Clinging to the top: natal dispersal tracks climate gradient in a trailing-edge population of a migratory songbird

PURPOSE

Trailing-edge populations at the low-latitude, receding edge of a shifting range face high extinction risk from climate change unless they are able to track optimal environmental conditions through dispersal.

METHODS

We fit dispersal models to the locations of 3165 individually-marked black-throated blue warblers (Setophaga caerulescens) in the southern Appalachian Mountains in North Carolina, USA from 2002 to 2023. Black-throated blue warbler breeding abundance in this population has remained relatively stable at colder and wetter areas at higher elevations but has declined at warmer and drier areas at lower elevations.

RESULTS

Median dispersal distance of young warblers was 917 m (range 23-3200 m), and dispersal tended to be directed away from warm and dry locations. In contrast, adults exhibited strong site fidelity between breeding seasons and rarely dispersed more than 100 m (range 10-1300 m). Consequently, adult dispersal kernels were much more compact and symmetric than natal dispersal kernels, suggesting adult dispersal is unlikely a driving force of declines in this population.

CONCLUSION

Our findings suggest that directional natal dispersal may mitigate fitness costs for trailing-edge populations by allowing individuals to track changing climate and avoid warming conditions at warm-edge range boundaries.

A review towards developing a hierarchical model for sustainable hospital solid waste management in rural areas of Zimbabwe

Management of solid waste from rural hospitals is amongst problems affecting Zimbabwe due to diseases, population, and hospital increase. Solid waste from rural hospitals is receiving little attention translating to environmental health problems. Therefore, 101 secondary sources were used to write a paper aiming to proffer a hierarchical model to achieve sustainable solid waste management at rural hospitals. Rural hospitals' solid waste encompasses electronic waste, sharps, pharmaceutical, pathological, radioactive, chemical, infectious, and general waste. General solid waste from rural hospitals is between 77.35 and 79% whilst hazardous waste is between 21 and 22.65%. Solid waste increase add burden to nearly incapacitated rural hospitals. Rural hospital solid waste management processes include storage, transportation, treatment methods like autoclaving and chlorination, waste reduction alternatives, and disposal. Disposal strategies involve open pits, open burning, dumping, and incineration. Rural hospital solid waste management is guided by legislation, policies, guidelines, and conventions. Effectiveness of legal framework is limited by economic and socio-political problems. Rural hospital solid waste management remain inappropriate causing environmental health risks. Developed hierarchical model can narrow the route to attain sustainable management of rural hospitals' solid waste. Proposed hierarchical model consists of five-layered strategies and acted as a guide for identifying and ranking approaches to manage rural hospitals' solid waste. Additionally, Zimbabwean government, Environmental Management Agency and Ministry of Health is recommended to collaborate to provide sufficient resources to rural hospitals whilst enforcing legal framework. Integration of all hierarchical model's elements is essential whereas all-stakeholder involvement and solid waste minimisation approaches are significant at rural hospitals.

Minimal clinically important change of knee flexion in people with knee osteoarthritis after non-surgical interventions using a meta-analytical approach

BACKGROUND

Minimal clinically important change (MCIC) represents the minimum patient-perceived improvement in an outcome after treatment, in an individual or within a group over time. This study aimed to determine MCIC of knee flexion in people with knee OA after non-surgical interventions using a meta-analytical approach.

METHODS

Four databases (MEDLINE, Cochrane, Web of Science and CINAHL) were searched for studies of randomised clinical trials of non-surgical interventions with intervention duration of ≤ 3 months that reported change in (Δ) (mean change between baseline and immediately after the intervention) knee flexion with Δ pain or Δ function measured using tools that have established MCIC values. The risk of bias in the included studies was assessed using version 2 of the Cochrane risk-of-bias tool for randomised trials (RoB 2). Bayesian meta-analytic models were used to determine relationships between Δ flexion with Δ pain and Δ function after non-surgical interventions and MCIC of knee flexion.

RESULTS

Seventy-two studies (k = 72, n = 5174) were eligible. Meta-analyses included 140 intervention arms (k = 61, n = 4516) that reported Δ flexion with Δ pain using the visual analog scale (pain-VAS) and Δ function using the Western Ontario and McMaster Universities Osteoarthritis Index function subscale (function-WOMAC). Linear relationships between Δ pain at rest-VAS (0-100 mm) with Δ flexion were - 0.29 (- 0.44; - 0.15) (β: posterior median (CrI: credible interval)). Relationships between Δ pain during activity VAS and Δ flexion were - 0.29 (- 0.41, - 0.18), and Δ pain-general VAS and Δ flexion were - 0.33 (- 0.42, - 0.23). The relationship between Δ function-WOMAC (out of 100) and Δ flexion was - 0.15 (- 0.25, - 0.07). Increased Δ flexion was associated with decreased Δ pain-VAS and increased Δ function-WOMAC. The point estimates for MCIC of knee flexion ranged from 3.8 to 6.4°.

CONCLUSIONS

The estimated knee flexion MCIC values from this study are the first to be reported using a novel meta-analytical method. The novel meta-analytical method may be useful to estimate MCIC for other measures where anchor questions are problematic.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42022323927.

Executive and hierarchical models for participatory response to health emergencies in the workplace: Lessons from COVID-19

Introduction

Workplaces are high-risk environments for epidemic transmission, and the COVID-19 pandemic has highlighted the significant impacts that health emergencies can have on both the healthcare system and the economy. This study presents executive and hierarchical models for participatory response to health emergencies in the workplace, with a focus on COVID-19.

Methods

The study was conducted in three phases. Content analysis of interviews with 101 stakeholders and national documents was used to identify key themes and dimensions for an executive model. A focus group discussion and review of international documents were then used to refine and expand the executive and hierarchical models. The alignment and trustworthiness of the final models, as well as feedback, were gathered from 117 informants working in various workplaces.

Results

The executive model highlighted that context understanding, management commitment, and participation play critical roles in developing tailored prevention and response plans, and adequate support is necessary for successful plan implementation. Monitoring and review processes should be established to ensure proper functioning. The hierarchical model emphasizes the need for collaborative efforts from various stakeholders to effectively implement pandemic prevention and participatory response plans.

Conclusion

Overall, the executive and hierarchical participatory models presented in this study provide a framework for effectively controlling pandemics and other health emergencies in the workplace, enhancing both health resilience and the sustainability of economic activities.

Multi-level color classification of post-consumer plastic packaging flakes by hyperspectral imaging for optimizing the recycling process

In a circular economy perspective, the development of fast and efficient sensor-based recognition strategies of plastic waste, not only by polymer but also by color, plays a crucial role for the production of high quality secondary raw materials in recycling plants. In this work, mixed colored flakes of high-density polyethylene (HDPE) from packaging waste were simultaneously classified by hyperspectral imaging working in the visible range (400-750 nm), combined with machine learning. Two classification models were built and compared: (1) Partial Least Square-Discriminant Analysis (PLS-DA) for 6 HDPE macro-color classes identification (i.e., white, blue, green, red, orange and yellow) and (2) hierarchical PLS-DA for a more accurate discrimination of the different HDPE color tones, providing as output 14 color classes. The obtained classification results were excellent for both models, with values of Recall, Specificity, Accuracy, and F-score in prediction close to 1. The proposed methodological approach can be utilized as sensor-based sorting logic in plastic recycling plants, tuning the output based on the required needs of the recycling plant, allowing to obtain a high-quality recycled HDPE of different colors, optimizing the plastic recycling process, in agreement with the principles of circular economy.

Accounting for clustering in automated variable selection using hospital data: a comparison of different LASSO approaches

BACKGROUND

Automated feature selection methods such as the Least Absolute Shrinkage and Selection Operator (LASSO) have recently gained importance in the prediction of quality-related outcomes as well as the risk-adjustment of quality indicators in healthcare. The methods that have been used so far, however, do not account for the fact that patient data are typically nested within hospitals.

METHODS

Therefore, we aimed to demonstrate how to account for the multilevel structure of hospital data with LASSO and compare the results of this procedure with a LASSO variant that ignores the multilevel structure of the data. We used three different data sets (from acute myocardial infarcation, COPD, and stroke patients) with two dependent variables (one numeric and one binary), on which different LASSO variants with and without consideration of the nested data structure were applied. Using a 20-fold sub-sampling procedure, we tested the predictive performance of the different LASSO variants and examined differences in variable importance.

RESULTS

For the metric dependent variable Duration Stay, we found that inserting hospitals led to better predictions, whereas for the binary variable Mortality, all methods performed equally well. However, in some instances, the variable importances differed greatly between the methods.

CONCLUSION

We showed that it is possible to take the multilevel structure of data into account in automated predictor selection and that this leads, at least partly, to better predictive performance. From the perspective of variable importance, including the multilevel structure is crucial to select predictors in an unbiased way under consideration of the structural differences between hospitals.

