Systematic reviewers planning quantitative meta-analysis usually choose between fixed effect meta-analysis, and random effects meta-analysis. An alternative method is called fixed effects (note the s in the name). This method has the unique property that the target estimand is defined by the variances of the studies found by the systematic review. This article considers each in relation to the quantitative analysis of data to be obtained by systematic review. For clarity, I refer to the traditional fixed effect method as the common effect method and the newer approach as fixed effects (plural). Case studies illustrate the post hoc nature of the fixed effects (plural) method, in which the population under study is determined by the data rather than by the protocol. Mathematical analysis shows that unlike common effect and random effects methods, the fixed effects (plural) method requires an additional, and unrealistic, assumption about the data obtained in systematic reviews. A simulation study demonstrates that confidence intervals from fixed effects (plural) meta-analysis do not account for the post hoc nature of the method. Fixed effects (plural) meta-analysis is neither a slot-in replacement for the common effect method nor for the random effects method of meta-analysis. Of the three methods considered here, the common effect method and the random effects method are potentially valid for the quantitative analysis of systematic reviews.

Cure models are a class of survival models used to analyze time-to-event data that allow the possibility that the event never occurs for a certain percentage , of the population. These methods allow for direct modelling of the cure rate and the influence of covariates on this rate. A common goal is to test whether the cure rate depends on a specific covariate or a set of covariates. The availability of methods to test the effect of covariates on the cure rate is limited in the literature. This paper proposes nonparametric hypothesis tests for the effect of covariates on the cure probability based on the martingale difference correlation. Two methods are used to approximate the null distribution of the test statistic: A permutation test and a chi-square test. The methodology is further extended to the case of two covariates using the partial martingale difference correlation. The performance of the proposed tests is evaluated through a simulation study under various scenarios, and the method is applied to a dataset on rheumatoid arthritis patients.

We investigate efficient estimation strategies for partially linear transformation models with time-dependent covariates under interval censoring. The unknown monotone function is approximated using a monotone -spline basis to enable flexible semiparametric modeling, and we develop a computationally efficient nested hybrid EM algorithm that integrates Newton's method with isotonic regression. To support large-sample inference, we propose a straightforward variance-covariance estimation procedure for the regression parameters and introduce a score test to assess the adequacy of the proportional hazards (PH) specification within the broader class of transformation models. The numerical performance of the penalized estimators is examined extensively and compared with both the time-invariant covariate model by Lu et al. and the semiparametric transformation model by Zeng et al. Finally, the proposed methodology is applied to data from the National Alzheimer's Coordinating Center (NACC) to demonstrate its practical utility in a real-world clinical setting.

Investigating changes and similarities in brain connectivity networks across task conditions is a central topic in neuroscience. We propose a novel framework for jointly estimating multiple graphical models using a hybrid Bayesian integration technique that can handle high-dimensional or large-scale data sets. This framework accommodates multiple graphical models and performs a series of conditional independence tests to infer the underlying network structures. Theoretical justification for consistency is established, and synthetic experiments demonstrate that our approach outperforms existing methods in both accuracy and robustness. We further apply this framework to an functional magnetic resonance imaging study of dynamic functional connectivity among regions of interest during an emotion-processing task. Results reveal that inter- and intra-module interactions involving the subcortical-cerebellum module are reduced during emotion processing compared to shape processing, highlighting this module's key role in emotional processing.

We revisit the widely used sample size formula for the Cox proportional hazards model, originally proposed by Schoenfeld in 1983. The classical derivation, based on the score test, evaluates the Fisher information under the null hypothesis, overlooking key conditions required for its validity. Using a Wald test framework, we demonstrate that the derivation relies on the risk set proportionality property, wherein the ratio of at-risk counts in the treatment and control arms at observed times matches the randomization ratio. This property typically holds under the null hypothesis or when event rates are low, given a sufficiently large sample size. Our analysis clarifies the asymptotic validity of the formula and shows that violations of this assumption can lead to substantial loss of efficiency, particularly under alternative hypotheses. In contrast, simulation-based approaches remain robust. A retrospective analysis of the ADAURA trial illustrates how simulation-based power analysis could have shortened the study duration compared to the formula-based approach, while still maintaining the type I error rate at the nominal level and preserving the coverage properties of the confidence interval. This work highlights the limitations of the Schoenfeld formula in realistic trial settings and recommends simulation-based methods for planning survival trials, especially when a large treatment effect is expected.

