, 0.39; 95% CI, 0.25-0.61). The greatest survival advantage associated with high pediatric readiness was experienced for Black children in the acute medical emergency cohort. Conclusions and Relevance: In this study, racial and ethnic disparities in mortality existed among children treated for acute medical emergencies but not traumatic injuries. Increased ED pediatric readiness was associated with reduced disparities; it was estimated that increasing the ED pediatric readiness levels of hospitals in the 3 lowest quartiles would result in an estimated 3-fold reduction in disparity for pediatric mortality. However, increased pediatric readiness did not eliminate disparities, indicating that organizations and initiatives dedicated to increasing ED pediatric readiness should consider formal integration of health equity into efforts to improve pediatric emergency care.

Defining the condensate landscape of fusion oncoproteins.

Fusion oncoproteins (FOs) arise from chromosomal translocations in ~17% of cancers and are often oncogenic drivers. Although some FOs can promote oncogenesis by undergoing liquid-liquid phase separation (LLPS) to form aberrant biomolecular condensates, the generality of this phenomenon is unknown. We explored this question by testing 166 FOs in HeLa cells and found that 58% formed condensates. The condensate-forming FOs displayed physicochemical features distinct from those of condensate-negative FOs and segregated into distinct feature-based groups that aligned with their sub-cellular localization and biological function. Using Machine Learning, we developed a predictor of FO condensation behavior, and discovered that 67% of ~3000 additional FOs likely form condensates, with 35% of those predicted to function by altering gene expression. 47% of the predicted condensate-negative FOs were associated with cell signaling functions, suggesting a functional dichotomy between condensate-positive and -negative FOs. Our Datasets and reagents are rich resources to interrogate FO condensation in the future.

A Workflow Guide to RNA-Seq Analysis of Chaperone Function and Beyond.

RNA sequencing (RNA-seq) is a powerful method of transcriptional analysis that allows for the sequence identification and quantification of cellular transcripts. RNA-seq can be used for differential gene expression (DGE) analysis, gene fusion detection, allele-specific expression, isoform and splice variant quantification, and identification of novel genes. These applications can be used for downstream systems biology analyses such as gene ontology or pathway analysis to provide insight into processes altered between biological conditions. Given the wide range of signaling pathways subject to chaperone activity as well as numerous chaperone functions in RNA metabolism, RNA-seq may provide a valuable tool for the study of chaperone proteins in biology and disease. This chapter outlines an example RNA-seq workflow to determine differentially expressed (DE) genes between two or more sample conditions and provides some considerations for RNA-seq experimental design.

Recent breakthroughs in computational structural biology harnessing the power of sequences and structures.

Recent advances in computational approaches and their integration into structural biology enable tackling increasingly complex questions. Here, we discuss several key areas, highlighting breakthroughs and remaining challenges. Theoretical modeling has provided tools to accurately predict and design protein structures on a scale currently difficult to achieve using experimental approaches. Molecular Dynamics simulations have become faster and more precise, delivering actionable information inaccessible by current experimental methods. Virtual screening workflows allow a high-throughput approach to discover ligands that bind and modulate protein function, while Machine Learning methods enable the design of proteins with new functionalities. Integrative structural biology combines several of these approaches, pushing the frontiers of structural and functional characterization to ever larger systems, advancing towards a complete understanding of the living cell. These breakthroughs will accelerate and significantly impact diverse areas of science.

Bioequivalence of Intravenous Alteplase from Two Different Manufacturing Processes in Healthy Male Volunteers: Results from a Two-Stage, Adaptive-Design Study.

OBJECTIVE: Alteplase is a recombinant tissue plasminogen activator used for thrombolytic treatment in several indications and is currently approved in Europe under the brand name Actilyse®. The current manufacturing process for alteplase was recently modified to meet increasing global demands. The aim of this randomized, open-label, adaptive two-stage design, two-way crossover study was to establish bioequivalence of alteplase derived from the two manufacturing processes (modified versus current). METHODS: The two alteplase formulations (modified and current, 0.2 mg/kg body weight) were compared in healthy male volunteers after intravenous infusion over a period of 30 min. The trial was put on hold after treatment of 12 subjects (Part A) and restarted as Part B (n = 18) with design adaptations, including a heparin bolus. RESULTS: Pharmacokinetic parameters of alteplase were determined from plasma concentration-time profiles. The pharmacokinetic parameters tested (AUC0-tz, Cmax, and AUC0-∞) for alteplase after single intravenous infusion demonstrated no differences between alteplase obtained from the modified and current processes. An analysis of variance (ANOVA) model was applied to test for bioequivalence. The geometric means ratio and the respective 92.83% confidence intervals (CIs) for all primary and secondary pharmacokinetic endpoints were well within the prespecified equivalence boundaries of 80-125%. The CIs also included unity, suggesting no statistically significant differences between the two treatments. CONCLUSIONS: The results show that alteplase exposure was virtually identical for the formulations tested, and statistical evaluation demonstrated bioequivalence of the formulations. Both formulations of alteplase were well tolerated by the subjects at the single intravenous doses in the trial. TRIAL REGISTRATION: Trial registration number: NCT04419493, 2019-004932-40 (EudraCT Number).