Bayesian multilevel multivariate logistic regression for superiority decision-making under observable treatment heterogeneity

BACKGROUND

In medical, social, and behavioral research we often encounter datasets with a multilevel structure and multiple correlated dependent variables. These data are frequently collected from a study population that distinguishes several subpopulations with different (i.e., heterogeneous) effects of an intervention. Despite the frequent occurrence of such data, methods to analyze them are less common and researchers often resort to either ignoring the multilevel and/or heterogeneous structure, analyzing only a single dependent variable, or a combination of these. These analysis strategies are suboptimal: Ignoring multilevel structures inflates Type I error rates, while neglecting the multivariate or heterogeneous structure masks detailed insights.

METHODS

To analyze such data comprehensively, the current paper presents a novel Bayesian multilevel multivariate logistic regression model. The clustered structure of multilevel data is taken into account, such that posterior inferences can be made with accurate error rates. Further, the model shares information between different subpopulations in the estimation of average and conditional average multivariate treatment effects. To facilitate interpretation, multivariate logistic regression parameters are transformed to posterior success probabilities and differences between them.

RESULTS

A numerical evaluation compared our framework to less comprehensive alternatives and highlighted the need to model the multilevel structure: Treatment comparisons based on the multilevel model had targeted Type I error rates, while single-level alternatives resulted in inflated Type I errors. Further, the multilevel model was more powerful than a single-level model when the number of clusters was higher. A re-analysis of the Third International Stroke Trial data illustrated how incorporating a multilevel structure, assessing treatment heterogeneity, and combining dependent variables contributed to an in-depth understanding of treatment effects. Further, we demonstrated how Bayes factors can aid in the selection of a suitable model.

CONCLUSION

The method is useful in prediction of treatment effects and decision-making within subpopulations from multiple clusters, while taking advantage of the size of the entire study sample and while properly incorporating the uncertainty in a principled probabilistic manner using the full posterior distribution.

An active inference model of hierarchical action understanding, learning and imitation

We advance a novel active inference model of the cognitive processing that underlies the acquisition of a hierarchical action repertoire and its use for observation, understanding and imitation. We illustrate the model in four simulations of a tennis learner who observes a teacher performing tennis shots, forms hierarchical representations of the observed actions, and imitates them. Our simulations show that the agent's oculomotor activity implements an active information sampling strategy that permits inferring the kinematic aspects of the observed movement, which lie at the lowest level of the action hierarchy. In turn, this low-level kinematic inference supports higher-level inferences about deeper aspects of the observed actions: proximal goals and intentions. Finally, the inferred action representations can steer imitative responses, but interfere with the execution of different actions. Our simulations show that hierarchical active inference provides a unified account of action observation, understanding, learning and imitation and helps explain the neurobiological underpinnings of visuomotor cognition, including the multiple routes for action understanding in the dorsal and ventral streams and mirror mechanisms.

Hierarchical inference as a source of human biases

The finding that human decision-making is systematically biased continues to have an immense impact on both research and policymaking. Prevailing views ascribe biases to limited computational resources, which require humans to resort to less costly resource-rational heuristics. Here, we propose that many biases in fact arise due to a computationally costly way of coping with uncertainty-namely, hierarchical inference-which by nature incorporates information that can seem irrelevant. We show how, in uncertain situations, Bayesian inference may avail of the environment's hierarchical structure to reduce uncertainty at the cost of introducing bias. We illustrate how this account can explain a range of familiar biases, focusing in detail on the halo effect and on the neglect of base rates. In each case, we show how a hierarchical-inference account takes the characterization of a bias beyond phenomenological description by revealing the computations and assumptions it might reflect. Furthermore, we highlight new predictions entailed by our account concerning factors that could mitigate or exacerbate bias, some of which have already garnered empirical support. We conclude that a hierarchical inference account may inform scientists and policy makers with a richer understanding of the adaptive and maladaptive aspects of human decision-making.

Time-specific hierarchical models for predicting crash frequency of reversible and high-occupancy vehicle lanes

Time-specific Safety Performance Functions (SPFs) were proposed to achieve accurate and dynamic crash frequency predictions. This study contributes to the literature by developing time-specific SPFs for freeways that include reversible lanes (RL) and freeways that include High-Occupancy Vehicle lanes (HOV) using Microwave Vehicle Detection System (MVDS) data from Virginia, Arizona and Washington States. Variables that capture the time-specific traffic turbulence were prepared and considered in the developed SPFs. Moreover, two different hierarchical models were proposed to identify factors associated with the different crash types or severity in crash frequency prediction. The results indicated that the variables representing the volume difference between reversible and general-purpose lanes (GPL) were positively associated with crash frequency. Further, the variable that indicated the design of the access point of the reversible lane was positively associated with crash frequency. The models comparison results showed that the hierarchical models outperformed the corresponding Poisson lognormal model with lower AIC and MAE values. This study also tested the proposed hierarchical models on High-Occupancy Vehicle freeway sections and reached the same conclusion on model comparison results. The significant variables representing the logarithm of volume were found to be significant and positive with crash frequency. Moreover, the difference in average speed between the HOV lanes and GPL was also found to be positive and significant with the crash frequency. In general, this study successfully identified the factors associated with the different crash types or severity in crash frequency prediction models.

Graph Neural Tree: A novel and interpretable deep learning-based framework for accurate molecular property predictions

Determining various properties of molecules is a critical step in drug discovery. Recently, with the improvement of large heterogeneous datasets and the development of deep learning approaches, more and more scientists have turned their attention to neural network-based virtual preliminary screening to reduce the time and monetary cost of drug discovery. However, the poor interpretability of deep learning masks causality, so models' conclusions are often beyond the comprehension of human users, which reduces the credibility of the model and makes it difficult for chemists to further narrow the huge chemical space based on models' results. Thus, this study develops a novel framework consisting of Graph Neural Networks for feature extraction, Curriculum-Based Learning Strategies for optimization, and a Learning Binary Neural Tree (LBNT) for prediction, to improve the performance of neural networks and reveal their decision-making process to chemists. The framework encodes molecular graph data with graph neural networks (GNNs), then retrains the encoder with curriculum-based learning strategies to reduce uncertainty and improve accuracy, and finally uses LBNT as the predictor, which joint retrains with the encoder after independently training, for prediction and visualization. The framework is validated on the public datasets and compared to single GNNs with normal training strategies as well as GNN encoders with common machine learning predictors instead of the LBNT predictor. The result reveals that the proposed framework enhances the point prediction accuracy of the completely trained GNN and reduces its uncertainty through curriculum-based learning, and further improves the accuracy by combining LBNT. Besides, compared with common machine learning tools, the LBNT predictor generally has the best performance because of joint retraining with the GNN encoder. The decision-making process of LBNT is also better and easier to explain than that of other models.

Hierarchical deep learning with Generative Adversarial Network for automatic cardiac diagnosis from ECG signals

Cardiac disease is the leading cause of death in the US. Accurate heart disease detection is critical to timely medical treatment to save patients' lives. Routine use of the electrocardiogram (ECG) is the most common method for physicians to assess the cardiac electrical activities and detect possible abnormal conditions. Fully utilizing the ECG data for reliable heart disease detection depends on developing effective analytical models. In this paper, we propose a two-level hierarchical deep learning framework with Generative Adversarial Network (GAN) for ECG signal analysis. The first-level model is composed of a Memory-Augmented Deep AutoEncoder with GAN (MadeGAN), which aims to differentiate abnormal signals from normal ECGs for anomaly detection. The second-level learning aims at robust multi-class classification for different arrhythmia identification, which is achieved by integrating the transfer learning technique to transfer knowledge from the first-level learning with the multi-branching architecture to handle the data-lacking and imbalanced data issues. We evaluate the performance of the proposed framework using real-world ECG data from the MIT-BIH arrhythmia database. Experimental results show that our proposed model outperforms existing methods that are commonly used in current practice.

Smaller p-values in genomics studies using distilled auxiliary information

Medical research institutions have generated massive amounts of biological data by genetically profiling hundreds of cancer cell lines. In parallel, academic biology labs have conducted genetic screens on small numbers of cancer cell lines under custom experimental conditions. In order to share information between these two approaches to scientific discovery, this article proposes a "frequentist assisted by Bayes" (FAB) procedure for hypothesis testing that allows auxiliary information from massive genomics datasets to increase the power of hypothesis tests in specialized studies. The exchange of information takes place through a novel probability model for multimodal genomics data, which distills auxiliary information pertaining to cancer cell lines and genes across a wide variety of experimental contexts. If the relevance of the auxiliary information to a given study is high, then the resulting FAB tests can be more powerful than the corresponding classical tests. If the relevance is low, then the FAB tests yield as many discoveries as the classical tests. Simulations and practical investigations demonstrate that the FAB testing procedure can increase the number of effects discovered in genomics studies while still maintaining strict control of type I error and false discovery rate.

