Biased Estimates Research Articles

Presence-only fisheries bycatch data produce biased species distribution predictions for Alcyonacean corals on British Columbia’s continental shelf and slope

Sedentary benthic species such as Alcyonacea corals form critical habitat for fishes and invertebrates. Assessing anthropogenic risks to these organisms requires unbiased, species distribution models (SDMs) that attempt to map probabilities of coral presence in relation to bio-physical ocean characteristics; however, in deep-water settings, the accuracy of SDMs is highly variable and dependent on spatial and taxonomic resolution. Here we investigated how data and model types affect SDM predictions of Alcyonacea probability of presence. We compared predictions from generalized additive models (GAMs) fitted to presence-absence observations over a stratified-random survey design with predictions from maximum entropy models (Maxent) fitted to presence-only bycatch records from commercial fisheries. We use a simulation analysis to show that both model types (i.e., GAM or Maxent) using presence-only bycatch data produced biased estimates of species distribution. Additionally, the maxent model fit using presence-only bycatch data produced biased estimates of performance metrics (AUC, TSS) that were overly optimistic. This study demonstrates a need for presence-absence data collected using a robust survey sampling design to fit SDMs that will better inform marine protected area placement while minimizing unnecessary economic losses, such as forgone fishing yield, due to sub-optimal placement.

Estimating Causal Effects on a Disease Progression Trait Using Bivariate Mendelian Randomisation.

Genome-wide association studies (GWAS) have provided large numbers of genetic markers that can be used as instrumental variables in a Mendelian Randomisation (MR) analysis to assess the causal effect of a risk factor on an outcome. An extension of MR analysis, multivariable MR, has been proposed to handle multiple risk factors. However, adjusting or stratifying the outcome on a variable that is associated with it may induce collider bias. For an outcome that represents progression of a disease, conditioning by selecting only the cases may cause a biased MR estimation of the causal effect of the risk factor of interest on the progression outcome. Recently, we developed instrument effect regression and corrected weighted least squares (CWLS) to adjust for collider bias in observational associations. In this paper, we highlight the importance of adjusting for collider bias in MR with a risk factor of interest and disease progression as the outcome. A generalised version of the instrument effect regression and CWLS adjustment is proposed based on a multivariable MR model. We highlight the assumptions required for this approach and demonstrate its utility for bias reduction. We give an illustrative application to the effect of smoking initiation and smoking cessation on Crohn's disease prognosis, finding no evidence to support a causal effect.

Shared parameter modeling of longitudinal data allowing for possibly informative visiting process and terminal event.

Joint modeling of longitudinal and time-to-event data, particularly through shared parameter models (SPMs), is a common approach for handling longitudinal marker data with an informative terminal event. A critical but often neglected assumption in this context is that the visiting/observation process is noninformative, depending solely on past marker values and visit times. When this assumption fails, the visiting process becomes informative, resulting potentially to biased SPM estimates. Existing methods generally rely on a conditional independence assumption, positing that the marker model, visiting process, and time-to-event model are independent given shared or correlated random effects. Moreover, they are typically built on an intensity-based visiting process using calendar time. This study introduces a unified approach for jointly modeling a normally distributed marker, the visiting process, and time-to-event data in the form of competing risks. Our model conditions on the history of observed marker values, prior visit times, the marker's random effects, and possibly a frailty term independent of the random effects. While our approach aligns with the shared-parameter framework, it does not presume conditional independence between the processes. Additionally, the visiting process can be defined on either a gap time scale, via proportional hazard models, or a calendar time scale, via proportional intensity models. Through extensive simulation studies, we assess the performance of our proposed methodology. We demonstrate that disregarding an informative visiting process can yield significantly biased marker estimates. However, misspecification of the visiting process can also lead to biased estimates. The gap time formulation exhibits greater robustness compared to the intensity-based model when the visiting process is misspecified. In general, enriching the visiting process with prior visit history enhances performance. We further apply our methodology to real longitudinal data from HIV, where visit frequency varies substantially among individuals.

