Bayesian modelling of effects of prenatal alcohol exposure on child cognition based on data from multiple cohorts

High levels of prenatal alcohol exposure (PAE) result in significant cognitive deficits in children, but the exact nature of the dose-response relationship is less well understood. To investigate this relationship, data were assembled from six longitudinal birth cohort studies examining the effects of PAE on cognitive outcomes from early school age through adolescence. Structural equation models (SEMs) are a natural approach to consider, because of the way they conceptualise multiple observed outcomes as relating to an underlying latent variable of interest, which can then be modelled as a function of exposure and other predictors of interest. However, conventional SEMs could not be fitted in this context because slightly different outcome measures were used in the six studies. In this paper we propose a multi-group Bayesian SEM that maps the unobserved cognition variable to a broad range of observed outcomes. The relation between these variables and PAE is then examined while controlling for potential confounders via propensity score adjustment. By examining different possible dose-response functions, the proposed framework is used to investigate whether there is a threshold PAE level that results in minimal cognitive deficit.     

Improving OpenURL Menus: User Testing of Revisions to SFX® Menus

Abstract

Twenty-one volunteers tested the usability of revisions to the Texas A&M University Libraries' SFX® OpenURL link resolver menus, including the addition of Ex Libris' new bXâ?¢ recommendation service and a plug-in which pulls additional information about the journal into the menu. The volunteers also evaluated the quality and desirability of the bX recommendations and discussed their preferences for help options and full-text format. Results of the usability testing are reported along with the resultant menu changes. This study will be of interest to librarians implementing or redesigning OpenURL menus as well as those interested in the user experience.     

Applications of Bayesian statistical methodology to clinical trial design: A case study of a phase 2 trial with an interim futility assessment in patients with knee osteoarthritis

Development of new pharmacological treatments for osteoarthritis that address unmet medical needs in a competitive market place is challenging. Bayesian approaches to trial design offer advantages in defining treatment benefits by addressing clinically relevant magnitude of effects relative to comparators and in optimizing efficiency in analysis. Such advantages are illustrated by a motivating case study, a proof of concept, and dose finding study in patients with osteoarthritis. Patients with osteoarthritis were randomized to receive placebo, celecoxib, or 1 of 4 doses of galcanezumab. Primary outcome measure was change from baseline WOMAC pain after 8 weeks of treatment. Literature review of clinical trials with targeted comparator therapies quantified treatment effects versus placebo. Two success criteria were defined: one to address superiority to placebo with adequate precision and another to ensure a clinically relevant treatment effect. Trial simulations used a Bayesian dose response and longitudinal model. An interim analysis for futility was incorporated. Simulations indicated the study had ≥85% power to detect a 14‐mm improvement and ≤1% risk for a placebo‐like drug to pass. The addition of the second success criterion substantially reduced the risk of an inadequate, weakly efficacious drug proceeding to future development. The study was terminated at the interim analysis due to inadequate analgesic efficacy. A Bayesian approach using probabilistic statements enables clear understanding of success criteria, leading to informed decisions for study conduct. Incorporating an interim analysis can effectively reduce sample size, save resources, and minimize exposure of patients to an inadequate treatment.     

Mother–Adolescent Conflict: Stability,Change, and Relations with Externalizing and Internalizing Behavior Problems

Stability and change in mother–adolescent conflict reactions (CRs) and the prediction of CRs from adolescents' earlier behavior problems (and vice versa) were examined with 131 mothers and their adolescents (63 boys). Dyads engaged in a 6‐min conflict discussion twice, 2 years apart [M age was 13 at Time 1 (T1)]. Non‐verbal expressive and verbal CRs during the conflict discussion were coded. Mothers, fathers, and teachers reported on adolescents' problem behaviors. There was inter‐individual (rank‐order) stability for adolescents' CRs whereas mothers' reactions were less stable. Mean levels of mothers' negativity, anger, and positive reactions and adolescents' negativity declined with time. Mothers’ CRs, more often than adolescents’ CRs, predicted and were predicted by adolescents’ problem behaviors in zero‐order correlations. In structural equation models with the stability of the constructs accounted for, adolescents' externalizing problems at T1 predicted higher maternal anger at Time 2 (T2). Mothers' anger and positive CRs at T1 predicted fewer T2 adolescents' internalizing problems. Stability and change in CRs are discussed.