Estimating vaccination coverage for routinely recommended vaccines among children aged 24 months and adolescents aged 13 through 17 years using data from immunization information systems in the United States

OBJECTIVE

To use a model-based approach to estimate vaccination coverage of routinely recommended childhood and adolescent vaccines for the United States.

METHODS

We used a hierarchical model with retrospective cohort data from eleven IIS jurisdictions, which contains vaccination records submitted by providers. Numerators included data from 2014 to 2019 at the county level for 2.4 million children at age 24 months and 14.4 million adolescents aged 13-17. Age-appropriate Census populations were used as denominators. Covariates associated with childhood and adolescent vaccinations were included in the model. Model-based estimates for each county were generated and aggregated to the national level to produce national vaccination coverage estimates and compared to National Immunization Survey (NIS) estimates of vaccination coverage. Trends of estimated vaccination coverage were compared between the model-based approach and NIS.

RESULTS

From 2014 to 18, model-based national vaccination coverage estimates were within ten percentage points of NIS-Child vaccination coverage estimates for most vaccines among children at age 24 months. One notable difference was higher model-based vaccination coverage estimates for hepatitis B birth dose compared to NIS-Child coverage estimates. From 2014 to 19, model-based national vaccination coverage estimates were within ten percentage points of NIS-Teen vaccination coverage estimates for most vaccines among adolescents aged 13-17 years. Model-based vaccination coverage estimates were notably lower for varicella, MMR, and Hepatitis B compared to NIS-Teen coverage estimates among adolescents. Trends in estimates of national vaccination coverage were similar between model-based estimates for children and adolescents as compared to NIS-Child and NIS-Teen, respectively.

CONCLUSIONS

A hierarchical model applied to data from IIS may be used to estimate coverage for routinely recommended vaccines among children and adolescents and allows for timely analyses of childhood and adolescent vaccines to quickly assess trends in vaccination coverage across the United States. Monitoring real-time vaccination coverage can help promote immunizations to protect children and adolescents against vaccine-preventable diseases.

A hierarchical machine learning model based on Glioblastoma patients' clinical, biomedical, and image data to analyze their treatment plans

AIM OF STUDY

Glioblastoma Multiforme (GBM) is an aggressive brain cancer in adults that kills most patients in the first year due to ineffective treatment. Different clinical, biomedical, and image data features are needed to analyze GBM, increasing complexities. Besides, they lead to weak performances for machine learning models due to ignoring physicians' knowledge. Therefore, this paper proposes a hierarchical model based on Fuzzy C-mean (FCM) clustering, Wrapper feature selection, and twelve classifiers to analyze treatment plans.

METHODOLOGY/APPROACH

The proposed method finds the effectiveness of previous and current treatment plans, hierarchically determining the best decision for future treatment plans for GBM patients using clinical data, biomedical data, and different image data. A case study is presented based on the Cancer Genome Atlas Glioblastoma Multiforme dataset to prove the effectiveness of the proposed model. This dataset is analyzed using data preprocessing, experts' knowledge, and a feature reduction method based on the Principal Component Analysis. Then, the FCM clustering method is utilized to reinforce classifier learning.

OUTCOMES OF STUDY

The proposed model finds the best combination of Wrapper feature selection and classifier for each cluster based on different measures, including accuracy, sensitivity, specificity, precision, F-score, and G-mean according to a hierarchical structure. It has the best performance among other reinforced classifiers. Besides, this model is compatible with real-world medical processes for GBM patients based on clinical, biomedical, and image data.

Hierarchical classification models and Handheld NIR spectrometer to human blood stains identification on different floor tiles

One of the most important types of evidence in certain criminal investigations is traces of human blood. For a detailed investigation, blood samples must be identified and collected at the crime scene. The present study aimed to evaluate the potential of the identification of human blood in stains deposited on different types of floor tiles (five types of ceramics and four types of porcelain tiles) using a portable NIR instrument. Hierarchical models were developed by combining multivariate analysis techniques capable of identifying traces of human blood (HB), animal blood (AB) and common false positives (CFP). The spectra of the dried stains were obtained using a portable MicroNIR spectrometer (Viavi). The hierarchical models used two decision rules, the first to separate CFP and the second to discriminate HB from AB. The first decision rule, used to separate the CFP, was based on the Q-Residual criterion considering a PCA model. For the second rule, used to discriminate HB and AB, the Q-Residual criterion were tested as obtained from a PCA model, a One-Class SIMCA model, and a PLS-DA model. The best results of sensitivity and specificity, both equal to 100%, were obtained when a PLS-DA model was employed as the second decision rule. The hierarchical classification models built for these same training sets using a PCA or SIMCA model also obtained excellent sensitivity results for HB classification, with values above 94% and 78% of specificity. No CFP samples were misclassified. Hierarchical models represent a significant advance as a methodology for the identification of human blood stains at crime scenes.

Unifying the analysis of continuous and categorical measures of weight loss and incorporating group effect: a secondary re-analysis of a large cluster randomized clinical trial using Bayesian approach

Background

Although frequentist paradigm has been the predominant approach to clinical studies for decades, some limitations associated with the frequentist null hypothesis significance testing have been recognized. Bayesian approaches can provide additional insights into data interpretation and inference by deriving posterior distributions of model parameters reflecting the clinical interest. In this article, we sought to demonstrate how Bayesian approaches can improve the data interpretation by reanalyzing the Rural Engagement in Primary Care for Optimizing Weight Reduction (REPOWER).

Methods

REPOWER is a cluster randomized clinical trial comparing three care delivery models: in-clinic individual visits, in-clinic group visits, and phone-based group visits. The primary endpoint was weight loss at 24 months and the secondary endpoints included the proportions of achieving 5 and 10% weight loss at 24 months. We reanalyzed the data using a three-level Bayesian hierarchical model. The posterior distributions of weight loss at 24 months for each arm were obtained using Hamiltonian Monte Carlo. We then estimated the probability of having a higher weight loss and the probability of having greater proportion achieving 5 and 10% weight loss between groups. Additionally, a four-level hierarchical model was used to assess the partially nested intervention group effect which was not investigated in the original REPOWER analyses.

Results

The Bayesian analyses estimated 99.5% probability that in-clinic group visits, compared with in-clinic individual visits, resulted in a higher percent weight loss (posterior mean difference: 1.8%[95% CrI: 0.5,3.2%]), a greater probability of achieving 5% threshold (posterior mean difference: 9.2% [95% CrI: 2.4, 16.0%]) and 10% threshold (posterior mean difference: 6.6% [95% CrI: 1.7, 11.5%]). The phone-based group visits had similar result. We also concluded that including intervention group did not impact model fit significantly.

Conclusions

We unified the analyses of continuous (the primary endpoint) and categorical measures (the secondary endpoints) of weight loss with one single Bayesian hierarchical model. This approach gained statistical power for the dichotomized endpoints by leveraging the information in the continuous data. Furthermore, the Bayesian analysis enabled additional insights into data interpretation and inference by providing posterior distributions for parameters of interest and posterior probabilities of different hypotheses that were not available with the frequentist approach.

Trial registration

ClinicalTrials.gov Identifier NCT02456636; date of registry: May 28, 2015.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-021-01499-0.

How to measure premature mortality? A proposal combining "relative" and "absolute" approaches

BACKGROUND

The concept of "premature mortality" is at the heart of many national and global health measurement and benchmarking efforts. However, despite the intuitive appeal of its underlying concept, it is far from obvious how to best operationalise it. The previous work offers at least two basic approaches: an absolute and a relative one. The former-and far more widely used- approach sets a unique age threshold (e.g. 65 years), below which deaths are defined as premature. The relative approach derives the share of premature deaths from the country-specific age distribution of deaths in the country of interest. The biggest disadvantage of the absolute approach is that of using a unique, arbitrary threshold for different mortality patterns, while the main disadvantage of the relative approach is that its estimate of premature mortality strongly depends on how the senescent deaths distribution is defined in each country.

METHOD

We propose to overcome some of the downsides of the existing approaches, by combining features of both, using a hierarchical model, in which senescent deaths distribution is held constant for each country as a pivotal quantity and the premature mortality distribution is allowed to vary across countries. In this way, premature mortality estimates become more comparable across countries with similar characteristics.

RESULTS

The proposed hierarchical models provide results, which appear to align with related evidence from  specific countries. In particular, we find a relatively high premature mortality for the United States and Denmark.

CONCLUSIONS

While our hybrid approach overcomes some of the problems of previous measures, some issues require further research, in particular the choice of the group of countries that a given country is assigned to and the choice of the benchmarks within the groups. Hence, our proposed method, combined with further study addressing these issues, could provide a valid alternative way to measure and compare premature mortality across countries.