Multi-digit touch II: perceptual weighting during integration follows a leading-finger priority.

When we run our hand across a surface, each finger typically repeats the sensory stimulation that the leading finger has already experienced. Because of this redundancy, the leading finger may attract more attention and contribute more strongly when tactile signals are integrated across fingers to form an overall percept. To test this hypothesis, we re-analysed data collected in a previous study (Arsanova et al., 1), where two probes were moved in different directions on two different fingerpads and participants reported the probes' average direction. Here, we evaluate the relative contribution of each finger to the percept and examine whether multi-digit integration gives priority to the leading finger. Although the hand actually remained static in these experiments, a 'functional leading finger' could be defined with reference to the average direction of the stimuli, and the direction of hand-object relative motion that this implied. When participants averaged the motion direction across fingers of the same hand, the leading finger received a higher weighting than the non-leading finger, even though this biased estimates of average direction. Importantly, this bias disappeared when averaging motion direction across the two hands. Both the reported average direction and its systematic relation to the difference between the individual stimulus directions were explained by a model of motion integration in which the sensory weighting of stimuli depends on the directions of the applied stimuli. Our finding supports the hypothesis that the leading finger, which often receives novel information in natural hand-object interactions, is prioritized in forming our tactile perception.

Earthquake Magnitude Correlations Expose Short-Term Catalog Incompleteness

Abstract The independence of earthquake magnitudes is a fundamental assumption and limitation in earthquake forecasting. To assess its validity, we examine correlations between the magnitude of successive earthquakes. We first investigate the 2019 Ridgecrest foreshock sequence and find a significant magnitude correlation as well as an unusually low Gutenberg–Richter b-value (≈0.7 using moment magnitudes). We demonstrate that these anomalous features are not indicative of a precursory phase, but a consequence of short-term incompleteness (STI) that is not detected by conventional methods to estimate catalog completeness. Synthetic simulations of this sequence support this explanation: imposing STI leads to significant magnitude correlation and biased b-value estimates. Expanding our investigation to seismicity across southern California reveals pervasive magnitude correlation due to STI, not limited to sequences with large earthquakes. Our findings suggest that magnitude correlation is the most effective indicator of STI, rather than a characteristic of the underlying earthquake-generating process.

Parental Expectation, Attitudes, and Home Numeracy Environment in Korea and in the U.S.: Potential Sources of Asian Math Advantages

This present study examined relations among parental math attitudes, expectations, and practice, and preschool children’s math achievement (i.e., parental math practice as a mediator) that may differ between Korean and U.S. samples. We examined measurement invariance to minimize the bias and inaccurate estimates in scores in two samples, which is a common barrier in cross-cultural studies. The Korean sample comprised 232 children (mean age = 54.58 months) and their parents from large urban cities and two other provinces. The U.S. sample included 146 preschool children (mean age = 52.49 months) and their parents in an urban area. We utilized measurement invariance to investigate whether the parental math practice has the same meaning across the two samples, along with mediation group comparisons. U.S. parents tended to have more positive math attitudes and higher expectations, while Korean parents were likely to be more involved in math practice with children at home. Korean children scored significantly higher on math achievement than the U.S. children. Among the Korean sample, higher parental math expectations were associated with higher levels of parental math practice, which was, in turn, associated with higher levels of child math outcomes after controlling for child age and SES. On the contrary, parental math expectations were unrelated to their math practice or child math outcomes in the U.S. sample. Parent math attitudes were not associated with any variables in both samples. These findings explain Asian math advantages that emerge early and offer insights into cultural processes (i.e., the importance of parental math practice) that may play a different role in children’s math outcomes.

