False discovery in the analysis of interactions as a function of the distributional and metric properties of the outcome

Lived experience suggests substantial heterogeneity in how people react to a common stimuli. Previous work has considered several potential sources of bias and confusion in studying interactions but relatively less attention has been devoted to the nature of the outcome variable in such studies. Here, we consider power and false discovery associated with estimates of interaction parameters as a function of the distributional and metric properties of the outcome. Focusing on a variety of models for non-continuously distributed outcomes (binary, count, and ordinal outcomes), we show that power analyses need to carefully attend to specific details of a given situation and that attempts to use the linear model for recovery can be catastrophic in some settings. Focusing on transformations of normally distributed variables (i.e., censoring and departures from interval scaling) we show that linear models produce spurious interaction effects when such effects are absent from the generating model. We also provide illustrations offering geometric intuition as to why interactions can be a challenge for these incorrectly specified linear models. In light of these findings, we make two specific suggestions. First, a careful consideration of the distributional properties of the outcome variable should be a standard component in interaction studies. Second, hand-tailored power calculations should be provided; to that end, we have written software to help researchers scrutinize their ability to study interactions in given data contexts.

Advantages of visualisations to evaluate and communicate adverse event information in randomised controlled trials

Background Randomised controlled trials (RCTs) provide valuable information and inform the development of harm profiles of new treatments. Harms are typically assessed through the collection of adverse events (AEs). Despite AEs being routine outcomes collected in trials, analysis and reporting of AEs in journal articles are continually shown to be suboptimal. One key challenge is the large volume of AEs, which can make evaluation and communication problematic. Prominent practice is to report frequency tables of AEs by arm. Visual displays offer an effective solution to assess and communicate complex information; however, they are rarely used and there is a lack of practical guidance on what and how to visually display complex AE data. Methods In this article, we demonstrate the use of two plots identified to be beneficial for wide use in RCTs, since both can display multiple AEs and are suitable to display point estimates for binary, count, or time-to-event AE data: the volcano and dot plots. We compare and contrast the use of data visualisations against traditional frequency table reporting, using published AE information in two placebo-controlled trials, of remdesivir for COVID-19 and GDNF for Parkinson disease. We introduce statistical programmes for implementation in Stata. Results/case study Visualisations of AEs in the COVID-19 trial communicated a risk profile for remdesivir which differed from the main message in the published authors’ conclusion. In the Parkinson’s disease trial of GDNF, the visualisation provided immediate communication of harm signals, which had otherwise been contained within lengthy descriptive text and tables. Asymmetry in the volcano plot helped flag extreme events that were less obvious from review of the frequency table and dot plot. The dot plot allowed a more comprehensive representation by means of a more detailed summary. Conclusions Visualisations can better support investigators to assimilate large volumes of data and enable improved informal between-arm comparisons compared to tables. We endorse increased uptake for use in trial publications. Care in construction of visual displays needs to be taken as there can be potential to overemphasise treatment effects in some circumstances.

Veridical Causal Inference for Comparative Effectiveness Research Using Medical Claims

Medical insurance claims are becoming increasingly common data sources to answer a variety of questions in biomedical research. Although comprehensive in terms of longitudinal characterization of disease development and progression for a potentially large number of patients, population-based studies using these datasets require thoughtful modification to sample selection and analytic strategies, relative to other types of studies. Along with complex selection bias and missing data issues, claims-based studies are purely observational, which limits effective understanding and characterization of the treatment differences between groups being compared. All these issues contribute to a crisis in reproducibility and replication of comparative findings. This paper offers some practical guidance to the full analytical process, demonstrates methods for estimating causal treatment effects on several types of outcomes common to such studies, such as binary, count, time to event and longitudinally varying repeated measures outcomes, and aims to increase transparency and reproducibility. We provide an online version of the paper with readily implementable code for the entire analysis pipeline to serve as a guided tutorial for practitioners. The online version can be accessed at https://rydaro.github.io/. The analytic pipeline is illustrated using a sub-cohort of patients with advanced prostate cancer from the large Clinformatics TM Data Mart Database (OptumInsight, Eden Prairie, Minnesota), consisting of 73 million distinct private payer insurees from 2001-2016.

Symmetric Stacking Based Fast Binary Counter Using Modified Gate Diffusion Input (M-GDI) Technique

In this paper, modified-Gate Diffusion Input (M-GDI) based binary counter is designed using symmetric stacking method. The binary counter is designed using 3-bit stacker circuit that groups the one bit together and symmetric method is used to form 6-bit stack. The 6-bit stack is converted to binary count to produce required counters. The M-GDI is used to further reduce the transistor count than the CMOS logic transistor count. Mainly the basic gates are developed using the M-GDI technique and the basic gates are replaced in the 6:3 counters to further improve counter-performance. The proposed 6:3 binary counter has no Exclusive or gate (XOR) gates on the critical path, which leads to faster performance of the circuit. The proposed counter is faster, also consumes less power than the traditional. By using this proposed counter in Wallace multiplier, the delay and power for higher-order multipliers is reduced. This paper proposes a novel symmetric stacking based fast binary counters using the modified gate diffusion input (M-GDI) technique. This paper proposes a novel binary counter.