RoBoT: a robust Bayesian hypothesis testing method for basket trials

A basket trial in oncology encompasses multiple "baskets" that simultaneously assess one treatment in multiple cancer types or subtypes. It is well-recognized that hierarchical modeling methods, which adaptively borrow strength across baskets, can improve over simple pooling and stratification. We propose a novel Bayesian method, RoBoT (Robust Bayesian Hypothesis Testing), for the data analysis and decision-making in phase II basket trials. In contrast to most existing methods that use posterior credible intervals to determine the efficacy of the new treatment, RoBoT builds upon a formal Bayesian hypothesis testing framework that leads to interpretable and robust inference. Specifically, we assume that the baskets belong to several latent subgroups, and within each subgroup, the treatment has similar probabilities of being more efficacious than controls, historical, or concurrent. The number of latent subgroups and subgroup memberships are inferred by the data through a Dirichlet process mixture model. Such model specification helps avoid type I error inflation caused by excessive shrinkage under typical hierarchical models. The operating characteristics of RoBoT are assessed through computer simulations and are compared with existing methods. Finally, we apply RoBoT to data from two recent phase II basket trials of imatinib and vemurafenib, respectively.

Additive and multiplicative probabilistic models of infant looking times

Additive and multiplicative regression models of habituation were compared regarding the fit to looking times from a habituation experiment with infants aged between 3 and 11 months. In contrast to earlier studies, the current study considered multiple probability distributions, namely Weibull, gamma, lognormal and normal distribution. In the habituation experiment the type of contrast between the habituation and the test trial was varied (luminance, color or orientation contrast), crossed with the number of habituation trials (1, 3, 5, or 7 habituation trials) and crossed with three age cohorts (4, 7, 10 months). The initial mean LT to dark stimuli (around 3.7 s) was considerably shorter than the mean LT to green and gray stimuli (around 5 s). Infants showed the strongest dishabituation to changes from dark to bright (luminance contrast) and weak-to-no dishabituation to a 90-degrees rotation of the gray stimuli (orientation contrast). The dishabituation was stronger after five and seven habituation trials, but the result was not statistically robust. The gamma distribution showed the best fit in terms of log-likelihood and mean absolute error and the best predictive performance. Furthermore, the gamma distribution showed small correlations between parameters relative to other models. The normal additive model showed an inferior fit and medium correlations between the parameters. In particular, the positive correlation between the initial looking time (LT) and the habituation rate was likely responsible for a different interpretation relative to the multiplicative models of the main effect of age on the habituation rate. Otherwise, the additive and multiplicative models provided similar statistical conclusions. The performance of the model versions without pooling and with partial pooling across participants (also called random-effects, multi-level or hierarchical models) were compared. The latter type of models showed worse data fit but more precise predictions and reduced correlations between the parameters. The performance of model variants with auto-regressive time structures were explored but showed considerably worse fit. The performance of quadratic models that allowed non-monotonic changes in LTs were investigated as well. However, when fitted with LT data, these models did not produce non-monotonic change in LTs. The study underscores the utility of partial-pooling models in terms of providing more accurate predictions. Further, it agrees with previous research in that a multiplicative LT model is preferable. Nevertheless, the current results suggest that the impact of the choice of an additive model on the statistical inference is less dramatic then previously assumed.

Incorporation of feeding functional group information informs explanatory patterns of long-term population changes in fish assemblages

The objective of this study was to evaluate long term trends of fish taxa in southern Lake Michigan while incorporating their functional roles to improve our understanding of ecosystem level changes that have occurred in the system over time. The approach used here highlighted the ease of incorporating ecological mechanisms into population models so researchers can take full advantage of available long-term ecosystem information. Long term studies of fish assemblages can be used to inform changes in community structure resulting from perturbations to aquatic systems and understanding these changes in fish assemblages can be better contextualized by grouping species according to functional groups that are grounded in niche theory. We hypothesized that describing the biological process based on partial pooling of information across functional groups would identify shifts in fish assemblages that coincide with major changes in the ecosystem (e.g., for this study, shifts in zooplankton abundance over time). Herein, we analyzed a long-term Lake Michigan fisheries dataset using a multi-species state space modeling approach within a Bayesian framework. Our results suggested the population growth rates of planktivores and benthic invertivores have been more variable than general invertivores over time and that trends in planktivores can be partially explained by ecosystem changes in zooplankton abundance. Additional work incorporating more ecosystem parameters (e.g., primary production, etc.) should be incorporated into future iterations of this novel modeling concept.

Slamming the sham: A Bayesian model for adaptive adjustment with noisy control data

It is not always clear how to adjust for control data in causal inference, balancing the goals of reducing bias and variance. We show how, in a setting with repeated experiments, Bayesian hierarchical modeling yields an adaptive procedure that uses the data to determine how much adjustment to perform. The result is a novel analysis with increased statistical efficiency compared with the default analysis based on difference estimates. We demonstrate this procedure on two real examples, as well as on a series of simulated datasets. We show that the increased efficiency can have real-world consequences in terms of the conclusions that can be drawn from the experiments. We also discuss the relevance of this work to causal inference and statistical design and analysis more generally.

The impact of stormwater biofilter design and operational variables on nutrient removal - a statistical modelling approach

Biofiltration systems can help mitigate the impact of urban runoff as they can treat, retain and attenuate stormwater. It is important to select the optimal design characteristics of biofilters (e.g., vegetation, filter media depth) to ensure high treatment performance. Operational conditions (e.g., infiltration rate) can also lead to significant changes in biofilter treatment performance over time. The impact of specific operational conditions on water quality treatment performance of stormwater biofilters is still not well understood. Furthermore, despite the importance of design characteristics and operational conditions on biofilter treatment performance, there is a lack of models that can be used to determine the optimal design and operation. In this paper, we developed a series of statistical models to predict the Total Phosphorus (TP) and Total Nitrogen (TN) removal performance of stormwater biofilters using various numbers of design characteristics and operational conditions. These statistical models were tested using data collected from four extensive laboratory-scale biofilter column studies. It was found that all models performed relatively well with a Nash-Sutcliffe Efficiency (NSE) of 0.42 - 0.61 for TP and 0.37 - 0.63 for TN. The most important design characteristics were filter media type and depth for TP treatment, and vegetation type and submerged zone depth for TN treatment. In addition, infiltration rate and inflow concentrations were the operational conditions that greatly influence outflow TP and TN concentrations from stormwater biofilters. As such, these variables need to be carefully considered when designing and operating stormwater biofilters. Sensitivity analysis results indicate that the model was quite sensitive to all regression coefficients and intercepts. Additional modelling exercises show that the model could be further simplified by reducing the number of cross-correlated parameters. These models can be used by practitioners for not just optimising the design, but also operating biofilters using real-time monitoring and control to achieve optimum performance.

Hierarchical Modelling of COVID-19 Death Risk in India in the Early Phase of the Pandemic

We improve upon the modelling of India's pandemic vulnerability. Our model is multidisciplinary and recognises the nested levels of the epidemic. We create a model of the risk of severe COVID-19 and death, instead of a model of transmission. Our model allows for socio-demographic-group differentials in risk, obesity and underweight people, morbidity status and other conditioning regional and lifestyle factors. We build a hierarchical multilevel model of severe COVID-19 cases, using three different data sources: the National Family Health Survey for 2015/16, Census data for 2011 and data for COVID-19 deaths obtained cumulatively until June 2020. We provide results for 11 states of India, enabling best-yet targeting of policy actions. COVID-19 deaths in north and central India were higher in areas with older and overweight populations, and were more common among people with pre-existing health conditions, or who smoke, or who live in urban areas. Policy experts may both want to 'follow World Health Organisation advice' and yet also use disaggregated and spatially specific data to improve wellbeing outcomes during the pandemic. The future uses of our innovative data-combining model are numerous.

Bayesian hierarchical dose-response meta-analysis of epidemiological studies: Modeling and target population prediction methods

When assessing the human risks due to exposure to environmental chemicals, traditional dose-response analyses are not straightforward when there are numerous high-quality epidemiological studies of priority cancer and non-cancer health outcomes. Given this wealth of information, selecting a single "best" study on which to base dose-response analyses is difficult and would potentially ignore much of the available data. Therefore, systematic approaches are necessary for the analysis of these rich databases. Examples are meta-analysis (and further, meta-regression), which are well established methods that consider and incorporate information from multiple studies into the estimation of risks due to exposure to environmental contaminants. In this paper, we propose a hierarchical, Bayesian meta-analysis approach for the dose-response analysis of multiple epidemiological studies. This paper is the second of two papers detailing this approach; the first covered "pre-analysis" steps necessary to prepare the data for dose-response modeling. This paper focuses on the hierarchical Bayesian approach to dose-response modeling and extrapolation of risk to populations of interest using the association between bladder cancer and oral inorganic arsenic (iAs) exposure as an illustrative case study. In particular, this paper addresses the modeling of both case-control and cohort studies with a flexible, logistic model in a hierarchical Bayesian framework that estimates study-specific slopes, as well as a pooled slope across all studies. This approach is akin to a random effects model in which no assumption is made a priori that there is a single, common slope for all included studies. Further, this paper also details extrapolation of the estimates of logistic slope to extra risk in a target population using a lifetable analysis and basic assumptions about background iAs exposure levels. In this case, the target population was the general United States population and information on all-cause mortality and incidence and mortality from bladder cancer was used to perform the lifetable analysis. The methods herein were developed for general use in investigating the association between any pollutant and observed health-effects in epidemiological studies. In order to demonstrate these methods, inorganic arsenic was chosen as a case study given the large epidemiological database that exists for this contaminant.