Fabrication of a Low-Cost, Fiber-Coupled, and Air-Spaced Fabry-Pérot Etalon.

Fabry-Pérot etalons (FPE) have found their way into many applications. In fields such as spectroscopy, telecommunications, and astronomy, FPEs are used for their high sensitivity as well as their exceptional filtering capability. However, air-spaced etalons with high finesse are usually built by specialized facilities. Their production requires a clean room, special glass handling, and coating machinery, meaning commercially available FPEs are sold for a high price. In this article, a new and cost-effective method to fabricate fiber-coupled FPEs with standard photonic laboratory equipment is presented. The protocol should serve as a step-by-step guide for the construction and characterization of these FPEs. We hope this will enable researchers to conduct fast and cost-effective prototyping of FPEs for various fields of application. The FPE, as presented here, is used for spectroscopic applications. As shown in the representative results section via proof of principle measurements of water vapor in ambient air, this FPE has a finesse of 15, which is sufficient for the photothermal detection of trace concentrations of gases.

Emergency Department Pediatric Readiness and Short-term and Long-term Mortality Among Children Receiving Emergency Care.

Importance: Emergency departments (EDs) with high pediatric readiness (coordination, personnel, quality improvement, safety, policies, and equipment) are associated with lower mortality among children with critical illness and those admitted to trauma centers, but the benefit among children with more diverse clinical conditions is unknown. Objective: To evaluate the association between ED pediatric readiness, in-hospital mortality, and 1-year mortality among injured and medically ill children receiving emergency care in 11 states. Design, Setting, and Participants: This is a retrospective cohort study of children receiving emergency care at 983 EDs in 11 states from January 1, 2012, through December 31, 2017, with follow-up for a subset of children through December 31, 2018. Participants included children younger than 18 years admitted, transferred to another hospital, or dying in the ED, stratified by injury vs medical conditions. Data analysis was performed from November 1, 2021, through June 30, 2022. Exposure: ED pediatric readiness of the initial ED, measured through the weighted Pediatric Readiness Score (wPRS; range, 0-100) from the 2013 National Pediatric Readiness Project assessment. Main Outcomes and Measures: The primary outcome was in-hospital mortality, with a secondary outcome of time to death to 1 year among children in 6 states. Results: There were 796 937 children, including 90 963 (11.4%) in the injury cohort (mean [SD] age, 9.3 [5.8] years; median [IQR] age, 10 [4-15] years; 33 516 [36.8%] female; 1820 [2.0%] deaths) and 705 974 (88.6%) in the medical cohort (mean [SD] age, 5.8 [6.1] years; median [IQR] age, 3 [0-12] years; 329 829 [46.7%] female, 7688 [1.1%] deaths). Among the 983 EDs, the median (IQR) wPRS was 73 (59-87). Compared with EDs in the lowest quartile of ED readiness (quartile 1, wPRS of 0-58), initial care in a quartile 4 ED (wPRS of 88-100) was associated with 60% lower in-hospital mortality among injured children (adjusted odds ratio, 0.40; 95% CI, 0.26-0.60) and 76% lower mortality among medical children (adjusted odds ratio, 0.24; 95% CI, 0.17-0.34). Among 545 921 children followed to 1 year, the adjusted hazard ratio of death in quartile 4 EDs was 0.59 (95% CI, 0.42-0.84) for injured children and 0.34 (95% CI, 0.25-0.45) for medical children. If all EDs were in the highest quartile of pediatric readiness, an estimated 288 injury deaths (95% CI, 281-297 injury deaths) and 1154 medical deaths (95% CI, 1150-1159 medical deaths) may have been prevented. Conclusions and Relevance: These findings suggest that children with injuries and medical conditions treated in EDs with high pediatric readiness had lower mortality during hospitalization and to 1 year.

A Geospatial Evaluation of 9-1-1 Ambulance Transports for Children and Emergency Department Pediatric Readiness.