A new Bayesian method for estimation of value at risk and conditional value at risk

AbstractValue at Risk (VaR) and Conditional Value at Risk (CVaR) have become the most popular measures of market risk in Financial and Insurance fields. However, the estimation of both risk measures is challenging, because it requires the knowledge of the tail of the distribution. Therefore, Extreme Value Theory initially seemed to be one of the best tools for this kind of problems, because using peaks-over-threshold method, we can assume the tail data approximately follow a Generalized Pareto distribution (GPD). The main objection to its use is that it only employs observations over the threshold, which are usually scarce. With the aim of improving the inference process, we propose a new Bayesian method that computes estimates built with all the information available. Informative prior Bayesian (IPB) method employs the existing relations between the parameters of the loss distribution and the parameters of the GPD that models the tail data to define informative priors in order to perform Metropolis–Hastings algorithm. We show how to apply IPB when the distribution of the observations is Exponential, stable or Gamma, to make inference and predictions. .Afterwards, we perform a thorough simulation study to compare the accuracy and precision of the estimates computed by IPB and the most employed methods to estimate VaR and CVaR. Results show that IPB provides the most accurate, precise and least biased estimates, especially when there are very few tail data. Finally, data from two real examples are analysed to show the practical application of the method.

Robust Regression as a Sensible Alternative to the Weighted Ordinary Least Squares Regression in case of Heteroskedasticity. A Tutorial

Background: The robust regression is rarely used in the statistical analyses in comparison with the Ordinary Least Squares regression and the Weighted Regression. In addition, in the frequent case of the heteroskedasticity of the residuals, a weighted regression carried out once is the main suggestion of the statistical books and the resulting reduced heteroscedasticity is usually considered sufficiently satisfactory. Methods: We showed the OLS regression analysis on data simulated with a well evident heteroskedasticity and an ad hoc outlier, followed by a weighted regression iteratively carried out by using iteratively reweighted least squares, an estimation method used also in several procedures of the robust regression analysis. Therefore, the link between the iteratively performed weighted regression and the robust regression becomes immediate. Furthermore, the same data have been analysed using some robust regression procedures.Results: It has been shown that in a simulated sample of heteroscedastic data with and without an obvious artificially created outlier the weighted regression performs worse with more biased parameter estimates than robust regression procedures (such as the robust MO procedure) as the presence of the outlier is not adequately neutralized. Discussion: In presence of a heteroskedastic pattern of the residuals, the suggestion to use robust regression procedures which can also deal with the almost sure presence of outliers seems more sensible. Among the robust regression procedures carried out, the performance of the robust MO procedure appears particularly appealing since it allows biostatisticians a more reasoned management of the outliers shown in a very illustrative “ad hoc” plot. Robust regression procedures represent a sensible alternative to OLS regression taking into account that its assumptions are practically not always fulfilled and that outliers, which are almost certainly present, are not only difficult to handle in classical OLS regression but can also provide highly biased estimates.

MCMC stopping rules in latent variable modelling.

Bayesian analysis relies heavily on the Markov chain Monte Carlo (MCMC) algorithm to obtain random samples from posterior distributions. In this study, we compare the performance of MCMC stopping rules and provide a guideline for determining the termination point of the MCMC algorithm in latent variable models. In simulation studies, we examine the performance of four different MCMC stopping rules: potential scale reduction factor (PSRF), fixed-width stopping rule, Geweke's diagnostic, and effective sample size. Specifically, we evaluate these stopping rules in the context of the DINA model and the bifactor item response theory model, two commonly used latent variable models in educational and psychological measurement. Our simulation study findings suggest that single-chain approaches outperform multiple-chain approaches in terms of item parameter accuracy. However, when it comes to person parameter estimates, the effect of stopping rules diminishes. We caution against relying solely on the univariate PSRF, which is the most popular method, as it may terminate the algorithm prematurely and produce biased item parameter estimates if the cut-off value is not chosen carefully. Our research offers guidance to practitioners on choosing suitable stopping rules to improve the precision of the MCMC algorithm in models involving latent variables.

County-level racial bias is associated with worse care for white and especially black older US adults: a cross-sectional observational study

ObjectiveTo assess the association of county-level bias about black and white people with patient experience, influenza immunisation, and quality of clinical care for black and white older US adults (age...

New Insight into the Role of Vitamin D in the Stroke Risk: A Meta-Analysis of Stratified Data by 25(OH)D Levels.