Analysis and prediction of climate forecasts in Northern Morocco: application of multilevel linear mixed effects models using R software

For many years, the application of mixed-effects modeling has received much attention for predicting scenarios in the fields of theoretical and applied sciences. In this study, a "new" Multilevel Linear Mixed-Effects (LME) model is proposed to analyze and predict multiply-nested and hierarchical data. Temperature and rainfall observation were carried out successively between 1979-2014 and 1984-2018; and the data input was organized on monthly basis for each year. Besides, a daily observation was made for "Dar Chaoui" zone of Northern Morocco. However, we chose in the first time a simple linear regression model, but the estimation has been just for fixed effects and ignoring the random effect. On the other hand, in multilevel linear mixed effects models, once the model has been formulated, methods are needed to estimate the model parameters. In this section, we first deal with the joint estimation of the fixed effects (β), random effects (ui) and then with estimation of the variance parameters (γ, ρ and σ2). The study revealed that the predicted values are very close to the real value. Besides, this model is capable of modelling the error, fixed and random parts of the sample. Moreover, in this range, the results showed that there is three standard deviations measures for fixed and random effects, also the variance measure, which demonstrate us a great prediction. In conclusion, this model gives a decisive precision of results that can be exploited in studies for forecast of water balance and/or soil erosion. These results can also be used to inhibit the risk of erosion with possible arrangements for the environment and human security.

Association of response rate and prevalence estimates of common mental disorders across 129 areas in a nationally representative survey of adults in Japan

OBJECTIVE

To investigate the association of area response rate with prevalence estimates of mental disorders in the 2nd World Mental Health Survey (WMHJ2).

METHODS

The sample of the WMHJ2 was selected from community residents in 129 areas from three regions of Japan. The surveys were conducted between 2013 and 2015, and 2450 (43.4%) responded. Mental disorders as well as three disorder classes (mood, anxiety, and substance use disorders) were identified using the WHO CIDI/DSM-IV. Response rates and 12-month and lifetime prevalences were calculated for each area. A generalized linear mixed model analysis was conducted to associate area response rate with the prevalence of mental disorders, controlling for sex, age, urbanity, and geographical region.

RESULTS

Area response rates ranged from 0.05 to 0.80 across the 129 areas. Area response rate was not significantly associated with 12-month or lifetime prevalence of mental disorder. Lifetime prevalences of substance use disorder were significantly lower in a survey with a higher response rate than a survey of the same area with a lower response rate.

CONCLUSION

Response rate may not strongly affect the prevalence estimates of mental disorders in a community-based survey of the prevalence of common mental disorders during a particular time frame. However, a lower response rate could be associated with overestimation of lifetime prevalence of substance use disorder. This needs further elucidation.

Bayesian Design for Pediatric Clinical Trials with Binary Endpoints When Borrowing Historical Information of Treatment Effect

The efficacy evaluation in pediatric population is an important component of drug development and is generally required by the regulatory agencies. It is often challenging to enroll pediatric subjects for a large trial especially when the incidence rate is low in certain disease areas. Bayesian framework can provide analytic avenues to effectively utilize historical information of the treatment effect and help make pediatric trials more efficient by reducing the sample size when there is evidence to suggest similarity of the treatment responses between the populations. Schoenfeld et al. (Clin Trials 6(4):297-304, 2009) proposed a Bayesian hierarchical model for efficacy extrapolation for continuous endpoints, which connects a single historical trial and the current trial by a variance parameter in the prior distribution. In this manuscript, we extend the existing model to borrow strength from multiple historical trials under the same assumptions and develop a quantitative method to borrow historical information more efficiently. Furthermore, we extend Schoenfeld's method based on continuous endpoints to binary endpoints with a hierarchical binomial model to extrapolate efficacy. Sensitivity analyses for the underlying assumptions are discussed with simulations and the methods are illustrated with a real case study, along with some practical considerations about how to choose the prior distribution.

HINT: A hierarchical independent component analysis toolbox for investigating brain functional networks using neuroimaging data

BACKGROUND

Independent component analysis (ICA) is a popular tool for investigating brain organization in neuroscience research. In fMRI studies, an important goal is to study how brain networks are modulated by subjects' clinical and demographic variables. Existing ICA methods and toolboxes don't incorporate subjects' covariates effects in ICA estimation of brain networks, which potentially leads to loss in accuracy and statistical power in detecting brain network differences between subjects' groups.

NEW METHOD

We introduce a Matlab toolbox, HINT (Hierarchical INdependent component analysis Toolbox), that provides a hierarchical covariate-adjusted ICA (hc-ICA) for modeling and testing covariate effects and generates model-based estimates of brain networks on both the population- and individual-level. HINT provides a user-friendly Matlab GUI that allows users to easily load images, specify covariate effects, monitor model estimation via an EM algorithm, specify hypothesis tests, and visualize results. HINT also has a command line interface which allows users to conveniently run and reproduce the analysis with a script.

COMPARISON TO EXISTING METHODS

HINT implements a new multi-level probabilistic ICA model for group ICA. It provides a statistically principled ICA modeling framework for investigating covariate effects on brain networks. HINT can also generate and visualize model-based network estimates for user-specified subject groups, which greatly facilitates group comparisons.

RESULTS

We demonstrate the steps and functionality of HINT with an fMRI example data to estimate treatment effects on brain networks while controlling for other covariates. Results demonstrate estimated brain networks and model-based comparisons between the treatment and control groups. In comparisons using synthetic fMRI data, HINT shows desirable statistical power in detecting group differences in networks especially in small sample sizes, while maintaining a low false positive rate. HINT also demonstrates similar or increased accuracy in reconstructing both population- and individual-level source signal maps as compared to some state-of-the-art group ICA methods.

CONCLUSION

HINT can provide a useful tool for both statistical and neuroscience researchers to evaluate and test differences in brain networks between subject groups.

A hierarchical model of perceptual multistability involving interocular grouping

Ambiguous visual images can generate dynamic and stochastic switches in perceptual interpretation known as perceptual rivalry. Such dynamics have primarily been studied in the context of rivalry between two percepts, but there is growing interest in the neural mechanisms that drive rivalry between more than two percepts. In recent experiments, we showed that split images presented to each eye lead to subjects perceiving four stochastically alternating percepts (Jacot-Guillarmod et al. Vision research, 133, 37-46, 2017): two single eye images and two interocularly grouped images. Here we propose a hierarchical neural network model that exhibits dynamics consistent with our experimental observations. The model consists of two levels, with the first representing monocular activity, and the second representing activity in higher visual areas. The model produces stochastically switching solutions, whose dependence on task parameters is consistent with four generalized Levelt Propositions, and with experiments. Moreover, dynamics restricted to invariant subspaces of the model demonstrate simpler forms of bistable rivalry. Thus, our hierarchical model generalizes past, validated models of binocular rivalry. This neuromechanistic model also allows us to probe the roles of interactions between populations at the network level. Generalized Levelt's Propositions hold as long as feedback from the higher to lower visual areas is weak, and the adaptation and mutual inhibition at the higher level is not too strong. Our results suggest constraints on the architecture of the visual system and show that complex visual stimuli can be used in perceptual rivalry experiments to develop more detailed mechanistic models of perceptual processing.

Factor analyses of multidimensional symptoms in a large group of patients with major depressive disorder, bipolar disorder, schizoaffective disorder and schizophrenia

BACKGROUND

There is an ongoing discussion about which neurobiological correlates or symptoms separate the major psychoses (i.e. Major Depressive Disorder MDD, Bipolar Disorder BD, and Schizophrenia SZ). Psychopathological factor analyses within one of these disorders have resulted in models including one to five factors. Factor analyses across the major psychoses using a comprehensive set of psychopathological scales in the same patients are lacking. It is further unclear, whether hierarchical or unitarian models better summarize phenomena.

