The main aim of many epidemiological studies is to estimate the causal effect of an exposure on an outcome. When data is obtained for such studies, there is potential for some of the exposure, confounders, mediators, effect modifiers, or outcomes to be measured with error. Where we have categorical variables, we refer to this measurement error as misclassification. If measurement error and misclassification are not appropriately accounted for, erroneous study conclusions may be reached. Quantitative bias analysis (QBA) can be applied to studies that have not accounted for measurement error and be used to quantify the potential impact of measurement error, or how much measurement error would be needed to result in changes to the study conclusions. Currently, QBA methods are not implemented as a standard practise, in some part due to a lack of awareness about accessible software for the purpose. With this review, we aim to identify the available software that implements a QBA for studies with measurement error or misclassification.
pvw implements the predictive value weighting approach for adjustment for misclassification in a binary covariate in a logistic regression model, as proposed by Lyles and Lin (2010). At present the command allows the user to specify fixed values for the sensitivity and specificity, and these are allowed to vary between cases and controls. In addition, the user must specify the covariates which are to be included in a model for the misclassified version of the variable given the outcome and other variables (Z|Y,C in the notation of Lyles and Lin (2010)). From this the command calculates predictive values, which are used to perform a weighted logistic regression model. Standard errors, p-values and confidence intervals are calculated using the bootstrap. The default behavior is to resample stratified on outcome, consistent with a case-control design. Specification of the cohort option performs unstratified bootstrap resampling.
Missing outcomes are a commonly occurring problem for cluster randomised trials, which can lead to biased and inefficient inference if ignored or handled inappropriately. Two approaches for analysing such trials are cluster-level analysis and individual-level analysis. In this study, we assessed the performance of unadjusted cluster-level analysis, baseline covariate adjusted cluster-level analysis, random effects logistic regression (RELR) and generalised estimating equations (GEE) when binary outcomes are missing under a baseline covariate dependent missingness mechanism. Missing outcomes were handled using complete records analysis (CRA) and multilevel multiple imputation (MMI). We analytically show that cluster-level analyses for estimating risk ratio (RR) using complete records are valid if the true data generating model has log link and the intervention groups have the same missingness mechanism and the same covariate effect in the outcome model. We performed a simulation study considering four different scenarios, depending on whether the missingness mechanisms are the same or different between the intervention groups and whether there is an interaction between intervention group and baseline covariate in the outcome model. Based on the simulation study and analytical results, we give guidance on the conditions under which each approach is valid.
Despite growing clinical and neurobiological interest in the brain mechanisms that process emotion in music, these mechanisms remain incompletely understood. Patients with frontotemporal lobar degeneration (FTLD) frequently exhibit clinical syndromes that illustrate the effects of breakdown in emotional and social functioning. Here we investigated the neuroanatomical substrate for recognition of musical emotion in a cohort of 26 patients with FTLD (16 with behavioural variant frontotemporal dementia, bvFTD, 10 with semantic dementia, SemD) using voxel-based morphometry. On neuropsychological evaluation, patients with FTLD showed deficient recognition of canonical emotions (happiness, sadness, anger and fear) from music as well as faces and voices compared with healthy control subjects. Impaired recognition of emotions from music was specifically associated with grey matter loss in a distributed cerebral network including insula, orbitofrontal cortex, anterior cingulate and medial prefrontal cortex, anterior temporal and more posterior temporal and parietal cortices, amygdala and the subcortical mesolimbic system. This network constitutes an essential brain substrate for recognition of musical emotion that overlaps with brain regions previously implicated in coding emotional value, behavioural context, conceptual knowledge and theory of mind. Musical emotion recognition may probe the interface of these processes, delineating a profile of brain damage that is essential for the abstraction of complex social emotions.
In this chapter, the authors set out the multiple imputation (MI) procedure, initially from an intuitive standpoint. They then give a more theoretical outline of MI from a Bayesian perspective and its frequentist properties in so-called congenial settings. The authors discuss the choice of the number of imputations. They consider some simple examples, deriving the frequentist variance of the MI estimator and relating it to the estimate obtained using Rubin's MI variance formula. More general settings, where the imputation model and substantive model are uncongenial, are also discussed. The authors then discuss how to construct congenial imputation models. It is probably true that in many missing data settings, there are alternative approaches that can be taken, and these may be more efficient than MI, sometimes with a stronger justification in a strictly statistical sense.
The prelims comprise: Half-Title Page Series Page Title Page Copyright Page Contents Preface to the second edition Data acknowledgements Acknowledgements Glossary
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)