Mendelian Randomization (MR) studies have emerged as a powerful tool for investigating causal relationships between modifiable risk factors and clinical outcomes, using genetic variants as instrumental variables. In the context of vitamin D research, MR is a promising approach to elucidate the effects of vitamin D on various health outcomes, including adverse cardiovascular events. However, the validity of MR analyses relies heavily on the strength of the genetic associations found. "Weak instrument bias", arising from instruments with low explanatory power for the exposure of interest, can lead to biased estimates and compromise causal inference. We have, herein, briefly reviewed the challenges posed by weak instrument bias in a large MR study on vitamin D [25(OH)D] and stroke, exploring implications for the study's validity and reliability of findings. We have then added an original meta-analysis stratified by 25(OH)D levels. By using aggregated data from a recent MR study, an original meta-analysis stratified by population mean levels of 25(OH)D has indicated that interventions based on vitamin D supplementations in population mean levels ranging from 50 to 70 nmol/L are likely to translate into a 13% reduction of stroke risk (pooled odds ratio=0.873, 95% CI: 0.764-0.997, p-value=0.04). MR studies are a valuable approach for discerning causal relationships between exposures, such as vitamin D, and health outcomes. However, the effectiveness of MR analyses depends on the robustness of the genetic instruments employed. By recognizing and addressing weak instrument bias in MR studies of vitamin D, researchers can enhance the credibility and utility of causal inference in understanding the health effects of this essential nutrient. A metaanalysis stratified by population mean levels of 25(OH)D has revealed the potential benefits of targeted interventions with vitamin D supplementations for stroke.

Derivation of outcome-dependent dietary patterns for low-income women obtained from survey data using a supervised weighted overfitted latent class analysis.

Poor diet quality is a key modifiable risk factor for hypertension and disproportionately impacts low-income women. Analyzing diet-driven hypertensive outcomes in this demographic is challenging due to the complexity of dietary data and selection bias when the data come from surveys, a main data source for understanding diet-disease relationships in understudied populations. Supervised Bayesian model-based clustering methods summarize dietary data into latent patterns that holistically capture relationships among foods and a known health outcome but do not sufficiently account for complex survey design. This leads to biased estimation and inference and lack of generalizability of the patterns. To address this, we propose a supervised weighted overfitted latent class analysis (SWOLCA) based on a Bayesian pseudo-likelihood approach that integrates sampling weights into an exposure-outcome model for discrete data. Our model adjusts for stratification, clustering, and informative sampling, and handles modifying effects via interaction terms within a Markov chain Monte Carlo Gibbs sampling algorithm. Simulation studies confirm that the SWOLCA model exhibits good performance in terms of bias, precision, and coverage. Using data from the National Health and Nutrition Examination Survey (2015-2018), we demonstrate the utility of our model by characterizing dietary patterns associated with hypertensive outcomes among low-income women in the United States.

A-349 Evaluation of i-STAT® Point of Care Blood Gas Cartridges and Competitor Blood Gas Devices Against Reference Standard for PCO2 and PO2