METHOD

Patients (n = 1182) who met DSM-IV criteria for MDD, BD, SZ or schizoaffective disorder were assessed with the SANS, SAPS, HAMA, HAM-D, and YMRS. The sample was split into two and analyzed using explorative and confirmatory factor analyses to extract psychopathological factors independent of diagnosis.

RESULTS

In the exploratory analysis of sample 1 (n = 593) we found 5 factors. The confirmatory analysis using sample 2 (n = 589) confirmed the 5-factor model (χ² = 1287.842, df = 571, p < .0001: CFI = 0.932; RMSEA = 0.033). The 5-factors were depression, negative syndrome, positive formal thought disorder, paranoid-hallucinatory syndrome, and increased appetite. Increased appetite was not related to medication. None of the factors was specific for one diagnosis. Second order factor analysis revealed two higher order factors: negative/affective (I) and positive symptoms (II).

CONCLUSION

This is the first study delineating psychopathological factors in a large group of patients across the spectrum of affective and psychotic disorders. In future neurobiological studies, we should consider transdiagnostic syndromes besides the traditional diagnoses.

Zero-inflated generalized Dirichlet multinomial regression model for microbiome compositional data analysis

There is heightened interest in using high-throughput sequencing technologies to quantify abundances of microbial taxa and linking the abundance to human diseases and traits. Proper modeling of multivariate taxon counts is essential to the power of detecting this association. Existing models are limited in handling excessive zero observations in taxon counts and in flexibly accommodating complex correlation structures and dispersion patterns among taxa. In this article, we develop a new probability distribution, zero-inflated generalized Dirichlet multinomial (ZIGDM), that overcomes these limitations in modeling multivariate taxon counts. Based on this distribution, we propose a ZIGDM regression model to link microbial abundances to covariates (e.g. disease status) and develop a fast expectation-maximization algorithm to efficiently estimate parameters in the model. The derived tests enable us to reveal rich patterns of variation in microbial compositions including differential mean and dispersion. The advantages of the proposed methods are demonstrated through simulation studies and an analysis of a gut microbiome dataset.

Emergence of cognitive priming and structure building from the hierarchical interaction of canonical microcircuit models

The concept of connectionism states that higher cognitive functions emerge from the interaction of many simple elements. Accordingly, research on canonical microcircuits conceptualizes findings on fundamental neuroanatomical circuits as well as recurrent organizational principles of the cerebral cortex and examines the link between architectures and their associated functionality. In this study, we establish minimal canonical microcircuit models as elements of hierarchical processing networks. Based on a combination of descriptive time simulations and explanatory state-space mappings, we show that minimal canonical microcircuits effectively segregate feedforward and feedback information flows and that feedback information conditions basic processing operations in minimal canonical microcircuits. Further, we derive and examine two prototypical meta-circuits of cooperating minimal canonical microcircuits for the neurocognitive problems of priming and structure building. Through the application of these findings to a language network of syntax parsing, this study embodies neurocognitive research on hierarchical communication in light of canonical microcircuits, cell assembly theory, and predictive coding.

Development of an algorithm for evaluating the impact of measurement variability on response categorization in oncology trials

BACKGROUND

Radiologic assessments of baseline and post-treatment tumor burden are subject to measurement variability, but the impact of this variability on the objective response rate (ORR) and progression rate in specific trials has been unpredictable on a practical level. In this study, we aimed to develop an algorithm for evaluating the quantitative impact of measurement variability on the ORR and progression rate.

METHODS

First, we devised a hierarchical model for estimating the distribution of measurement variability using a clinical trial dataset of computed tomography scans. Next, a simulation method was used to calculate the probability representing the effect of measurement errors on categorical diagnoses in various scenarios using the estimated distribution. Based on the probabilities derived from the simulation, we developed an algorithm to evaluate the reliability of an ORR (or progression rate) (i.e., the variation in the assessed rate) by generating a 95% central range of ORR (or progression rate) results if a reassessment was performed. Finally, we performed validation using an external dataset. In the validation of the estimated distribution of measurement variability, the coverage level was calculated as the proportion of the 95% central ranges of hypothetical second readings that covered the actual burden sizes. In the validation of the evaluation algorithm, for 100 resampled datasets, the coverage level was calculated as the proportion of the 95% central ranges of ORR results that covered the ORR from a real second assessment.

RESULTS

We built a web tool for implementing the algorithm (publicly available at http://studyanalysis2017.pythonanywhere.com/ ). In the validation of the estimated distribution and the algorithm, the coverage levels were 93 and 100%, respectively.

CONCLUSIONS

The validation exercise using an external dataset demonstrated the adequacy of the statistical model and the utility of the developed algorithm. Quantification of variation in the ORR and progression rate due to potential measurement variability is essential and will help inform decisions made on the basis of trial data.

Total phosphorus-precipitation and Chlorophyll a-phosphorus relationships of lakes and reservoirs mediated by soil iron at regional scale

Phosphorus is a critical element determining trophic status and Chlorophyll a (Chl a) level in natural lakes and reservoirs, and total phosphorus (TP) concentrations can be predicted from data on phosphorus loading, hydraulic flushing rate and sedimentation. Due to their interactions with phosphorus, iron (hydr) oxides in suspended particles, originally derived from watershed soil, can strongly influence the phosphorus sedimentation and phosphorus bioavailability in water columns. Thus, the TP-precipitation relationship and the response of Chl a to TP are likely associated with watersheds soil iron. To test this assumption, we built hierarchical linear models for summer observation of natural lakes and reservoirs across a large geographic gradient. The intercepts and slopes of TP-precipitation relationships are higher in natural lakes than those in reservoirs, and these model coefficients exhibit latitudinal variations that are explained by the natural soil iron gradient. Soil iron, operating at a regional level, significantly mediates the effect of precipitation on TP concentration in both natural lakes and reservoirs, and drives the latitudinal variation in the Chl a-TP relationships for reservoirs. Our results imply that the increase in extreme precipitation events anticipated under future climate conditions may substantially mitigate eutrophication in tropical and subtropical reservoirs, but may worsen conditions in temperate lakes.

Borrowing from Historical Control Data in Cancer Drug Development: A Cautionary Tale and Practical Guidelines

Some clinical trialists, especially those working in rare or pediatric disease, have suggested borrowing information from similar but already-completed clinical trials. This paper begins with a case study in which relying solely on historical control information would have erroneously resulted in concluding a significant treatment effect. We then attempt to catalog situations where borrowing historical information may or may not be advisable using a series of carefully designed simulation studies. We use an MCMC-driven Bayesian hierarchical parametric survival modeling approach to analyze data from a sponsor's colorectal cancer study. We also apply these same models to simulated data comparing the effective historical sample size, bias, 95% credible interval widths, and empirical coverage probabilities across the simulated cases. We find that even after accounting for variations in study design, baseline characteristics, and standard-of-care improvement, our approach consistently identifies Bayesianly significant differences between the historical and concurrent controls under a range of priors on the degree of historical data borrowing. Our simulation studies are far from exhaustive, but inform the design of future trials. When the historical and current controls are not dissimilar, Bayesian methods can still moderate borrowing to a more appropriate level by adjusting for important covariates and adopting sensible priors.

A Bayesian hierarchical model for network meta-analysis of multiple diagnostic tests

To compare the accuracy of multiple diagnostic tests in a single study, three designs are commonly used (i) the multiple test comparison design; (ii) the randomized design, and (iii) the non-comparative design. Existing meta-analysis methods of diagnostic tests (MA-DT) have been focused on evaluating the performance of a single test by comparing it with a reference test. The increasing number of available diagnostic instruments for a disease condition and the different study designs being used have generated the need to develop efficient and flexible meta-analysis framework to combine all designs for simultaneous inference. In this article, we develop a missing data framework and a Bayesian hierarchical model for network MA-DT (NMA-DT) and offer important promises over traditional MA-DT: (i) It combines studies using all three designs; (ii) It pools both studies with or without a gold standard; (iii) it combines studies with different sets of candidate tests; and (iv) it accounts for heterogeneity across studies and complex correlation structure among multiple tests. We illustrate our method through a case study: network meta-analysis of deep vein thrombosis tests.

Urbanization and Internet addiction in a nationally representative sample of adult community residents in Japan: A cross-sectional, multilevel study

This study examines the relationship between urbanization and Internet addiction (IA) and association with other psychopathology and social support, in a nationally representative sample in Japan. Data from the World Mental Health Japan Second Survey were used. There were 2450 survey respondents, with an average response rate of 43.4%. Respondents' living areas were divided into three groups on the basis of urbanization (operationalized as city size). IA was measured using the Compulsive Internet Use Scale (CIUS). Three variables - psychological distress, past-12-month mental disorder, and social support - were measured using established instruments. A multilevel model was conducted to clarify the association between urbanization and IA (continuous scores and prevalence), before and after adjusting for possible individual-level and area-level variables and demographic variables. CIUS scores were significantly higher in large cities than in small municipalities before adjusting for psychological distress, social support, and past-12-month mental disorder. After adjustment, these associations attenuated substantially: urbanization was no longer significantly associated with odds of mild/severe IA, while the relationship held for continuous CIUS scores. Thus, residence in large cities is associated with higher odds of IA in Japan; psychological distress, social support, and past-12-month mental disorder partly explain this association.