OBJECTIVE: Whether ambulance transport patterns are optimized to match children to high-readiness emergency departments (EDs) and the resulting effect on survival are unknown. We quantified the number of children transported by 9-1-1 emergency medical services (EMS) to high-readiness EDs, additional children within 30 minutes of a high-readiness ED, and the estimated effect on survival. METHODS: This was a cross-sectional study using data from the National EMS Information System for 5,461 EMS agencies in 28 states from 1/1/2012 through 12/31/2019, matched to the 2013 National Pediatric Readiness Project assessment of ED pediatric readiness. We performed a geospatial analysis of children 0 to 17 years requiring 9-1-1 EMS transport to acute care hospitals, including day-, time-, and traffic-adjusted estimates for driving times to all EDs within 30 minutes of the scene. We categorized receiving hospitals by quartile of ED pediatric readiness using the weighted Pediatric Readiness Score (wPRS, range 0-100) and defined a high-risk subgroup of children as a proxy for admission. We used published estimates for the survival benefit of high readiness EDs to estimate the number of lives saved. RESULTS: There were 808,536 children transported by EMS, of whom 253,541 (31.4%) were high-risk. Among the 2,261 receiving hospitals, the median wPRS was 70 (IQR 57-85, range 26-100) and the median number of receiving hospitals within 30 minutes was 4 per child (IQR 2-11, range 1 to 53). Among all children, 411,685 (50.9%) were taken to EDs in the highest quartile of pediatric readiness, and 180,547 (22.3%) children transported to lower readiness EDs were within 30 minutes of a high readiness ED. Findings were similar among high-risk children. Based on high-risk children, we estimated that 3,050 pediatric lives were saved by transport to high-readiness EDs and an additional 1,719 lives could have been saved by shifting transports to high readiness EDs within 30 minutes. CONCLUSIONS: Approximately half of children transported by EMS were taken to high-readiness EDs and an additional one quarter could have been transported to such an ED, with measurable effect on survival.

Cdc37 as a Co-chaperone to Hsp90.

The co-chaperone p50/Cdc37 is an important partner for Hsp90, assisting in molecular chaperone activities, particularly with regard to the regulation of protein kinases. Analysis of the structure of Hsp90-Cdc37-kinase complexes demonstrates the way in which Cdc37 interacts with and controls the folding of a large proportion of intracellular protein kinases. This co-chaperone thus stands at the hub of a multitude of intracellular signaling networks. Indeed, the influence of Cdc37 reaches beyond the housekeeping pathways of protein folding into the regulation of a wide range of cellular processes. This co-chaperone has attracted attention as a potential intermediate in carcinogenesis. Cdc37 is an attractive potential target in cancer due to (1) high expression in a number of tumor types and (2) control of multiple signaling pathways. These properties indicate (3) a potential for selectivity due to its elevated expression in malignant cells and (4) robustness, as the co-chaperone may control multiple growth signaling pathways and thus be less prone to evolution of resistance than less versatile oncoproteins. Cdc37 may also be involved in other aspects of pathophysiology and has been shown to be secreted in exosomes. Protein aggregation disorders have been linked to age-related declines in molecular chaperones and co-chaperones. Cdc37 also appears to be a potential agent in longevity due to its links to protein folding and autophagy, and it will be informative to study the role of Cdc37 maintenance/decline in aging organisms.

Responsibility Gaps and Black Box Healthcare AI: Shared Responsibilization as a Solution.

As sophisticated artificial intelligence software becomes more ubiquitously and more intimately integrated within domains of traditionally human endeavor, many are raising questions over how responsibility (be it moral, legal, or causal) can be understood for an AI's actions or influence on an outcome. So called "responsibility gaps" occur whenever there exists an apparent chasm in the ordinary attribution of moral blame or responsibility when an AI automates physical or cognitive labor otherwise performed by human beings and commits an error. Healthcare administration is an industry ripe for responsibility gaps produced by these kinds of AI. The moral stakes of healthcare are often life and death, and the demand for reducing clinical uncertainty while standardizing care incentivizes the development and integration of AI diagnosticians and prognosticators. In this paper, we argue that (1) responsibility gaps are generated by "black box" healthcare AI, (2) the presence of responsibility gaps (if unaddressed) creates serious moral problems, (3) a suitable solution is for relevant stakeholders to voluntarily responsibilize the gaps, taking on some moral responsibility for things they are not, strictly speaking, blameworthy for, and (4) should this solution be taken, black box healthcare AI will be permissible in the provision of healthcare.

A Call for Behavioral Science in Embedded Bioethics.

Bioethicists today are taking a greater role in the design and implementation of emerging technologies by "embedding" within the development teams and providing their direct guidance and recommendations. Ideally, these collaborations allow ethical considerations to be addressed in an active, iterative, and ongoing process through regular exchanges between ethicists and members of the technological development team. This article discusses a challenge to this embedded ethics approach-namely, that bioethical guidance, even if embraced by the development team in theory, is not easily actionable in situ. Many of the ethical problems at issue in emerging technologies are associated with preexisting structural, socioeconomic, and political factors, making compliance with ethical recommendations sometimes less a matter of choice and more a matter of feasibility. Moreover, incentive structures within these systemic factors maintain them against reform efforts. The authors recommend that embedded bioethicists utilize principles from behavioral science (such as behavioral economics) to better understand and account for these incentive structures so as to encourage the ethically responsible uptake of technological innovations.