Abstract Background The i-STAT System provides laboratory quality results in minutes at the patient’s bedside. Accurate and rapid test results are critical for clinical decision making in the presence of blood gas disorders, including oxygenation and acid-base status, where the partial pressure oxygen (PO2) and partial pressure carbon dioxide (PCO2) are needed. The purpose of this study was to compare the analytical performance of the PO2 and PCO2 tests in the i-STAT G3+ and i-STAT CG8+ cartridges to the theoretical PO2 and PCO2 in prepared reference standards. Two other blood gas instruments, a benchtop device and a laboratory device were also compared to the reference PO2 and PCO2. Methods Venous whole blood samples were contrived using saturation tonometry with U.S. NIST (National Institute of Science and Technology) traceable gas tanks to prepare the reference standards, which were value assigned to theoretical PO2 or PCO2 levels based on the molar composition of the gas mixture used. Eleven levels spanning the reportable range of each PO2 (5 mmHg - 800 mmHg) and PCO2 (5 mmHg - 130 mmHg) were prepared and tested in duplicate on the i-STAT G3+ and i-STAT CG8+ cartridges, benchtop device, and the laboratory device. Passing-Bablok linear regression analysis was performed to evaluate the slope and correlation coefficient, comparing each blood gas device against the reference PO2 and PCO2 values. Study designs followed CLSI (Clinical and Laboratory Standards Institute) EP09C-ED3:2018, Measurement Procedure Comparison and Bias Estimation using Patient Samples, 3rd Edition. Passing-Bablok regression analysis was also performed for the i-STAT cartridges and benchtop device against the laboratory device. Results The regression analysis was performed against the reference standards. For PO2, slopes for the i-STAT cartridges ranged from 0.93 - 0.97, the slope for the benchtop device was 0.99, and the slope for the laboratory device was 0.97. Correlation coefficients for all devices were 1.00. For PCO2, slopes for the i-STAT cartridges ranged from 0.98 - 1.02, the slope for the benchtop device was 0.85, and the slope for the laboratory device was 1.01. Correlation coefficients were 1.00 for the i-STAT cartridges and laboratory device, and 0.99 for the benchtop device. The regression analysis was also performed against the laboratory device. For PO2, slopes for the i-STAT cartridges ranged from 0.96 - 1.00, and the slope for the benchtop device was 1.02. Correlation coefficients for all devices were 1.00. For PCO2, slopes for the i-STAT cartridges ranged from 0.97 - 1.01, and the slope for the benchtop device was 0.85. Correlation coefficients were 1.00 for the i-STAT cartridges and 0.99 for the benchtop device. Conclusions The study demonstrated that the i-STAT G3+ and i-STAT CG8+ cartridges used with the i-STAT System were shown to provide laboratory quality results within 2 minutes, showing good agreement to both reference standards and laboratory quality devices. For PCO2, the i-STAT System was shown to have better agreement to both theoretical reference standards and the laboratory device than the benchtop device. These studies were funded by Abbott Laboratories.

B-008 An Evaluation of the Analytical Performance of the New Beckman Coulter DxC 500i Clinical Analyzer

Abstract Background The DxC 500i Clinical Analyzer* is Beckman Coulter’s latest integrated Clinical Chemistry and Immunoassay analyzer. It provides reliability, scalability, and workflow efficiency via a flexible independent maintenance mode, increased uptime, and a dynamic sample handler to optimize sample management. Standard Beckman Coulter Chemistry and Immunoassay consumables and reagents ensure network standardization and commutability. The purpose of this study was to evaluate the analytical performance of the new DxC 500i analyzer and to compare the performance against the Beckman Coulter stand-alone Clinical Chemistry and Immunoassay analyzers on the market. Methods To assess performance over a range of assay methodologies, several DxC 500i Beckman Coulter Immunoassays and Chemistry assays were evaluated. Precision was assessed using human samples following CLSI EP05-A3 guidelines. Linear range was assessed following CLSI EP06-A guidelines. Method Comparison and Bias Estimation studies compared the DxC 500i Clinical Analyzer against the Access 2 Immunoassay System, the AU480 and the DxC 700 AU Clinical Chemistry analyzers, following CLSI EP09c-ED3 guidelines. To investigate carryover potential as samples are processed across the integrated system, a carryover study was designed. This focused on evaluating potential sample-to-sample carryover caused by the AU sample probe and its effects on immunoassay samples during reflex scenarios. Results DxC 500i Chemistry results: Glucose Serum (OSR6x21) Assessments of repeatability and within laboratory precision at multiple critical analyte concentrations showed total imprecision of <3% and within-run imprecision of <3%. The Glucose serum assay demonstrated acceptable linearity throughout the 10-800 mg/dL analytical measuring range. Method comparison studies compared the DxC 500i analyzer against the AU480 and DxC 700 AU systems. A Weighted Deming regression of serum patient samples (n>100 measured across the analytical range) yielded a 0.986 slope, 0.227 mg/dL intercept, and R=0.9999 correlation coefficient against the DxC 700 AU and a 0.995 slope, 0.356 mg/dL intercept, and R=0.9999 correlation coefficient against the AU480. DxC 500i Immunoassay results: Cortisol (33600) Beckman Coutler’s Cortisol assay exhibited total imprecision of <12%CV at <5 μg/dL or <10CV% at ≥5 μg/dL for serum. The Cortisol assay demonstrated acceptable linearity throughout the 0.4-60 μg/dL analytical measuring range. A method comparison study compared the DxC 500i against the Access 2 system. A Weighted Deming regression of serum patient samples (n>100 measured across the analytical range) yielded a 0.9785 slope, 0.2764 μg/dL intercept, and R=0.9932 correlation coefficient. DxC 500i Carryover Assessment The Total βhCG (5th IS) assay was used to evaluate the potential of chemistry analyzer to immunoassay analyzer sample carryover. The study used High hCG serum samples (>1,000,000 mIU/mL) and Low hCG serum samples (4-6 mIU/mL). The DxC 500i exhibited a cumulative carryover of ≤2 parts per million (PPM) for serum. Conclusions These studies demonstrate excellent analytical performance for the new Beckman Coulter DxC 500i Clinical Analyzer* and confirm comparable performance to current Beckman Coulter standalone Clinical Chemistry and Immunoassay analyzers on the market. *Pending clearance by the United States Food and Drug Administration; not yet available for in vitro diagnostic use in the U.S. Product availability and regulatory status depends on country registration per applicable regulations.