Exploring the impacts of speed variances on safety performance of urban elevated expressways using GPS data

Speed variation on urban expressways has been frequently noted as a key factor associated with high crash risk. However, it was often difficult to capture the safety impact of speed variance with spaced sensor measurements. As an alternative, this paper aims to leverage the use of the floating car data (FCD) to capture the speed variance in a morning rush hour on urban elevated expressways and examine its effect on safety. A semi-automatic filtering process was introduced to distinguish taxi GPS data points on the elevated expressways from the ones on the surface roads under the expressways. Subsequently, the standard deviation of the cross-sectional speed mean (SDCSM) and the cross-section speed standard deviation (MCSSD) were derived to capture the spatial and temporal speed variances, respectively. Together with other explanatory variables, both hierarchical and non-hierarchical Poisson-gamma measurement error models were developed to model the crash frequencies of the expressways. The modeling results showed that the hierarchical model performed better and both SDCSM and MCSSD were found to be positively related to the crash occurrence. This secures the need for addressing the impact of speed variation when modeling crashes occurred on the elevated expressways.

Improving estimates of built-up area from night time light across globally distributed cities through hierarchical modeling

Built-up area has become an important indicator for studying urban environments, but mapping built-up area at the regional/global scale remains challenging due to the complexity of impervious surface features. Nighttime light data (NTL) is one of the major remote sensing data sources for regional/global built-up or impervious surface mapping. A single regression relationship between fractional built-up/impervious area and NTL or various indices derived based on NTL and vegetation index (e.g., NDVI) data had been established in many previous studies. However, due to the varying geographical, climatic, and socio-economic characteristics of cities, the same regression relationship may vary significantly across cities. In this study, we examined the regression relationship between percentage of built-up area (pBUA) and vegetation adjusted nighttime light urban index (VANUI) for 120 randomly selected cities around the world with a hierarchical hockey-stick regression model. We found that there is a substantial variability in the slope (0.658 ± 0.318), the threshold VANUI (-1.92 ± 0.769, log scale) after which the linear relationship holds, and the coefficient of determination R² (0.71 ± 0.14) among globally distributed cities. A small proportion of this substantial variability can be attributed to socio-economic status (e.g., total population, GDP per capita) and landscape structures (e.g., compactness and fragmentation). Due to these variations, our hierarchical model or no-pooling model (i.e., fit each city individually) can significantly improve model prediction accuracy (17% in terms of root mean squared error) over a complete-pooling model. We, however, recommend hierarchical models as they can provide meaningful priors for future modeling under a Bayesian framework, and achieve higher prediction accuracy than no-pooling models when sample size is small.

Time-varying predatory behavior is primary predictor of fine-scale movement of wildland-urban cougars

BACKGROUND

While many species have suffered from the detrimental impacts of increasing human population growth, some species, such as cougars (Puma concolor), have been observed using human-modified landscapes. However, human-modified habitat can be a source of both increased risk and increased food availability, particularly for large carnivores. Assessing preferential use of the landscape is important for managing wildlife and can be particularly useful in transitional habitats, such as at the wildland-urban interface. Preferential use is often evaluated using resource selection functions (RSFs), which are focused on quantifying habitat preference using either a temporally static framework or researcher-defined temporal delineations. Many applications of RSFs do not incorporate time-varying landscape availability or temporally-varying behavior, which may mask conflict and avoidance behavior.

METHODS

Contemporary approaches to incorporate landscape availability into the assessment of habitat selection include spatio-temporal point process models, step selection functions, and continuous-time Markov chain (CTMC) models; in contrast with the other methods, the CTMC model allows for explicit inference on animal movement in continuous-time. We used a hierarchical version of the CTMC framework to model speed and directionality of fine-scale movement by a population of cougars inhabiting the Front Range of Colorado, U.S.A., an area exhibiting rapid population growth and increased recreational use, as a function of individual variation and time-varying responses to landscape covariates.

RESULTS

We found evidence for individual- and daily temporal-variability in cougar response to landscape characteristics. Distance to nearest kill site emerged as the most important driver of movement at a population-level. We also detected seasonal differences in average response to elevation, heat loading, and distance to roads. Motility was also a function of amount of development, with cougars moving faster in developed areas than in undeveloped areas.

CONCLUSIONS

The time-varying framework allowed us to detect temporal variability that would be masked in a generalized linear model, and improved the within-sample predictive ability of the model. The high degree of individual variation suggests that, if agencies want to minimize human-wildlife conflict management options should be varied and flexible. However, due to the effect of recursive behavior on cougar movement, likely related to the location and timing of potential kill-sites, kill-site identification tools may be useful for identifying areas of potential conflict.

A utility-based design for randomized comparative trials with ordinal outcomes and prognostic subgroups

A design is proposed for randomized comparative trials with ordinal outcomes and prognostic subgroups. The design accounts for patient heterogeneity by allowing possibly different comparative conclusions within subgroups. The comparative testing criterion is based on utilities for the levels of the ordinal outcome and a Bayesian probability model. Designs based on two alternative models that include treatment-subgroup interactions are considered, the proportional odds model and a non-proportional odds model with a hierarchical prior that shrinks toward the proportional odds model. A third design that assumes homogeneity and ignores possible treatment-subgroup interactions also is considered. The three approaches are applied to construct group sequential designs for a trial of nutritional prehabilitation versus standard of care for esophageal cancer patients undergoing chemoradiation and surgery, including both untreated patients and salvage patients whose disease has recurred following previous therapy. A simulation study is presented that compares the three designs, including evaluation of within-subgroup type I and II error probabilities under a variety of scenarios including different combinations of treatment-subgroup interactions.

Modeling the Structure of Acute Sport-Related Concussion Symptoms: A Bifactor Approach

OBJECTIVES

Concussions cause diverse symptoms that are often measured through a single symptom severity score. Researchers have postulated distinct dimensions of concussion symptoms, raising the possibility that total scores may not accurately represent their multidimensional nature. This study examined to what degree concussion symptoms, assessed by the Sport Concussion Assessment Tool 3 (SCAT3), reflect a unidimensional versus multidimensional construct to inform how the SCAT3 should be scored and advance efforts to identify distinct phenotypes of concussion.

METHODS

Data were aggregated across two prospective studies of sport-related concussion, yielding 219 high school and college athletes in the acute (<48 hr) post-injury period. Item-level ratings on the SCAT3 checklist were analyzed through exploratory and confirmatory factor analyses. We specified higher-order and bifactor models and compared their fit, interpretability, and external correlates.

RESULTS

The best-fitting model was a five-factor bifactor model that included a general factor on which all items loaded and four specific factors reflecting emotional symptoms, torpor, sensory sensitivities, and headache symptoms. The bifactor model demonstrated better discriminant validity than the counterpart higher-order model, in which the factors were highly correlated (r=.55-.91).

CONCLUSIONS

The SCAT3 contains items that appear unidimensional, suggesting that it is appropriate to quantify concussion symptoms with total scores. However, evidence of multidimensionality was revealed using bifactor modeling. Additional work is needed to clarify the nature of factors identified by this model, explicate their clinical and research utility, and determine to what degree the model applies to other stages of injury recovery and patient subgroups. (JINS, 2018, 24, 793-804).

Reference effect measures for quantifying, comparing and visualizing variation from random and fixed effects in non-normal multilevel models, with applications to site variation in medical procedure use and outcomes

BACKGROUND

Multilevel models for non-normal outcomes are widely used in medical and health sciences research. While methods for interpreting fixed effects are well-developed, methods to quantify and interpret random cluster variation and compare it with other sources of variation are less established. Random cluster variation, sometimes referred to as general contextual effects (GCE), may be the main focus of a study; therefore, easily interpretable methods are needed to quantify GCE. We propose a Reference Effect Measure (REM) approach to 1) quantify GCE and compare it to individual subject and cluster covariate effects, and 2) quantify relative magnitudes of GCE and variation from sets of measured factors.

METHODS

To illustrate REM, we consider a two-level mixed logistic model with patients clustered within hospitals and a random intercept for hospitals. We compare patients at hospitals at given percentiles of the estimated random effect distribution to patients at a median or 'reference' hospital. These estimates are then compared numerically and graphically to individual fixed effects to quantify GCE in the context of effects of other measured variables (aim 1). We then extend this approach by comparing variation from the random effect distribution to variation from sets of fixed effects to understand their magnitudes relative to overall outcome variation (aim 2).