Counts in epidemiology often deviate from equidispersion and exhibit spatial, temporal, and nonlinear structure that the Poisson model cannot accommodate. We introduce a gamma-count structured additive regression model that strategically integrates penalized complexity priors in two critical aspects: (i) a principled penalized complexity prior on the dispersion parameter of the gamma-count distribution, which naturally shrinks toward the base Poisson model when the data support equidispersion, and (ii) scale-dependent penalized complexity hyperpriors on the smoothing variances for nonlinear, spatial, and temporal effects. By formulating the model within a latent Gaussian framework, we enable efficient approximate Bayesian inference through integrated nested Laplace approximations. Simulation studies across under-, equi-, and over-dispersed regimes show that the penalized complexity prior for dispersion parameter combined with scale-dependent hyperpriors yields accurate estimation of dispersion and smooth effects, favorable predictive scores, and robust inference. In empirical applications to larynx cancer mortality in Germany, COVID-19 incidence in Georgia (USA), and lung and bronchus cancer in Iowa (USA), the gamma-count structured additive regression model exhibits competitive or enhanced fit relative to Poisson and negative binomial counterparts, while elucidating interpretable nonlinear and spatial structures. This framework delivers robust, spatially resolved estimates of disease burden in the presence of non-equidispersion, thereby facilitating evidence-based resource allocation, epidemiological surveillance, and monitoring of health disparities, contributing to Sustainable Development Goals (SDGs) 3 (Good Health and Well-Being) and 10 (Reduced Inequalities). For geographically targeted analyses, it further supports informed decision-making in urban and community planning, aligning with SDG 11 (Sustainable Cities and Communities).

We develop methods to analyze clustered competing risks data when the event types are only available in a training dataset and are missing in the main study. We propose to estimate the exposure effects through the cause-specific proportional hazards frailty model where random effects are introduced into the model to account for the within-cluster correlation. We propose a weighted penalized partial likelihood method where the weights represent the probabilities of the occurrence of events, and the weights can be obtained by fitting a classification model for the event types on the training dataset. Alternatively, we propose an imputation approach where the missing event types are imputed based on the predictions from the classification model. We derive the analytical variances, and evaluate the finite sample properties of our methods in an extensive simulation study. As an illustrative example, we apply our methods to estimate the associations between tinnitus and metabolic, sensory and metabolic+sensory hearing loss in the Conservation of Hearing Study Audiology Assessment Arm.

In the context of longitudinal data regression modeling, individuals often have two or more response indicators, and these response indicators are typically correlated to some extent. Additionally, in the field of clinical medicine, the response indicators of longitudinal data are often ordinal. For the joint modeling of multivariate ordinal longitudinal data, methods based on mean regression (MR) are commonly used to study latent variables. However, for data with non-normal errors, MR methods often perform poorly. As an alternative to MR methods, composite quantile regression (CQR) can overcome the limitations of MR methods and provide more robust estimates. This article proposes a joint relative composite quantile regression method (joint relative CQR) for multivariate ordinal longitudinal data and investigates its application to a set of longitudinal medical datasets on dementia. Firstly, the joint relative CQR method for multivariate ordinal longitudinal data is constructed based on the pseudo composite asymmetric Laplace distribution (PCALD) and latent variable models. Secondly, the parameter estimation problem of the model is studied using MCMC algorithms. Finally, Monte Carlo simulations and a set of longitudinal medical datasets on dementia validate the effectiveness of the proposed model and method.

Whenever treatment is scarce, the question of how to allocate resources arises. One option is to allocate based on conditional survival benefit, defined as the contrast between an individual's expected survival with and without treatment. Estimating conditional survival benefit from observational data may present the following three challenges: (i) time-dependent confounding, arising when treatments are assigned based on longitudinal health markers which are in turn affected by past assignments; (ii) multiple time scales, namely time since becoming eligible for treatment, relevant for comparability of patients, and calendar time, representing when treatment decisions are made; and (iii) multiple versions of treatment, leading to multiple counterfactual outcomes with treatment. Building on previous work where cross-sections and inverse probability of treatment weighting were combined to estimate survival benefit on the treated population, we propose a strategy to dynamically estimate conditional survival benefit for all patients awaiting scarce treatment at any time treatment becomes available, accounting for multiple versions of treatment. After delineating identifiability assumptions, we show in a simulation study that our proposed method improves the estimation of conditional survival benefit compared to simpler methods. The proposed method is then applied to liver transplant data from the Eurotransplant region.

Generalized pairwise comparison (GPC) methods are extensions of the Mann-Whitney approach that allow comparisons of outcomes through prioritized ranking, and they have been widely applied in randomized controlled trials (RCTs). Importantly, GPC methods can be adapted to handle censored time-to-event data. GPC methods are based on assigning scores to pairs of subjects where all pairs of treatment and control subjects are evaluated: the outcome of each treatment subject is compared with each control subject. The GPC test statistic can be expressed as a treatment effect for the therapeutic intervention by measures such as the net benefit, win odds, win ratio (WR), or probability index. As the focus for this study, the WR has an alternative interpretation as the inverse of the hazard ratio under proportional hazards. However, its estimate could be biased in the presence of substantial censoring. Censoring increases the number of indeterminate treatment and control pairs, where the win or loss is undetermined due to the censored observation(s) and a definitive score cannot be assigned. We propose a novel method leveraging pseudo-observations to address the issue of uninformative pairs resulting from censoring for a time-to-event outcome. We compare the performance of our method with existing GPC methods in simulations under various censoring scenarios. For equal drop-out and administrative censoring, our method provides results that are comparable to existing GPC methods. However, for unequal drop-out, which is common in clinical trials, the performance of our approach relative to existing methods depends on the censoring proportion and distribution. The proposed approach reduced bias and root mean squared error relative to Gehan and Latta under several censoring conditions, but these improvements did not extend to gains in statistical power. Lastly, we illustrate this new GPC approach using two reconstructed RCT datasets.