A-306 High Sensitivity Capability of an Anti-müllerian Hormone Assay on the Beckman Coulter DxI 9000 Immunoassay Analyzer*, a Next Step Toward Enabling Ovarian Failure Staging?

Abstract Background Measurement of anti-müllerian hormone (AMH) has proven value in the evaluation of ovarian reserve using assays such as the Beckman Coulter Access AMH assay. If improved sensitivity of AMH measurement can be achieved, enhanced clinical utility during the assessment of reproductive aging and ovarian response to enable staging of ovarian failure is possible (Practice Committee of the American Society of Reproductive Medicine 2020, Cappola et al. 2023). The Beckman Coulter DxI 9000 Immunoassay Analyzer* has been designed with new technology that can be leveraged to improve assay sensitivity. Such technological advancements include the new Lumi-Phos PRO chemiluminescent substrate which yields markedly increased signal-to-noise, a new luminometer with increased dynamic range, and improved low-volume pipetting capabilities. Data herein summarize results from analytical verification studies of the Access AMH assay on the DxI 9000 Immunoassay Analyzer. Results demonstrate an example of DxI 9000 technologies enabling sensitivity claims that are up to 20-fold improved. * The official name is DxI 9000 Access Immunoassay Analyzer. Methods A method comparison study was performed based on CLSI guideline EP09c, 3rd ed. to compare results for the Access AMH assay on DxI 9000 to the predicate device, the Access AMH assay on the Access 2 Immunoassay System. Within-laboratory precision was also evaluated following CLSI EP05-A3. Finally, performance at low analyte levels was evaluated following CLSI EP17-A2, and linearity was evaluated following CLSI EP06-Ed2. Results Method comparison of the Access AMH assay on DxI 9000 compared to the Access AMH assay on the Access 2 yielded excellent agreement with a Passing-Bablok slope of 1.02 for N=126 samples. Further, bias estimates for a number of key medical decision points suggest minimal bias (≤ 2%) across the assay range, and minimal non-linearity was observed. 20-day imprecision studies yielded CV’s between 2 and 5% across a broad range of concentrations evaluated. Sensitivity studies yielded estimates of limit of quantitation (LoQ) between 0.001 and 0.003 ng/mL. This represents an approximately 10- to 20-fold improvement in sensitivity claims compared to the AMH assay on the Access 2 and other commercially available automated immunoassay analyzers. Conclusions The data herein present performance data for the Access AMH assay on the DxI 9000 Immunoassay Analyzer. The assay demonstrates excellent accuracy relative to the predicate device, while showing good linearity and imprecision. Of note, the Access AMH assay exhibits exceptional low-end performance on the DxI 9000 analyzer as demonstrated by the ability to enable sensitivity claims that are up to 20-fold improved in comparison to existing immunoassays for the measurement of AMH. The sensitivity capabilities of the Access AMH assay on DxI 9000 suggest an opportunity for the addition of further intended uses for scenarios where extremely low AMH measurements can provide additional value of clinical significance. Further studies are needed to fully understand additional clinical utilities of the Access AMH assay with high sensitivity capability on DxI 9000, including the assessment of reproductive aging for menopause diagnosis or ovarian response for fertility treatment.