RESULTS

Using an example of initiation of rhythm control treatment in atrial fibrillation (AF) patients within the Veterans Affairs (VA), we use REM to demonstrate that random variation across hospitals (GCE) in initiation of treatment is substantially greater than that due to most individual patient factors, and explains at least as much variation in treatment initiation as do all patient factors combined. These results are contrasted with a relatively small GCE compared with patient factors in 1 year mortality following hospitalization for AF patients.

CONCLUSIONS

REM provides a means of quantifying random effect variation (GCE) with multilevel data and can be used to explore drivers of outcome variation. This method is easily interpretable and can be presented visually. REM offers a simple, interpretable approach for evaluating questions of growing importance in the study of health care systems.

Integrated Bayesian analysis of rare exonic variants to identify risk genes for schizophrenia and neurodevelopmental disorders

BACKGROUND

Integrating rare variation from trio family and case-control studies has successfully implicated specific genes contributing to risk of neurodevelopmental disorders (NDDs) including autism spectrum disorders (ASD), intellectual disability (ID), developmental disorders (DDs), and epilepsy (EPI). For schizophrenia (SCZ), however, while sets of genes have been implicated through the study of rare variation, only two risk genes have been identified.

METHODS

We used hierarchical Bayesian modeling of rare-variant genetic architecture to estimate mean effect sizes and risk-gene proportions, analyzing the largest available collection of whole exome sequence data for SCZ (1,077 trios, 6,699 cases, and 13,028 controls), and data for four NDDs (ASD, ID, DD, and EPI; total 10,792 trios, and 4,058 cases and controls).

RESULTS

For SCZ, we estimate there are 1,551 risk genes. There are more risk genes and they have weaker effects than for NDDs. We provide power analyses to predict the number of risk-gene discoveries as more data become available. We confirm and augment prior risk gene and gene set enrichment results for SCZ and NDDs. In particular, we detected 98 new DD risk genes at FDR < 0.05. Correlations of risk-gene posterior probabilities are high across four NDDs (ρ>0.55), but low between SCZ and the NDDs (ρ<0.3). An in-depth analysis of 288 NDD genes shows there is highly significant protein-protein interaction (PPI) network connectivity, and functionally distinct PPI subnetworks based on pathway enrichment, single-cell RNA-seq cell types, and multi-region developmental brain RNA-seq.

CONCLUSIONS

We have extended a pipeline used in ASD studies and applied it to infer rare genetic parameters for SCZ and four NDDs ( https://github.com/hoangtn/extTADA ). We find many new DD risk genes, supported by gene set enrichment and PPI network connectivity analyses. We find greater similarity among NDDs than between NDDs and SCZ. NDD gene subnetworks are implicated in postnatally expressed presynaptic and postsynaptic genes, and for transcriptional and post-transcriptional gene regulation in prenatal neural progenitor and stem cells.

Bayesian inference for biomarker discovery in proteomics: an analytic solution

This paper addresses the question of biomarker discovery in proteomics. Given clinical data regarding a list of proteins for a set of individuals, the tackled problem is to extract a short subset of proteins the concentrations of which are an indicator of the biological status (healthy or pathological). In this paper, it is formulated as a specific instance of variable selection. The originality is that the proteins are not investigated one after the other but the best partition between discriminant and non-discriminant proteins is directly sought. In this way, correlations between the proteins are intrinsically taken into account in the decision. The developed strategy is derived in a Bayesian setting, and the decision is optimal in the sense that it minimizes a global mean error. It is finally based on the posterior probabilities of the partitions. The main difficulty is to calculate these probabilities since they are based on the so-called evidence that require marginalization of all the unknown model parameters. Two models are presented that relate the status to the protein concentrations, depending whether the latter are biomarkers or not. The first model accounts for biological variabilities by assuming that the concentrations are Gaussian distributed with a mean and a covariance matrix that depend on the status only for the biomarkers. The second one is an extension that also takes into account the technical variabilities that may significantly impact the observed concentrations. The main contributions of the paper are: (1) a new Bayesian formulation of the biomarker selection problem, (2) the closed-form expression of the posterior probabilities in the noiseless case, and (3) a suitable approximated solution in the noisy case. The methods are numerically assessed and compared to the state-of-the-art methods (t test, LASSO, Battacharyya distance, FOHSIC) on synthetic and real data from proteins quantified in human serum by mass spectrometry in selected reaction monitoring mode.

A tutorial on bridge sampling

The marginal likelihood plays an important role in many areas of Bayesian statistics such as parameter estimation, model comparison, and model averaging. In most applications, however, the marginal likelihood is not analytically tractable and must be approximated using numerical methods. Here we provide a tutorial on bridge sampling (Bennett, 1976; Meng & Wong, 1996), a reliable and relatively straightforward sampling method that allows researchers to obtain the marginal likelihood for models of varying complexity. First, we introduce bridge sampling and three related sampling methods using the beta-binomial model as a running example. We then apply bridge sampling to estimate the marginal likelihood for the Expectancy Valence (EV) model-a popular model for reinforcement learning. Our results indicate that bridge sampling provides accurate estimates for both a single participant and a hierarchical version of the EV model. We conclude that bridge sampling is an attractive method for mathematical psychologists who typically aim to approximate the marginal likelihood for a limited set of possibly high-dimensional models.

Heterogeneous reciprocal graphical models

We develop novel hierarchical reciprocal graphical models to infer gene networks from heterogeneous data. In the case of data that can be naturally divided into known groups, we propose to connect graphs by introducing a hierarchical prior across group-specific graphs, including a correlation on edge strengths across graphs. Thresholding priors are applied to induce sparsity of the estimated networks. In the case of unknown groups, we cluster subjects into subpopulations and jointly estimate cluster-specific gene networks, again using similar hierarchical priors across clusters. We illustrate the proposed approach by simulation studies and three applications with multiplatform genomic data for multiple cancers.

A hierarchical modeling approach for assessing the safety of exposure to complex antiretroviral drug regimens during pregnancy

Combination antiretroviral regimens have achieved tremendous success in reducing perinatal HIV transmission, and have become standard of care in pregnant women with HIV. However, the large variety of combination antiretroviral regimens utilized in practice raises the question of whether some of these highly potent drugs pose other risks to the pregnancy or infant. While HIV-infected pregnant women are almost always exposed to multiple antiretrovirals concurrently, standard safety screening strategies typically consider each individual antiretroviral separately, which fails to account for potential confounding due to simultaneous exposure to other antiretrovirals. In this paper, we evaluate a hierarchical modeling approach which groups antiretrovirals by drug class to screen for the safety of antiretrovirals taken during pregnancy, while still providing individual antiretroviral drug effect estimates. In simulation studies, we observed that the hierarchical approach may be advantageous as compared to considering each antiretroviral drug separately or simultaneously evaluating all antiretrovirals in a fixed effect model, particularly when there is prior evidence suggesting drugs from the same class behave similarly on the outcome. The characteristics of the hierarchical approach are illustrated in an application evaluating risk of preterm birth using a study including over 2000 pregnancies representing over 100 antiretroviral combinations, each involving up to three drug classes.

Diagnostic test accuracy of antigenaemia assay for PCR-proven cytomegalovirus infection-systematic review and meta-analysis

OBJECTIVES

We aimed to assess diagnostic test accuracy of antigenaemia assay for PCR-proven cytomegalovirus (CMV) infection.

METHODS

We systematically searched studies that provide data both on sensitivity and specificity of the CMV antigenaemia assay using the PCR as the reference standard. Adults, children, infants, individuals who were immunocompromised for any reason, symptomatic patients and asymptomatic individuals were all included. A hierarchical summary receiver operating characteristics model was used for diagnostic meta-analysis. Study quality was assessed by Revised Tool for the Quality Assessment of Diagnostic Accuracy Studies. Protocol registration identification is CRD42016035892.

RESULTS

We identified 75 eligible articles including 9058 CMV PCR-positive individuals and 22 232 PCR-negative individuals. The diagnostic odds ratio for positive antigenaemia was 30 (95% CI 24-38, I² = 28%) and the area under the hierarchical summary receiver operating characteristic curve was 0.86 (95% CI 0.83-0.88). The summary estimates of sensitivity and specificity were 0.65 (95% CI 0.59-0.70) and 0.94 (95% CI 0.93-0.95), respectively. The positive likelihood ratio of 10.9 (95% CI 8.5-14.0) suggested that a positive result from the antigenaemia assay greatly increased the probability of PCR-proven CMV infection, but a negative likelihood ratio of 0.38 (95% CI 0.32-0.44) indicated that a negative result led to a small decrease in the probability of PCR-proven CMV infection. Sensitivity and subgroup analyses replicated these results.

CONCLUSIONS

The antigenaemia assay overlooked 35% of PCR-proven CMV infections; hence, a negative result of an antigenaemia assay could not rule out a CMV infection.