Firth-Type Penalized Methods of the Modified Poisson and Least-Squares Regression Analyses for Binary Outcomes.

The modified Poisson and least-squares regression analyses for binary outcomes have been widely used as effective multivariable analysis methods to provide risk ratio and risk difference estimates in clinical and epidemiological studies. However, there is no certain evidence that assessed their operating characteristics under small and sparse data settings and no effective methods have been proposed for these regression analyses to address this issue. In this article, we show that the modified Poisson regression provides seriously biased estimates under small and sparse data settings. In addition, the modified least-squares regression provides unbiased estimates under these settings. We further show that the ordinary robust variance estimators for both of the methods have certain biases under situations that involve small or moderate sample sizes. To address these issues, we propose the Firth-type penalized methods for the modified Poisson and least-squares regressions. The adjustment methods lead to a more accurate and stable risk ratio estimator under small and sparse data settings, although the risk difference estimator is not invariant. In addition, to improve the inferences of the effect measures, we provide an improved robust variance estimator for these regression analyses. We conducted extensive simulation studies to assess the performances of the proposed methods under real-world conditions and found that the accuracies of the point and interval estimations were markedly improved by the proposed methods. We illustrate the effectiveness of these methods by applying them to a clinical study of epilepsy.

Publication bias in pharmacogenetics of statin-associated muscle symptoms, an umbrella review with a meta-epidemiological study

Background and aimsStatin-associated muscle symptoms (SAMS) are a major cause of treatment discontinuation. Clinical Pharmacogenetics Implementation Consortium (CPIC) recommend dose adjustment for statin treatment according to known SLCO1B1 genotype to reduce SAMS. We hypothesized that the association between SLCO1B1 genotype and SAMS is misestimated because of publication bias. MethodsWe searched published systematic reviews on the association between SLCO1B1 genotype and SAMS. To assessed publication bias, we used funnel plot visual inspection, Egger’s test, and the Bayes Factor (BFPublication-bias) from Robust Bayesian Meta-Analysis (RoBMA). We compared the odds ratios (ORUncorrected) from meta-analyses before and after correcting for publication bias using trim-and-fill (ORTrim&Fill) and RoBMA (ORRoBMA) methods. ResultsWe included 8 cohort and 11 case-control studies, totaling 62 OR of three SLCO1B1 genotypes and six statin drugs. In the primary analysis, the funnel plot was suggestive of publication bias, confirmed by Egger’s test (p=0.001) and RoBMA (BFPublication-bias=18). Correcting the estimate for publication bias resulted in loss of the association, from a significant ORUncorrected (1.31 95%CI [1.13–1.53]) to corrected ORs suggesting no difference: ORTrim&Fill (1.07 95%CI [0.89–1.30]) and ORRoBMA (1.02 95%CI [1.00–1.33]). This suggested that publication bias overestimated the association by 18% and 23%, respectively. Similar results were found for genotype rs4149056, simvastatin and atorvastatin. ConclusionsThe effect of the SLCO1B1 genotype on the risk of developing SAMS is overestimated in the published literature, especially rs4149056. This could lead prescribers to incorrectly decreasing statin doses or even avoiding statin use, leading to a loss of the potential cardiovascular benefit of statins.

Deployment of Public Charging Stations for BEVs Using an Agent-Based Modeling Approach

A low utilization rate of public chargers and unmatched deployment of public charging stations (CSs) are partly attributed to inappropriate modeling of charging behavior and biased charging demand estimation. This study proposes an optimization methodology for public CS deployment, considering real charging behavior and interactions between battery electric vehicle (BEV) users and CSs. Realistic charging choice behavior is modeled based on surveys, and a dynamic charging decision chain is simulated, allowing interactions between BEV users and CSs through an agent-based modeling (ABM) approach. The charging-related activities are triggered by state of charge (SOC) levels randomly generated from distributions derived from real BEV operating data, including the random SOC levels at the start of a trip, the SOC level that prompts the user to charge the BEV, and the SOC level at which the user stops charging the BEV. A bi-level programming model is proposed to optimize the deployment schemes for building new CSs considering the existing CSs, to determine the location and the capacity of new CSs. The objective is to minimize the total time cost per BEV user, including travel time, charging time and waiting time in the queue. An application is conducted, for the deployment of fast CSs in Washington State, USA. The results show that our method could provide effective guidance for allocating new CSs that are good supplements to the existing heavy-load CSs to share their charging load and relieve their serious queuing problems. The optimized deployment scheme can efficiently alleviate long waiting times at existing CSs and leads to a more balanced utilization among CSs. The proposed approach is expected to contribute to better planning and deployment of public CSs, satisfaction of the booming charging demand and increased utilization of public CSs.

Sub‐Pixel Displacement Estimation With Deep Learning: Application to Optical Satellite Images Containing Sharp Displacements

AbstractOptical image correlation is a powerful method for remotely constraining ground movement from optical satellite imagery related to natural disasters (e.g., earthquakes, volcanoes, landslides). This approach enables the characterization, and identification of the causal factors and mechanisms underlying such processes. By employing sub‐pixel correlation algorithms, one can obtain highly accurate (m‐to‐cm level) displacement fields at high spatial resolution (dm‐to‐cm) by comparing satellite images acquired before and after a period of movement. However, this method generally assumes a homogeneous translation of all pixels within a given correlation window, which will lead to biased estimates of ground displacement if the real case is not well represented by such a simplification, especially when resolving ground displacements next to sharp gradients in displacement, such as those found in the near‐field of earthquake surface ruptures. In this paper, we present an innovative deep learning method estimating sub‐pixel displacement maps from optical satellite images for the retrieval of ground displacement. From the generation of a realistic simulated database, comprising Landsat‐8 satellite image pairs containing simulated sub‐pixel shifts and sharp discontinuities, we develop a Convolutional Neural Network able to retrieve sub‐pixel displacements. The comparison to state‐of‐the‐art correlation methods shows that our pipeline significantly reduces by 32% the estimation bias around fault ruptures, leading to more accurate characterization of the near‐field strain in surface rupturing earthquakes. Application of our model to the 2019 Ridgecrest earthquake demonstrates the ability of our model to accurately and quickly resolve ground displacement using real satellite images. Code is made available at https://gricad‐gitlab.univ‐grenoble‐alpes.fr/montagtr/cnn4l‐discontinuities.

Efficient private SCO for heavy-tailed data via averaged clipping

AbstractWe consider stochastic convex optimization for heavy-tailed data with the guarantee of being differentially private (DP). Most prior works on differentially private stochastic convex optimization for heavy-tailed data are either restricted to gradient descent (GD) or performed multi-times clipping on stochastic gradient descent (SGD), which is inefficient for large-scale problems. In this paper, we consider a one-time clipping strategy and provide principled analyses of its bias and private mean estimation. We establish new convergence results and improved complexity bounds for the proposed algorithm called AClipped-dpSGD for constrained and unconstrained convex problems. We also extend our convergent analysis to the strongly convex case and non-smooth case (which works for generalized smooth objectives with H$$\ddot{\text {o}}$$ o ¨ lder-continuous gradients). All the above results are guaranteed with a high probability for heavy-tailed data. Numerical experiments are conducted to justify the theoretical improvement.